0% found this document useful (0 votes)
29 views488 pages

PowerHA SystemMirror For IBM I Cookbook

Uploaded by

zoe
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
29 views488 pages

PowerHA SystemMirror For IBM I Cookbook

Uploaded by

zoe
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 488

Front cover

PowerHA SystemMirror
for IBM i Cookbook

Take advantage of PowerHA to configure


and manage high availability

Find guidance on planning and


implementing PowerHA

Benefit from the latest PowerHA


solution enhancements

Hernando Bedoya
Abdell Ali-Darwish
Ingo Dimmer
Sabine Jordan
KyoSeok Kim
Akinori Mogi
Nandoo Neerukonda
Tomasz Piela
Marc Rauzier

ibm.com/redbooks
International Technical Support Organization

PowerHA SystemMirror for IBM i Cookbook

January 2012

SG24-7994-00
Note: Before using this information and the product it supports, read the information in “Notices” on
page ix.

First Edition (January 2012)

This edition applies to Version 7, Release 1, Modification 0 of IBM i (5770-SS1) and related licensed porgram
products.

© Copyright International Business Machines Corporation 2012. All rights reserved.


Note to U.S. Government Users Restricted Rights -- Use, duplication or disclosure restricted by GSA ADP Schedule
Contract with IBM Corp.
Contents

Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .x

IBM Redbooks promotions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
The team who wrote this book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv
Now you can become a published author, too! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvi
Comments welcome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvi
Stay connected to IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii

Part 1. Introduction and background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Chapter 1. Introduction to PowerHA SystemMirror for i . . . . . . . . . . . . . . . . . . . . . . . . . 3


1.1 IBM i Business Continuity Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1.1 PowerHA SystemMirror for i 7.1 availability capabilities . . . . . . . . . . . . . . . . . . . . . 6
1.1.2 PowerHA SystemMirror for i: 2011 enhancements . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2 Choosing a solution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Chapter 2. Implementing an independent auxiliary storage pool . . . . . . . . . . . . . . . . 13


2.1 IASP technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.1.1 Name space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.1.2 Relational Database directory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.1.3 Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.1.4 Object creation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.1.5 System-wide statement cache (SWSC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.2 Creating an IASP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3 Moving applications to an IASP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.3.1 Object considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.3.2 Accessing objects in an IASP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.3.3 Considerations for specific environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.3.4 Steps for application migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

Chapter 3. IBM i clustering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35


3.1 Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2 Cluster nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.3 Device domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.4 Cluster resource group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.5 Advanced node failure detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

Chapter 4. PowerHA architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45


4.1 PowerHA technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.1.1 Switched disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.1.2 Host-based replication (geographic mirroring) . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.1.3 Storage-based replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.1.4 Administrative domain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.2 ASP copy descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.3 ASP sessions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.3.1 Start/end . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.3.2 Changing attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

© Copyright IBM Corp. 2012. All rights reserved. iii


Part 2. Concepts and planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

Chapter 5. Geographic Mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67


5.1 Concept of geographic mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.2 Synchronous geographic mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.2.1 Synchronous geographic mirroring with synchronous mirroring mode . . . . . . . . . 74
5.2.2 Synchronous geographic mirroring with asynchronous mirroring mode . . . . . . . . 74
5.3 Asynchronous geographic mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.4 Switched disk for local HA and geographic mirroring for DR . . . . . . . . . . . . . . . . . . . . 77
5.4.1 Switched disks between logical partitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.4.2 Combining geographic mirroring and switched disks . . . . . . . . . . . . . . . . . . . . . . 78

Chapter 6. DS8000 Copy Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81


6.1 DS8000 storage concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
6.1.1 Hardware overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
6.1.2 Array site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
6.1.3 Array . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
6.1.4 Rank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
6.1.5 Extent pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.1.6 Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
6.1.7 Volume groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
6.1.8 Host connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
6.1.9 Logical subsystems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
6.2 Metro Mirror. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
6.2.1 Metro Mirror overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
6.2.2 Metro Mirror operations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
6.2.3 PPRC paths and links . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.2.4 Metro Mirror and IBM PowerHA SystemMirror for i. . . . . . . . . . . . . . . . . . . . . . . . 98
6.3 Global Mirror . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
6.3.1 Global Mirror overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
6.3.2 Global Mirror operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
6.3.3 Global Mirror and IBM PowerHA SystemMirror for i . . . . . . . . . . . . . . . . . . . . . . 104
6.4 LUN-level switching. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.5 FlashCopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
6.6 FlashCopy SE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

Chapter 7. Storwize V7000 and SAN Volume Controller Copy Services . . . . . . . . . . 113
7.1 Storwize V7000/SAN Volume Controller storage concepts. . . . . . . . . . . . . . . . . . . . . 114
7.1.1 Hardware overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
7.1.2 Storage virtualization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
7.1.3 I/O processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
7.1.4 Copy Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
7.2 Metro Mirror. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
7.2.1 Bandwidth thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
7.2.2 Remote copy relationship states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
7.2.3 Consistency groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
7.3 Global Mirror . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
7.4 FlashCopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
7.4.1 I/O indirection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
7.4.2 Background copy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
7.4.3 FlashCopy relationship states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
7.4.4 Thin-provisioned FlashCopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
7.4.5 Multi-target and reverse FlashCopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

iv PowerHA SystemMirror for IBM i Cookbook


Chapter 8. Planning for PowerHA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
8.1 Requirements for PowerHA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
8.1.1 Licensing considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
8.1.2 PRPQ ordering information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
8.1.3 Power Systems requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
8.1.4 Virtual I/O Server considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
8.1.5 Storage considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
8.2 PowerHA Copy Services Support considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
8.2.1 Global Mirror symmetrical and asymmetrical configurations. . . . . . . . . . . . . . . . 141
8.2.2 FlashCopy NoCopy/full copy/incremental/reverse . . . . . . . . . . . . . . . . . . . . . . . 143
8.3 Sizing and performance considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
8.3.1 Geographic mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
8.3.2 Virtual I/O Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
8.3.3 Copy Service bandwidth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
8.3.4 FlashCopy space-efficient relation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

Chapter 9. PowerHA user interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157


9.1 Command line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
9.1.1 The Work with Cluster (WRKCLU) command . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
9.1.2 The Configure Device ASP command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
9.1.3 Configure Geographic Mirroring command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
9.2 PowerHA GUI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
9.2.1 Accessing the PowerHA GUI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
9.2.2 Managing the cluster. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
9.3 Cluster Resource Services GUI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
9.4 High Availability Solution Manager GUI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

Chapter 10. Advanced Copy Services for PowerHA . . . . . . . . . . . . . . . . . . . . . . . . . . 167


10.1 Advanced Copy Services interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
10.2 DS storage management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
10.3 FlashCopy on Global Mirror target site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
10.4 Metro/Global Mirror and TPC-R support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
10.5 Custom programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
10.6 IBM i full-system FlashCopy replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192

Part 3. Implementation examples and best practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

Chapter 11. Creating a PowerHA base environment . . . . . . . . . . . . . . . . . . . . . . . . . . 199


11.1 Creating a cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
11.2 Setting up cluster monitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
11.3 Creating an IASP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
11.4 Setting up an administrative domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

Chapter 12. Configuring and managing Geographic Mirroring . . . . . . . . . . . . . . . . . 235


12.1 Setting up geographic mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
12.2 Managing geographic mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
12.2.1 Administrative domain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
12.2.2 Planned switchover . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
12.2.3 Deconfiguring geographic mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260

Chapter 13. Configuring and managing DS8000 Copy Services . . . . . . . . . . . . . . . . 263


13.1 Setting up IBM i DS8000 Copy Services. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
13.1.1 Configuring IBM i DS8000 Metro Mirror (GUI and CL commands) . . . . . . . . . . 264
13.1.2 Configuring IBM i DS8000 FlashCopy (CL commands) . . . . . . . . . . . . . . . . . . 294

Contents v
13.1.3 Configuring IBM i DS8000 Global Mirror (CL commands) . . . . . . . . . . . . . . . . 311
13.1.4 Configuring IBM i DS8000 LUN-level switching . . . . . . . . . . . . . . . . . . . . . . . . 317
13.2 Managing IBM i DS8000 Copy Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
13.2.1 Switchover and switchback for a Metro Mirror or Global Mirror planned outage 320
13.2.2 Using CL commands for DS8000 LUN-level switching . . . . . . . . . . . . . . . . . . . 346
13.2.3 Failing over and back for an unplanned outage . . . . . . . . . . . . . . . . . . . . . . . . 350
13.2.4 Detaching and reattaching a remote copy ASP session . . . . . . . . . . . . . . . . . . 362
13.2.5 Managing FlashCopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366

Chapter 14. Configuring and managing CSVC/V7000 Copy Services . . . . . . . . . . . . 371


14.1 SVC/V7000 Copy Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
14.1.1 Setting up an IBM i SVC/V7000 Copy Services environment . . . . . . . . . . . . . . 373
14.1.2 Configuring IBM i SVC/V7000 remote Copy Services . . . . . . . . . . . . . . . . . . . 375
14.1.3 Configuring IBM i SVC/V7000 FlashCopy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
14.2 Managing IBM i SVC/V7000 Copy Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
14.2.1 Displaying and changing a remote copy ASP session . . . . . . . . . . . . . . . . . . . 387
14.2.2 Suspending a remote copy ASP session . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
14.2.3 Detaching and reattaching a remote copy ASP session . . . . . . . . . . . . . . . . . . 388
14.2.4 Planned switchover . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
14.2.5 Unplanned failover . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391
14.2.6 Displaying and changing a FlashCopy ASP session . . . . . . . . . . . . . . . . . . . . 394
14.2.7 Reversing a FlashCopy ASP session . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
14.2.8 Using incremental FlashCopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395

Chapter 15. Best practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397


15.1 Clustering configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
15.1.1 PowerHA license consideration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
15.1.2 Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
15.1.3 Independent ASP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
15.1.4 Communications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
15.1.5 Failover wait time and default action. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
15.1.6 Administrative domain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
15.2 Journaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
15.2.1 Journal performance impact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
15.2.2 Journal management effort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
15.3 Best practices for planned site switches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
15.3.1 Regular tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
15.3.2 Check cluster and replication health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
15.3.3 Reverse replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
15.3.4 Ending applications using the IASP before a switchover . . . . . . . . . . . . . . . . . 408
15.4 Best practices for unplanned site switches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
15.5 Best practices for reducing IASP vary on times . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
15.5.1 Keeping as few DB files in SYSBAS as possible . . . . . . . . . . . . . . . . . . . . . . . 409
15.5.2 Synchronizing UIDs/GIDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
15.5.3 Access path rebuild. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
15.5.4 System Managed Access Path Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
15.6 Switching mirroring while detached. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
15.7 Resolving a cluster partition condition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
15.8 IBM i hosting environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
15.9 Upgrading IBM i and PowerHA release . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
15.9.1 Performing a rolling upgrade in a clustering environment. . . . . . . . . . . . . . . . . 414
15.9.2 Performing release upgrade while retaining current production system . . . . . . 415
15.10 Integration with BRMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415

vi PowerHA SystemMirror for IBM i Cookbook


15.10.1 Normal BRMS usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
15.10.2 Production IASP copy save-to-tapes on backup nodes . . . . . . . . . . . . . . . . . 416
15.10.3 Run production IASP copy saves to tapes after a roles switch. . . . . . . . . . . . 421
15.10.4 Specific system synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
15.10.5 User-defined IASP timestamps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
15.10.6 SYSBASE save-to-tape considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
15.11 Hardware replacement in a PowerHA environment . . . . . . . . . . . . . . . . . . . . . . . . 426
15.11.1 Server replacement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
15.11.2 Storage system replacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
15.12 Problem data collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
15.12.1 IBM i clustering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
15.12.2 PowerHA GUI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
15.12.3 The Must Gather Data Collector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
15.12.4 PowerVM Virtual I/O Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
15.12.5 DS8000 Copy Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
15.12.6 SVC/V7000 Copy Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449

Appendix A. IBM i data resilience options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451


IBM i full-system storage-based Copy Services solutions . . . . . . . . . . . . . . . . . . . . . . . . . 452
Cloning IBM i. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452
Full system FlashCopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
Full system replication by Metro Mirror or Global Mirror . . . . . . . . . . . . . . . . . . . . . . . . 453
Logical replication solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
Comparison characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457

Related publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459


IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
Online resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
Help from IBM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461

Contents vii
viii PowerHA SystemMirror for IBM i Cookbook
Notices

This information was developed for products and services offered in the U.S.A.

IBM may not offer the products, services, or features discussed in this document in other countries. Consult
your local IBM representative for information on the products and services currently available in your area. Any
reference to an IBM product, program, or service is not intended to state or imply that only that IBM product,
program, or service may be used. Any functionally equivalent product, program, or service that does not
infringe any IBM intellectual property right may be used instead. However, it is the user's responsibility to
evaluate and verify the operation of any non-IBM product, program, or service.

IBM may have patents or pending patent applications covering subject matter described in this document. The
furnishing of this document does not give you any license to these patents. You can send license inquiries, in
writing, to:
IBM Director of Licensing, IBM Corporation, North Castle Drive, Armonk, NY 10504-1785 U.S.A.

The following paragraph does not apply to the United Kingdom or any other country where such
provisions are inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATION
PROVIDES THIS PUBLICATION "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR
IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF NON-INFRINGEMENT,
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Some states do not allow disclaimer of
express or implied warranties in certain transactions, therefore, this statement may not apply to you.

This information could include technical inaccuracies or typographical errors. Changes are periodically made
to the information herein; these changes will be incorporated in new editions of the publication. IBM may make
improvements and/or changes in the product(s) and/or the program(s) described in this publication at any time
without notice.

Any references in this information to non-IBM websites are provided for convenience only and do not in any
manner serve as an endorsement of those websites. The materials at those websites are not part of the
materials for this IBM product and use of those websites is at your own risk.

IBM may use or distribute any of the information you supply in any way it believes appropriate without incurring
any obligation to you.

Information concerning non-IBM products was obtained from the suppliers of those products, their published
announcements or other publicly available sources. IBM has not tested those products and cannot confirm the
accuracy of performance, compatibility or any other claims related to non-IBM products. Questions on the
capabilities of non-IBM products should be addressed to the suppliers of those products.

This information contains examples of data and reports used in daily business operations. To illustrate them
as completely as possible, the examples include the names of individuals, companies, brands, and products.
All of these names are fictitious and any similarity to the names and addresses used by an actual business
enterprise is entirely coincidental.

COPYRIGHT LICENSE:

This information contains sample application programs in source language, which illustrate programming
techniques on various operating platforms. You may copy, modify, and distribute these sample programs in
any form without payment to IBM, for the purposes of developing, using, marketing or distributing application
programs conforming to the application programming interface for the operating platform for which the sample
programs are written. These examples have not been thoroughly tested under all conditions. IBM, therefore,
cannot guarantee or imply reliability, serviceability, or function of these programs.

© Copyright IBM Corp. 2012. All rights reserved. ix


Trademarks
IBM, the IBM logo, and ibm.com are trademarks or registered trademarks of International Business Machines
Corporation in the United States, other countries, or both. These and other IBM trademarked terms are
marked on their first occurrence in this information with the appropriate symbol (® or ™), indicating US
registered or common law trademarks owned by IBM at the time this information was published. Such
trademarks may also be registered or common law trademarks in other countries. A current list of IBM
trademarks is available on the Web at http://www.ibm.com/legal/copytrade.shtml

The following terms are trademarks of the International Business Machines Corporation in the United States,
other countries, or both:
Active Memory™ i5/OS® PowerHA®
AIX® IBM® PowerVM®
AS/400® iCluster® POWER®
DB2® iSeries® Redbooks®
Distributed Relational Database Micro-Partitioning® Redbooks (logo) ®
Architecture™ OS/400® Storwize®
DRDA® POWER Hypervisor™ System i®
DS6000™ Power Systems™ System Storage®
DS8000® POWER6+™ System x®
FlashCopy® POWER6® System z®
Global Technology Services® POWER7® Tivoli®

The following terms are trademarks of other companies:

Microsoft, Windows, and the Windows logo are trademarks of Microsoft Corporation in the United States,
other countries, or both.

Java, and all Java-based trademarks and logos are trademarks or registered trademarks of Oracle and/or its
affiliates.

SAP, and SAP logos are trademarks or registered trademarks of SAP AG in Germany and in several other
countries.

Intel, Intel logo, Intel Inside, Intel Inside logo, Intel Centrino, Intel Centrino logo, Celeron, Intel Xeon, Intel
SpeedStep, Itanium, and Pentium are trademarks or registered trademarks of Intel Corporation or its
subsidiaries in the United States and other countries.

Linux is a trademark of Linus Torvalds in the United States, other countries, or both.

Other company, product, or service names may be trademarks or service marks of others.

x PowerHA SystemMirror for IBM i Cookbook


IBM REDBOOKS PROMOTIONS

IBM Redbooks promotions

Find and read thousands of


IBM Redbooks publications
Search, bookmark, save and organize favorites
Get up-to-the-minute Redbooks news and announcements
Link to the latest Redbooks blogs and videos

Get the latest version of the Redbooks Mobile App

Download
Android
iOS

Now

Promote your business


in an IBM Redbooks
publication
®
Place a Sponsorship Promotion in an IBM
®
Redbooks publication, featuring your business
or solution with a link to your web site.

Qualified IBM Business Partners may place a full page


promotion in the most popular Redbooks publications.
Imagine the power of being seen by users who download ibm.com/Redbooks
millions of Redbooks publications each year! About Redbooks Business Partner Programs
THIS PAGE INTENTIONALLY LEFT BLANK
Preface

IBM® PowerHA® SystemMirror for i is the IBM high-availability disk-based clustering solution
for the IBM i 7.1 operating system. When combined with IBM i clustering technology,
PowerHA for i delivers a complete high-availability and disaster-recovery solution for your
business applications running in the IBM System i® environment. PowerHA for i enables you
to support high-availability capabilities with either native disk storage or IBM DS8000® or
DS6000™ storage servers or IBM Storwize® V7000 and SAN Volume Controllers.

The latest release of IBM PowerHA SystemMirror for i delivers a brand new web-based
PowerHA graphical user interface that effectively combines the solution-based and
task-based activities for your HA environment, all in a single user interface.

This IBM Redbooks® publication gives a broad understanding of PowerHA for i. This book is
divided into three major parts:
 Part 1, “Introduction and background” on page 1, provides a general introduction to
clustering technology, independent ASPs, PowerHA SystemMirror products, and PowerHA
Architecture.
 Part 2, “Concepts and planning” on page 65, describes and explains the various interfaces
that PowerHA for i has. It also explains the HA concepts as they pertain to Geographic
Mirroring, DS8000 Copy Services, Storwize V7000 and SAN Volume Controller Copy
Services, and Advanced Copy Services. It also shows you licensing and ordering
information and outlines several considerations to remember when sizing and
performance of the HA solution that you are planning to deploy using IBM PowerHA
SystemMirror for i.
 Part 3, “Implementation examples and best practices” on page 197, walks you through
several scenarios with a step-by-step approach for configuring and managing your
IBM PowerHA SystemMirror for i solution. For each scenario, we show you how to
perform a planned switchover, and we also discuss the procedures for unplanned failover.
In Chapter 15, “Best practices” on page 397, we share our recommendations and best
practices to follow in a Highly Available environment that uses various components of the
IBM PowerHA SystemMirror for i product.

If you are new to high availability, we recommend that you follow the general structure and
flow of this book so that you can start by learning the concepts and progress into the
implementation scenarios.

If you are familiar with high availability building blocks or have previous experience with any
of the solutions that we discuss in this book, then we recommend that you familiarize yourself
with the changes that are new to the latest release of the product.

Since the original writing of this book, IBM iCluster® for IBM Power Systems™, a logical
replication software product for high availability and disaster recovery that runs on the IBM i
operating system, has been acquired by Rocket Software, Inc. HA Assist, a derivative offering
of the iCluster product, was also included in the sale to Rocket Software. HA Assist for IBM i
is a specially priced and packaged offering of iCluster, which can be used in conjunction with
PowerHA SystemMirror for i for specific situations.

© Copyright IBM Corp. 2012. All rights reserved. xiii


The team who wrote this book
This book was produced by a team of specialists from around the world working at the
International Technical Support Organization, Poughkeepsie Center.

Hernando Bedoya is a Senior IT Specialist at STG Lab Services and Training, in Rochester,
Minnesota. He writes extensively and teaches IBM classes worldwide in all areas of DB2® for
i. Before joining STG Lab Services he worked in the ITSO for nine years writing multiple IBM
Redbooks publications. He also worked for IBM Colombia as an IBM AS/400® IT Specialist
doing pre-sales support for the Andean countries. He has 25 years of experience in the
computing field and has taught database classes in Colombian universities. His areas of
expertise are database technology, performance, and data warehousing. He has a master’s
degree in Computer Science from EAFIT, Colombia.

Abdel Ali-Darwish is a IT Specialist for IBM i and a Technical Solution Architect working in
the IBM Global Technology Services® support organization in Lima, Perú. He has eight years
of experience with IBM i systems. Over the years he has participated in several pre-sales and
post sales support roles including the midrange platform. He currently provides technical
pre-sales support to IBM Multicountry for Power Systems with emphasis on IBM i. His current
responsibilities include designing solutions and configurations, providing marketing
information, and serving as a subject matter expert in technical and delivery assessments.

Ingo Dimmer is an IBM Consulting IT Specialist for IBM i and a PMI Project Management
Professional working in the IBM STG Europe storage support organization in Mainz,
Germany. He has 12 years of experience in enterprise storage support from working in IBM
post-sales and pre-sales support. His areas of expertise include IBM i external disk and tape
storage solutions, I/O performance, and high availability. He has been an author of several
white paper and IBM Redbooks publications. He holds a degree in electrical engineering from
the Gerhard-Mercator University Duisburg.

Sabine Jordan is a Consulting IT Specialist working in IBM Germany. She has worked as a
Technical Specialist in the IBM i area for more than 20 years, specializing in high availability
since 2004. She has worked on IBM PowerHA System Mirror for i implementations for both
SAP and non-SAP environments using geographic mirroring and DS8000 remote copy
services. Among these implementations, she has created concepts for the design and
implemented the entire project (cluster setup, application changes), in addition to performing
customer education and testing. In addition, Sabine presents and delivers workshops (internal
and external) on IBM PowerHA System Mirror for i and high availability and disaster recovery.

KyoSeok Kim is a Senior IT Specialist for IBM i working for IBM Global Technology Services
in Korea. He has 17 years of experience with AS400, iSeries®, System i, i5/OS®, and IBM i.
He is second-level support (Top Gun) for IBM i and Power System for IBM Korea. His areas of
expertise include internal, performance, database, and IBM i problem determination. He
holds a master’s degree in computer engineering and a Bachelor of Science degree in
physics from Korea University.

Akinori Mogi is an IBM Consulting IT Specialist for IBM i and the Power Systems platform.
He works in the Technical Sales division in IBM Japan. He has over 20 years of experience in
AS/400, iSeries, System i, and IBM i pre-sales and post sales support activities as a technical
expert. His areas of expertise are high-availability solutions, virtualization, and system
performance of the IBM i environment. He has recently joined Power Systems Technical
Sales is responsible for ATS and FTSS. Akinori is also an instructor of information technology
at a university.

Nandoo Neerukonda is an IBM i Consultant specializing in performance management, query


optimization on DB2 for i, high availability, systems programming, and security. He has 15

xiv PowerHA SystemMirror for IBM i Cookbook


years of experience with IBM i and its predecessors and has worked for Countrywide
Financial (now Bank of America) and Penske Truck Leasing. He currently provides consulting
services through his own corporation, Metixis, Inc. He is a speaker at COMMON and a
co-author for two other IBM Redbooks, DB2 Universal Database for iSeries Administration:
The Graphical Way on V5R3, SG24-6092, and End to End Performance Management on IBM
i, SG24-7808. Nandoo can be contacted via email at [email protected].

Tomasz Piela is an IT Specialist working in the IBM support organization in Katowice,


Poland. He has 13 years of experience with IBM i support, consulting and solution
implementation. He holds a degree in computer science from Silesian University of
Technology in Gliwice. His areas of expertise include IBM i system performance, HA, and DR
solutions for IBM i.

Marc Rauzier is an IBM Certified IT Specialist working for IBM France in Global Technology
Services - Services Delivery organization in Lyon. He has more than 20 years of experience
in information technology, focusing on AS/400, iSeries, System i, i5/OS, and IBM i. He is
responsible for the architecture, design, and implementation of IBM Power Systems and the
IBM i based-solution for strategic outsourcing of customers in France. His areas of expertise
include IBM i, HMC, VIOS, SAN, 35xx series tape libraries, and DS8x00 series external
storage related to the IBM i environment.

Figure 1 From left to right: Ingo Dimmer, Akinori Mogi, Tomasz Piela, Sabine Jordan, Abdell
Ali-Darwish, Nandoo Neerukonda, Marc Rauzier, and KyoSeok Kim

Thanks to the following people for their contributions to this project:

Jenifer Servais
Linda Robinson
International Technical Support Organization

Troy Biesterfeld
Tom Crowley
Jenny Dervin
Steven Finnes
Amanda Fogarty
James Lembke
Curt Schemmel
Christopher Wilk
Ben Rabe
IBM Development Rochester

Laural Bauer
Selwyn Dickey
Jerry Evans

Preface xv
Brian Walker
Tim Klubertanz
John Stroh
Cindy Mestad
Steven Ransom
John Stroh
Charles Farrell
IBM STG Lab Services Rochester

Now you can become a published author, too!


Here’s an opportunity to spotlight your skills, grow your career, and become a published
author—all at the same time! Join an ITSO residency project and help write a book in your
area of expertise, while honing your experience using leading-edge technologies. Your efforts
will help to increase product acceptance and customer satisfaction, as you expand your
network of technical contacts and relationships. Residencies run from two to six weeks in
length, and you can participate either in person or as a remote resident working from your
home base.

Find out more about the residency program, browse the residency index, and apply online at:
ibm.com/redbooks/residencies.html

Comments welcome
Your comments are important to us!

We want our books to be as helpful as possible. Send us your comments about this book or
other IBM Redbooks publications in one of the following ways:
 Use the online Contact us review Redbooks form found at:
ibm.com/redbooks
 Send your comments in an email to:
[email protected]
 Mail your comments to:
IBM Corporation, International Technical Support Organization
Dept. HYTD Mail Station P099
2455 South Road
Poughkeepsie, NY 12601-5400

xvi PowerHA SystemMirror for IBM i Cookbook


Stay connected to IBM Redbooks
 Find us on Facebook:
http://www.facebook.com/IBMRedbooks
 Follow us on Twitter:
http://twitter.com/ibmredbooks
 Look for us on LinkedIn:
http://www.linkedin.com/groups?home=&gid=2130806
 Explore new Redbooks publications, residencies, and workshops with the IBM Redbooks
weekly newsletter:
https://www.redbooks.ibm.com/Redbooks.nsf/subscribe?OpenForm
 Stay current on recent Redbooks publications with RSS Feeds:
http://www.redbooks.ibm.com/rss.html

Preface xvii
xviii PowerHA SystemMirror for IBM i Cookbook
Part 1

Part 1 Introduction and


background
For years our clients have been asking when IBM will offer a hardware solution for high
availability. Over the past decade and with each subsequent release of the operating system,
we introduced the building blocks that eventually enabled us to deliver a complete integrated
IBM i solution with IBM PowerHA SystemMirror for i. We are pleased to be able to offer our
customers a complete set of IBM solution options that address their high-availability and
disaster-recovery needs.

IBM recently made significant investments to further enhance PowerHA SystemMirror for i as
its strategic high availability product for the IBM i platform.

With the October 2011 announcement, PowerHA SystemMirror for i now also supports
IBM System Storage® SAN Volume Controller and IBM Storwize V7000 for storage-based
replication solutions. A new PowerHA GUI and further enhancements for IASP-based
high-availability solutions complement the new PowerHA functionality.

This book is structured in three parts, with Part 1 covering an introduction and the architecture
of IBM PowerHA SystemMirror for i, Part 2, “Concepts and planning” on page 65, providing
concepts and information about planning, and Part 3, “Implementation examples and best
practices” on page 197, providing implementation examples and scenarios along with
best practices.

© Copyright IBM Corp. 2012. All rights reserved. 1


2 PowerHA SystemMirror for IBM i Cookbook
1

Chapter 1. Introduction to PowerHA


SystemMirror for i
This chapter provides an overview of the IBM PowerHA SystemMirror for i business continuity
solutions, including new enhancements introduced with the October 2011 announcement.

© Copyright IBM Corp. 2012. All rights reserved. 3


1.1 IBM i Business Continuity Solutions
Increasing business demands for application availability require more and more customers of
any size to look after a solution that can help eliminate planned and unplanned downtimes for
their IT services.

An unplanned outage that in duration or recovery time exceeds business expectations can
have severe implications, including unexpected loss of reputation, customer loyalty, and
revenue. Customers who did not effectively plan for the risk of an unplanned outage, never
completed their installation of an HA solution, or did not have a tested tape recovery plan in
place are especially exposed to negative business impacts.

Figure 1-1 shows an example how a high-availability solution can help you to significantly
reduce planned and unplanned application downtimes.

Why You May Need High Availability?

Timeline Sun. Mon. Mon. Tue. Wed. Wed. Fri.


>>>>>>> Night Noon Evening Morning Morning Noon Noon

unplanned
outage Failure Reload Restore Journal Partial Reconcile
(without HA) Started continues Recovery Production Journal Data

unplanned
outage Move Users < Active Production Resync HA to Prod
Repair
with HA to HA Server on HA Server > Move users back to Prod

planned
Move Users Power Off Production < Active Production Resync HA to Prod
outage
to HA Server Server for Maintenance on HA Server > Move users back to Prod
with HA

Outage
Outage comparisonwith
comparison withand
and without
without HA
HA
Unavailable Uptime

Figure 1-1 An example of why you might need high availability

4 PowerHA SystemMirror for IBM i Cookbook


To address customer needs for high availability for their IBM i environment, IBM announced
the PowerHA for i licensed program (LP) with IBM i 6.1, which with IBM i 7.1 is now called
IBM PowerHA SystemMirror for i. The IBM PowerHA SystemMirror for i solution offers a
complete end-to-end integrated clustering solution for high availability (HA) and disaster
recovery (DR). PowerHA provides a data and application resiliency solution that is an
integrated extension of IBM i operation system and storage management architecture
(Figure 1-2) and has the design objective of providing application high availability through
both planned and unplanned outages.

PowerHA SystemMirror for i (5770-HAS)

Cross Site Mirroring (XSM)


“HA Switchable Resources” - IBM i option 41
End-to-End Solution

Storage Command Line Interface


Cluster Resources in IBM i

Switched IASPs Geographic Mirroring Metro Mirror Global Mirror FlashCopy LUN Level Switching1
• Single copy of • IBM i storage mgmt • SAN hardware • SAN • SAN • Single copy of IASP
IASP is switched page level replication replication hardware hardware switched between
between LPARs • Sync. or asynchronous1 • Synchronous replication replication LPARs or server
• Internal or • Internal or external • IBM DS8000/ • Asynchronous • Point in time • IBM DS8000/
external storage storage DS6000 or • IBM DS8000/ • IBM DS8000/ DS6000
• Supports direct, VIOS & SVC/V7000 DS6000 or DS6000 or
IBM i hosted storage SVC/V7000 SVC/V7000

1
requires IBM i 7.1 or later
SVC/V7000 storage-based replication requires IBM i 7.1 with 5799-HAS PRPQ

Figure 1-2 PowerHA SystemMirror for i architecture

The key characteristic that our PowerHA customers love is that the solution is automated.
Because the data resiliency is completely managed within the IBM i storage management
architecture, there is no operator involvement, just as there is no operator involvement with
RAID 5 or disk mirroring. Geographic mirroring offers IBM i customers an IBM i-based
page-level replication solution for implementing high availability and disaster recovery with
any kind of IBM i-supported internal or external storage solution. With IBM System Storage
DS8000/DS6000 or SAN Volume Controller (SVC)/Storwize V7000 storage servers our
clients are able to exploit storage-based remote replication functions for high availability and
disaster recovery, LUN-level switching for local high availability, and FlashCopy® for reducing
save window outages by enabling the creation of a copy that is attached to a separate
partition for off-line backup to tape.

Chapter 1. Introduction to PowerHA SystemMirror for i 5


We offer IBM i customers a full menu of HA/DR solution choices (Figure 1-3).

IBM i Multi-system Data Resiliency

Switched IASP Geographic Mirroring Metro Mirror

DS8000 DS8000

LPAR-1
Production HA
LPAR-1 LPAR-1
*SYSBAS Metro
*SYSBAS IASP Mirror IASP *SYSBAS
IASP *SYSBAS *SYSBAS

LPAR-2 IASP IASP


Tape Backup

*SYSBAS

Global Mirror LUN-level Switching PowerHA SystemMirror for I combined solution with
System Storage Copy Services using
FlashCopy, LUN level switching and remote replication

DS8000 DS8000 Production HA


System i cluster services
*SYSBAS

Consistency *SYSBAS *SYSBAS Cluster Admin Domain


Group (C.G.) Mi rror
HA DR Glob al
*SYSBAS*SYSBAS
IASP
Global
*SYSBAS IASP Mirror IASP *SYSBAS
Fla
shC
IASP opy

LUN Level Switching


IASP

DS8000

Figure 1-3 PowerHA SystemMirror for i multi-system data resiliency solutions

For more detailed information about IBM PowerHA SystemMirror for i architecture and
business resiliency solutions see Chapter 4, “PowerHA architecture” on page 45.

1.1.1 PowerHA SystemMirror for i 7.1 availability capabilities


This section describes these enhancements included with PowerHA SystemMirror for i 7.1:
 PowerHA SystemMirror for i Editions
IBM PowerHA SystemMirror for i is now offered in two editions for IBM i 7.1:
– IBM PowerHA SystemMirror for i Standard Edition (5770-HAS *BASE) for local
datacenter replication only
– IBM PowerHA SystemMirror for i Enterprise Edition (5770-HAS option 1) for local or
multi-site replication
 PowerHA versioning
To use any of the new PowerHA SystemMirror for i enhancements all nodes in the cluster
need to be upgraded to IBM i 7.1.
 Clustering GUI support change
The clustering GUI plug-in for System i Navigator from High Availability Switchable
Resources licensed program (IBM i option 41) has been removed in IBM i 7.1. Clustering
HA environments can continue to be configured and managed using the PowerHA for i
licensed product (5770-HAS), CL commands, and the IBM Systems Director Navigator
for i web interface.

6 PowerHA SystemMirror for IBM i Cookbook


 N_Port ID virtualization support
Using NPIV with PowerHA does not require dedicated Fibre Channel IOAs for each
SYSBAS and IASP. Instead, virtual adapters can be defined for the partitions.
 Asynchronous Geographic Mirroring
Asynchronous geographic mirroring is a new function supported by PowerHA
SystemMirror for i Enterprise Edition with IBM i 7.1 extending the previously available
synchronous geographic mirroring option, which for performance reasons is practically
limited to metro area distances up to 40 km.
 LUN-level switching
One copy of iASP switched between two partitions/systems managed by a cluster
resource group device domain and located in IBM System Storage DS8000 or DS6000
series. An ASP session is not required for LUN-level switching, as there is no replication
for the IASP involved.
 Space-efficient FlashCopy
PowerHA for SystemMirror for i with IBM i 7.1 newly supports space-efficient FlashCopy of
the IBM System Storage DS8000 series. The IBM System Storage DS8000 series
FlashCopy SE licensed feature allows creation of space-efficient FlashCopy target
volumes that can help to reduce the required physical storage space for the FlashCopy
target volumes. These volumes are typically needed only for a limited time (such as for the
duration of a backup to tape).
 Better detection of cluster node outages
With IBM i 7.1, PowerHA SystemMirror for i now allows advanced node failure detection
by cluster nodes. This is done by registering with an HMC or Virtual I/O Server (VIOS)
management partition on IVM-managed systems. This way clustering gets notified in case
of severe partition or system failures to trigger a cluster failover event instead of causing a
cluster partition condition.
 Improved Geographic Mirroring full synchronization performance
Performance improvements have been implemented in IBM i 7.1 for geographic mirroring
full synchronization. The achievable performance improvement varies based on the IASP
data. IASPs with a large number of small objects see more benefit than those with a
smaller number of large objects.
 Cluster administrative domain enhancements
PowerHA SystemMirror for i is required to support these new administration domain
Monitored Resource Entries (MREs):
– Authorization lists (*AUTL)
– Printer device descriptions (*PRTDEV) for LAN or virtual printers
 IBM HA Assist for i
This is a new licensed product (5733-HAA) for IBM i 6.1 and later that was announced
with IBM i 6.1.1 as an extension for PowerHA only. IBM HA Assist for i is based on
iCluster code to replicate objects not supported for IASPs or by the cluster administrative
domain. It is primarily targeted at customers with existing applications that cannot be fully
migrated to an IASP environment.
 IPv6 support
PowerHA SystemMirror for i on IBM i 7.1 now fully supports IPv6 or a mix of IPv6 and
IPv4. All HA-related APIs, commands, and GUIs have been extended for field names
holding either a 32-bit IPv4 or a 128-bit IPv6 address.

Chapter 1. Introduction to PowerHA SystemMirror for i 7


 New CL commands for programming cluster automation
With PowerHA SystemMirror for i, these new CL commands are introduced in IBM i 7.1 to
better support CL programming for cluster automation management:
– RTVCLU (Retrieve Cluster)
– RTVCRG (Retrieve Cluster Resource Group)
– RTVASPCPYD (Retrieve ASP Copy Description)
– RTVASPSSN (Retrieve ASP Session)
– PRTCADMRE (Print Cluster Administrative Domain Managed Resource Entry)
For further information about the new PowerHA CL commands see the IBM i 7.1
Information Center at the following web page:
http://publib.boulder.ibm.com/infocenter/iseries/v7r1m0/topic/rbam6/HAS.htm

1.1.2 PowerHA SystemMirror for i: 2011 enhancements


These new functions are delivered with the October 2011 announcement for PowerHA
SystemMirror for i:
 New 5799-HAS PRPQ for IBM PowerHA SystemMirror for i:
– Support for managing IBM System Storage SAN Volume Controller (SVC) and IBM
Storwize V7000 Copy Services functions of FlashCopy, Metro Mirror and Global Mirror.
– IBM i CL command CFGDEVASP for configuring an independent auxiliary storage pool.
– IBM i CL command CFGGEOMIR for configuring geographic mirroring.
– New PowerHA GUI providing the facility to handle the high-availability solution starting
from a single screen. It currently supports these:
• Geographic mirroring
• Switched disk (IOA)
• DS8000/DS6000 FlashCopy
• Metro Mirror and Global Mirror
– 5799-HAS PRPQ is English only and requires 5770-HAS PTF SI44148.
 N-2 support for clustering
This allows you to skip one level of IBM i such as upgrading a V5R4M0 system within a
clustered environment directly to i 7.1 by skipping i 6.1.
 Duplicate library error handling
A message ID CPDB8EB is displayed in the QSYSOPR message queue for a library
name conflict between SYSBAS and a varying-on IASP. The vary on can be continued or
cancelled after the duplicate library issue is resolved.

8 PowerHA SystemMirror for IBM i Cookbook


1.2 Choosing a solution
Today our customers have many choices and need to evaluate which is their best
high-availability and disaster-recovery solution. We suggest that the criteria for choosing the
correct solution must be based on business needs like the recovery point objective (RPO),
recovery time objective (RTO), geographic dispersion requirements, staffing, skills, and
day-to-day administrative efforts. Figure 1-4 shows typical recovery time objectives for various
recovery solutions associated with seven tiers of business continuity (BC).

Recovery from a disk image Recovery from tape copy

PowerHA BC Tier 7 – Server or Storage replication with end-to-end automated server recovery

Metro /
Cost / Value

BC Tier 6 – Real-time continuous data replication, server or storage


Global Mirror
BC Tier 5 – Application/database level replication and transaction integrity

FlashCopy BC Tier 4 – Point in Time replication for Backup/Restore

BC Tier 3 – Tape Libraries, Electronic Vaulting


BC Tier 2 - Hot Site, Restore from Tape
BC Tier 1 – Restore
15 min. 1-4 h 4-8 h 8-12 h 12-16 h 24 h days from Tape
Recovery Time Objective
Figure 1-4 Seven tiers of disaster recovery

When you start thinking about implementing HA for your IBM i environment, consider how this
criteria apply to your situation before deciding which solution fits your needs best:
 Types of outages to be addressed
– Unplanned outages (for example, a hardware failure)
– Planned outages (for example, a software upgrade)
– Backups (for example, creating a copy of disk for an online save to tape)
– Disasters (for example, site loss, power grid outage, and so on)
 Recovery objectives
– Recovery time objective (RTO): The time to recovery from an outage
– Recovery point objective (RPO): The amount of tolerable data loss (expressed as a
time duration)

Chapter 1. Introduction to PowerHA SystemMirror for i 9


IBM i data resiliency solutions are either based on logical replication or hardware replication
(Figure 1-5). Unlike the previously mentioned PowerHA IASP hardware-based replication
solutions, logical replication solutions like IBM iCluster or high availability business partner
(HABP) replication solutions send journal entries via TPC/IP from the production system to a
backup system where the journal entries are applied to the database. Appendix A, “IBM i data
resilience options” on page 451, provides further information, including a comparison of the
IBM i data resiliency solutions.

IBM i Data Resiliency Choices


IBM HA/DR data replication options

iCluster
Asynchronous
APPLICATION APPLICATION Journal-based replication
Logical
over TCP/IP
replication IBM i IBM i Disk subsystem agnostic

Synchronous or Asynchronous
Geographic
OS-based APPLICATION APPLICATION IBM i-based replication
Mirroring
over TCP/IP
replication IBM i IBM i Disk subsystem agnostic

Synchronous or Asynchronous
Storage-based replication
Storage-based APPLICATION APPLICATION controlled by DS8000
over Fibre Channel or FCIP for
replication IBM i IBM i DS8000 and
Metro Mirror SVC/V7000

Global Mirror

Figure 1-5 IBM i HA/DR data replication options

Solution considerations
In this section we explain concepts that can help you to decide which solution your
business requires.

A storage-based synchronous replication method is one in which the application state is


directly tied to the act of data replication, just as it is when performing a write operation to
local disk. You can think of the primary and secondary IASP copies as local disk from the
application perspective. This aspect of a synchronous replication approach means that all
data written to the production IASP is also written to the backup IASP copy and the
application waits just as though it were a write to local disk. The two copies cannot be out of
sync, and also the distance between the production and backup copies, in addition to the
bandwidth of the communication link, will have an influence on application performance. The
farther apart the production and backup copies, the longer the synchronous application steps
will need to wait before proceeding to the next application step. For a longer distance
exceeding the limits of a metro area network consider using an asynchronous hardware
replication solution to prevent or minimize performance impacts for critical applications. The
huge benefit in comparison to a logical replication approach is that the two copies are

10 PowerHA SystemMirror for IBM i Cookbook


identical minus the data in the “pipe,” and therefore the secondary copy is ready to be varied
on for use on a secondary node in the cluster.

The cluster administrative domain is the PowerHA function that ensures that the set of
objects that are not in an IASP are synchronized across the nodes in the cluster. Thus, the
application has the resources that it needs to function on each node in the cluster. Clustering
solutions deployed with iASPs and using either storage-based copy services or geographic
mirroring replication require little in the way of day-to-day administrative maintenance and
were designed from the beginning for role-swap operations. We define an HA environment as
one in which the primary and secondary nodes of the cluster switch roles on a regular and
sustained basis.

Rule of thumb: If your business does not conduct regular and sustained role swaps, your
business does not have a high-availability solution deployment.

The logical replication in the IBM i environment is based on IBM i journaling technology,
including the option of remote journaling. A key characteristic of logical replication is that only
those objects that are journalled by IBM i (that is, database, IFS, data area, data queue) can
be replicated in near real time.

Synchronous remote journaling provides synchronous replication for the above-mentioned


objects, but all other objects are captured via the audit journal and then replicated to the
target system. The practical ramification of this type of replication approach is that there are
administrative activities required to ensure that the production and backup copes of data are
the same prior to a role-swap operation. Another issue is that there can be a significant
out-of-sync condition between the primary and secondary copies of data while the backup
server works to apply the data sent from the primary trying to catch up. The benefit of the
logical replication approach is that the production and backup systems can be virtually any
distance from each other and the backup copy can be used for read operations.

In addition, because one can choose to replicate a subset of objects, the bandwidth
requirements are typically not as great in comparison to a hardware-based replication approach.

Chapter 1. Introduction to PowerHA SystemMirror for i 11


12 PowerHA SystemMirror for IBM i Cookbook
2

Chapter 2. Implementing an independent


auxiliary storage pool
Independent auxiliary storage pools (IASPs) are a fundamental building block for
implementing Power HA System Mirror for IBM i. In this chapter we provide you with a brief
overview of the concept in addition to step-by-step instructions to create them. In addition, we
describe the steps necessary to move an existing application environment into an IASP and
successfully run it in this new environment.

© Copyright IBM Corp. 2012. All rights reserved. 13


2.1 IASP technology
IBM i has used the concept of single-level storage since its first release. All space available
on disks and in main memory is treated as one continuous address range where users or
programs are not actually aware of the location of the information that they want to access.

As the need to segregate groups of programs and data on the same system emerged, the
concept of pools developed and was included as part of the operating system. The pools
were referred to as auxiliary storage pools (ASPs) because they pertained to areas of
auxiliary storage (disk space). The new command structures within the operating system
used the letters ASP when referring to the auxiliary storage pools.

Enhancements to the concept of pools has led to independent auxiliary storage pools
introduced with OS/400® V5R1. These are pools that can be brought online, taken offline,
and accessed independently of the other pools on the system. They can even be logically or
physically switched between systems or logical partitions.

These are the disk pools that are available today:


 System disk pool (disk pool 1)
The system disk pool contains the load source and all configured disks that are not
assigned to any other disk pool.
 Basic disk pools (disk pool 2 to 32)
Basic disk pools can be used to separate objects from the system disk pool. For example,
you can separate your journal receivers from database objects. Basic disk pools and data
contained in them are always accessible when the system is up and running.
 Primary disk pool
This is an independent disk pool that defines a collection of directories and libraries and
might have other secondary disk pools associated with it. Primary disk pools and any
associated secondary pools can be taken offline or brought online independent of system
activity on other disk pools. Data in a primary disk pool can only be accessed by jobs on
the system if the disk pool is brought online and the job gets connected to the IASP.
 Secondary disk pool
This is an independent disk pool that defines a collection of directories and libraries and
must be associated with a primary disk pool. It is comparable to basic disk pools because
it is again used to separate specific application objects like journal receivers from your
main application objects.
 User-defined file system disk pool
This is an independent disk pool that contains only user-defined file systems (UDFSs).
 Disk pool groups
Disk pool groups consist of a primary disk pool and zero or more secondary disk pools.
Each disk pool is independent in regard to data storage, but in the disk pool group they
combine to act as one entity (for example, they are varied on and off together and
switchover is done for the entire disk pool group). Making disk pools available to the users
is accomplished by using the disk pool group name.

14 PowerHA SystemMirror for IBM i Cookbook


Figure 2-1 illustrates the hierarchy of ASPs on a system.

ASPs

System User

SYSBAS

Basic Independent
2-32 33-255

UDFS
Primary
Disk pool
group Secondary
Secondary

Figure 2-1 ASP hierarchy

There is a difference between basic ASPs and independent ASPs when it comes to data
overflow. Basic ASPs overflow and independent ASPs do not. An overflow of a basic user
ASP occurs when the ASP fills. The excess data spills into the system ASP. IASPs are
designed so that they cannot overflow. Otherwise, they would not be considered independent
or switchable. An IASP is allowed to fill up, and the application that is responsible for filling it
up simply halts. There is no automatic cancellation of the responsible job. If this job is running
from a single-threaded JOBQ, in a single-threaded subsystem all further processing is
stopped until user action is initiated.

When an IASP fills up, the job that generates the data that filled up the disk pool might not be
complete. The system generates an MCH2814 message indicating this condition. This might
have serious ramifications. Jobs that only read data are still able to work, but any job trying to
add data to the IASP is on hold. The system does not automatically cancel the offending job.
If the job is from a single-threaded JOBQ or a single-threaded subsystem, other jobs behind it
are held up until the offending job is handled. Possible scheduling impacts might occur.

2.1.1 Name space


Prior to the introduction of library-capable IASPs, any thread, including the primary or only
thread for a job, can reference the following libraries by name:
 The QTEMP library for the thread’s job, but not the QTEMP library of any other job
 All libraries within the system ASP
 All libraries within all existing basic user ASPs

Chapter 2. Implementing an independent auxiliary storage pool 15


This set of libraries formed the library name space for the thread and was the only possible
component of that name space. Although there was not a formal term for this name space
component, it is now referred to as the *SYSBAS component of the name space. It is a
required component of every name space.

With library-capable IASPs, a thread can reference, by name, all of the libraries in the IASPs
of one ASP group. This adds a second, but optional, component to the name space and is
referred to as the ASP group component of the name space. A thread that does not have an
ASP group component in its name space has its library references limited to the *SYSBAS
component. A thread with an ASP group component to its library name space can reference
libraries in both the *SYSBAS and the ASP group components of its name space.

Library names no longer must be unique on a system. However, to avoid ambiguity in name
references, library names must be unique within every possible name space. Because
*SYSBAS is a component of every name space, presence of a library name in *SYSBAS
precludes its use within any IASP. Because all libraries in all IASPs of an ASP group are part
of a name space, for which the ASP group is a component, existence of a library name within
one IASP of an ASP group precludes its use within any other IASP of the same ASP group.
Because a name space can have only one ASP group component, a library name that is not
used in *SYSBAS can be used in any or all ASP groups.

IBM i has a file interface and an SQL interface to its databases. The file interface uses the
name space to locate database objects. For compatibility, SQL maintains a catalog for each
ASP group. This catalog resides in the primary IASP of the ASP group. The catalog is built
from the objects that are in a name space that has the ASP group and *SYSBAS as its two
components. The names database and the name space are somewhat interchangeable
because they refer to the same set of database objects.

Each name space is treated as a separate relational database by SQL. It is required that all
RDBs whose data is accessible by SQL are defined in the RDB directory on the system.

Note that the name space is a thread attribute and can be specified when a job is started.
When it is referenced as a job attribute, it technically means the “thread attribute for the initial
thread of a single-threaded job.”

2.1.2 Relational Database directory


The Relational Database (RDB) directory allows an application requester (AR) to accept an
RDB name from the application and translate this name into the appropriate IP address or
host name and port. In addition, the RDB directory can also specify the user's preferred
outbound connection security mechanism. The relational database directory can also
associate an Application Requester Driver (ARD) program with an RDB name.

Each IBM i system in the distributed relational database network must have a relational
database directory configured. There is only one relational database directory on a system.
Each AR in the distributed relational database network must have an entry in its relational
database directory for its local RDB and one for each remote and local user RDB that the AR
accesses. Any system in the distributed RDB network that acts only as an application server
does not need to include the RDB names of other remote RDBs in its directory.

16 PowerHA SystemMirror for IBM i Cookbook


The RDB name assigned to the local RDB must be unique from any other RDB in the
network. Names assigned to other RDBs in the directory identify remote RDBs or local user
databases. The names of remote RDBs must match the name that an ASP uses to identify its
local system database or one of its user databases, if configured. If the local system RDB
name entry for an application server does not exist when it is needed, one is created
automatically in the directory. The name used is the current system name displayed by the
Display Network Attributes (DSPNETA) command.

Figure 2-2 gives an example of the RDB directory on a system with an IASP configured.
Notice that there is one entry present with a remote location of local. This is the RDB entry
representing the database in SYSBASE. In addition, an RDB entry gets created by the
operating system when you vary on an IASP. In our case this is the entry IASP1 with a
remote location of loopback. By default, the relational database name of an IASP is identical
to the IASP device name, but you can also choose another name here. When migrating an
application environment with a large number of accesses through the RDB name, you might
want to change the SYSBASE RBD name to a different value and use the “old” SYSBASE
RDB name as the database name for the IASP. This way, you do not have to change RDB
access to your environment.

Work with Relational Database Directory Entries

Position to . . . . . .

Type options, press Enter.


1=Add 2=Change 4=Remove 5=Display details 6=Print details

Remote
Option Entry Location Text

IASP1 LOOPBACK Entry added by system


S10C78FP *LOCAL Entry added by system

Bottom
F3=Exit F5=Refresh F6=Print list F12=Cancel F22=Display entire field
(C) COPYRIGHT IBM CORP. 1980, 2009.
Figure 2-2 Work with Relational Database Directory Entries

Although the objects in the system RDB are logically included in a user RDB, certain
dependencies between database objects have to exist within the same RDB. These include:
 A view into a schema must exist in the same RDB as its referenced tables, views,
or functions.
 An index into a schema must exist in the same RDB as its referenced table.
 A trigger or constraint into a schema must exist in the same RDB as its base table.
 Parent table and dependent table in a referential constraint both have to exist in the
same RDB.
 A table into a schema has to exist in the same RDB as any referenced distinct types.

Chapter 2. Implementing an independent auxiliary storage pool 17


Other dependencies between the objects in the system RDB and the user RDB are allowed.
For example, a procedure in a schema in a user RDB might reference objects in the system
RDB. However, operations on such an object might fail if the other RDB is not available, such
as when the underlaying IASP is varied off and then varied on to another system. A user RDB
is local to IBM i while the IASP is varied on. But as an IASP can be varied off on one server
and then varied on to another server, a user RDB might be local to a given server at one point
in time and remote at a different point in time.

2.1.3 Connections
In an SQL environment, SQL CONNECT is used to specify the correct database. To achieve
the best performance, make sure that the database being connected to corresponds with
your current library name space. You can use SETASPGRP or the INLASPGRP parameter in
your job description to achieve this. If the SQL CONNECT function is not operating within the
same library name space, the application uses Distributed Relational Database
Architecture™ (DRDA®) support, which can affect performance.

2.1.4 Object creation


While it is possible to create files, tables, and so on, into QSYS2, the corresponding library in
the independent disk pool prevents this from occurring. Most applications that create data in
QSYS2 do not realize it and fail when running in an independent disk pool.

Consider the example in Example 2-1 with library demo10 residing in an IASP and the job
running the SQL being attached to an IASP. In this example, the view ICTABLES is not built in
the current library (DEMO10) as you would expect. It is built in the library of the first table that is
mentioned, which is QSYS2 (where SYSTABLES is located). It fails when accessing the
independent disk pool because creation of objects in QSYS2XXXXX is prevented. In the
example mentioned, you must explicitly specify that you want to create the view either in
QSYS2 or in a user library in the IASP.

Example 2-1 Create view on SYSTABLES


CHGCURLIB DEMO10
create view ICTABLES(Owner, tabname, type) as select table_schema, TABLE_NAME, TABLE_TYPE
from SYSTABLES where table_name like’IC%’

2.1.5 System-wide statement cache (SWSC)


A separate SWSC is created and maintained on each IASP. Multiple sets of system
cross-reference and SQL catalog tables are defined and maintained on each IASP.

The IASP version of QSYS and QSYS2 contains cross-reference and SQL catalog tables
with merged views of all the SQL and database objects that are accessible when connected
to the IASP.

2.2 Creating an IASP


To create an IASP, you need to have at least one unconfigured disk available on your system.
The IASP can either be created using Systems Director Navigator for IBM i or using the new
CL command CFGDEVASP, available with the 5799-HAS PRPQ.

18 PowerHA SystemMirror for IBM i Cookbook


If you want to use Systems Director Navigator for IBM i to create an IASP, the following tasks
have to be performed before doing so:
1. Ensure that the IBM i user profile that you are using to access disk units has
these authorities:
– *ALLOBJ: All object authority
– *SERVICE
2. Start DST via the service panel function 21.
3. Sign on to DST using your service tools user ID and password.
4. When the Use Dedicated Service Tools (DST) display is shown, select option 5 (Work with
DST environment) and press Enter. The Work with DST Environment display is shown.
5. At the Work with DST Environment menu, select option 6 (Service tools security data).
6. At the Work with Service Tools Security Data menu, select option 6 (Change password
level). Make sure that the password level is set to Secure Hash Algorithm (SHA)
encryption or password level 2, and press F12.
7. At the Work with DST Environment display, select option 3 (Service tools user IDs) to work
with service tools user IDs.
8. Create a service tools user ID that matches the IBM i user profile and that also has the
same password in uppercase. The service tools user ID and password must match the
IBM i user profile and password of the user using IBM Systems Director Navigator for
IBM i. For example, if the user profile and password combination is BOB and my1pass,
then the DST user ID and password combination must be BOB and MY1PASS. If the
service tool user ID that you intend to use existed before changing the password level,
then you have to change its password before using IBM i Systems Director Navigator.
9. Give this service tools user ID at least these authorities:
– Disk units: operation
– Disk units: administration
10.Press Enter to enable these changes and exit DST.

Chapter 2. Implementing an independent auxiliary storage pool 19


When starting IBM i Systems Director Navigator, you can find the Configuration and Service
tasks in the main task menu (Figure 2-3).

Figure 2-3 System Director Navigator main menu

20 PowerHA SystemMirror for IBM i Cookbook


After choosing Configuration and Service, a a list of options displays (Figure 2-4):
1. To create a new independent ASP, choose Disk Pools.

Figure 2-4 System Director Navigator: Configuration and service tasks

You are presented with the current disk pool configuration (Figure 2-5).

Figure 2-5 System Director Navigator: Current disk pool configuration

Chapter 2. Implementing an independent auxiliary storage pool 21


2. From the Select Action pull-down menu, choose New Disk Pool (Figure 2-6). Click Go
after making your selection.

Figure 2-6 System Director Navigator: New disk pool

3. In the next window, choose Primary as the type of disk pool (Figure 2-7). You can then
enter the name of your IASP. The database name defaults to the name of the IASP, but
you can also enter a different name for the database. Be aware that the IASP database
name cannot be identical to the system ASP database name.

Figure 2-7 System Director Navigator: Disk pool details

22 PowerHA SystemMirror for IBM i Cookbook


4. If the Protect data in this disk pool check box is marked, the GUI will give you the ability
to select disks for the IASP that are either parity protected or that can be mirrored. If the
check box is not marked, you can only add unprotected disks to your IASP. Because we
marked the check box in our example, we can choose whether we want to add parity
protected disk or disk that can be mirrored (Figure 2-8).

Figure 2-8 System Director Navigator: Add Disk Units

5. Choosing Add Disks to be Mirrored then gives us a list of disks that can be put into the
new IASP (Figure 2-9). Make sure to select all the disks that you want to have in your
IASP and click Add.

Figure 2-9 System Director Navigator: Choose disks

Chapter 2. Implementing an independent auxiliary storage pool 23


6. The summary window gives you an overview of the new disk pool configuration that
you are about to create (Figure 2-10). Verify that is what you wanted to achieve and
click Finish.

Figure 2-10 System Director Navigator: New Disk Pool Summary

7. The IASP creation then starts (Figure 2-11). Be aware that this screen does not
automatically refresh. Refresh must be done manually.

Figure 2-11 System Director Navigator: Disk pool creation

In 11.3, “Creating an IASP” on page 215, we show how to create an IASP using the new
PowerHA GUI.

24 PowerHA SystemMirror for IBM i Cookbook


Alternatively, you can use a CL command to create your iASP. Figure 2-12 shows the
parameter required for CFGDEVASP.

Configure Device ASP (CFGDEVASP)

Type choices, press Enter.

ASP device . . . . . . . . . . . > IASP1 Name


Action . . . . . . . . . . . . . > *CREATE *CREATE, *DELETE
ASP type . . . . . . . . . . . . *PRIMARY *PRIMARY, *SECONDARY, *UDFS
Protection . . . . . . . . . . . *NO *NO, *YES
Encryption . . . . . . . . . . . *NO *NO, *YES
Disk units . . . . . . . . . . . *SELECT Name, *SELECT
+ for more values

Bottom
F1=Help F9=Calculate Selection F11=View 2 F12=Cancel

Figure 2-12 CFGDEVASP command to create IASP

Specifying *SELECT for the disk unit parameter shows the screen shown in Figure 2-13. It
provides you with a list of unconfigured disks on your system. Choose which ones you want to
add to your IASP and press Enter to create the IASP.

Select Non-Configured Disk Units

ASP device . . . . . . . . . . . . . . . : IASP1


Selected capacity . . . . . . . . . . . : 0
Selected disk units . . . . . . . . . . : 0

Type options, press Enter.


1=Select

Resource
Opt Name Serial Number Type Model Capacity Rank Eligible
1 DD007 YQKJGD54BUK6 6B22 0050 19088 002 Yes
1 DD006 YDP4V2FVUK63 6B22 0050 19088 002 Yes
1 DD008 YUNHA7W9URJL 6B22 0050 19088 002 Yes
1 DD005 YWPZGH6N8LA9 6B22 0050 19088 002 Yes

Bottom
F1=Help F9=Calculate Selection F11=View 2 F12=Cancel
Configuration of ASP device IASP1 is 8% complete.
Figure 2-13 CFGDEVASP command: Select disks to put into IASP

A message on the bottom of the screen shows the progress of the IASP creation. Your
screen is locked as long as the creation of the IASP is running. You can also follow the status
of the IASP creation using DSPASPSTS.

Chapter 2. Implementing an independent auxiliary storage pool 25


2.3 Moving applications to an IASP
Having successfully created in IASP, the next step is to look at the considerations for
migrating your application into it. We discuss what object types can be moved into an IASP
and which cannot, how you get access to objects in an IASP, and various aspects of your
application environment with an IASP.

2.3.1 Object considerations


To understand the steps necessary to move an application into an IASP we first need to have
an understanding of which objects can be located in an IASP and which objects cannot.
Table 2-1 shows which object types can be put into an IASP.

Table 2-1 Object types that can be put into IASP


*ALRTBL *FIFO *MGTCOL *QRYDFN

*BLKSF *FILE *MODULE *SBSD

*BNDDIR *FNTRSC *MSGF *SCHIDX

*CHTFMT *FNTTBL *MSGQ *SPADCT

*CHRSF *FORMDF *NODGRP *SQLPKG

*CLD *FTR *NODL *SQLUDT

*CLS *GSS *OVL *SVRPGM

*CMD *IGCDCT *OUTQ *STFM

*CRQD *JOBD *PAGDFN *SVRSTG

*CSI *JOBQ *PAGSEG *SYMLNK

*DIR *JRN *PDG *TBL

*DSTFMT *JRNRCV *PGM *USRIDX

*DTAARA *LIB *PNLGRP *USRQ

*DTADCT *LOCALE *PSFCFG *USRSPC

*DTAQ *MEDDFN *QMFORM *VLDL

*FCT *MENU *QMQRY *WSCST

For some of these object types, special considerations have to be taken into account:
 If network attributes reference the alert table, then this object must reside in the
system ASP.
 If an active subsystem references a class object, then this class object must reside in the
system ASP.
 Database files that are either multiple-system database files or that have DataLink fields
that are created as link control cannot be located in an independent disk pool. If an active
subsystem references the file object, *FILE must exist in the system disk pool, for
example, the sign-on display file.
 Subsystem descriptions can be placed into in IASP. However, if you want to start a
subsystem, its subsystem description has to be in the system ASP at that point in time.

26 PowerHA SystemMirror for IBM i Cookbook


We therefore recommended putting subsystem descriptions into an extra library located in
the system ASP.
 The same is true for job descriptions. There are a number of cases in which job
descriptions can only be used when they are located in the system ASP, for example,
when they are used for autostart jobs.
 Job queues in an IASP are operationally identical to job queues in the system ASP. Users
can manipulate the jobs (submit, hold, release, and so on) or the job queues themselves
(clear, hold, release, delete, and so on). However, the internal structures holding the real
content of the job queue still reside in the system ASP. Job entries in a job queue in an
IASP do not survive the vary off and vary on of an IASP and therefore are lost when
switching over from production to backup system.
 Journals and journal receivers must be located in the same IASP group as objects being
journaled. Journal receivers can be moved to a secondary IASP within that IASP group.
 A library that is specified by CRTSBSD SYSLIBLE() must exist in the system ASP. In
addition, libraries referenced in the system values QSYSLIBL or QUSRLIBL cannot be
located in an IASP.
 If network attributes reference a message queue, then that message queue must be
located in the system ASP.
 Programs referenced in subsystem descriptions (for example, in routing entries or
prestarted jobs) must be found in the system ASP when that subsystem is activated. The
same is true if a program is associated with the attention key.

Table 2-2 shows which object types cannot be put into an IASP.

Table 2-2 Object types that cannot be put into an IASP


*AUTHLR *CTLD *IGCTBL *NWSD

*AUTL *DDIR *IPXD *PRDAVL

*CFGL *DEVD *JOBSCD *RCT

*CNNL *DOC *LIND *SOCKET

*COSD *EDTD *MODD *SSND

*CRG *EXITRG *M36 *S36

*CSPMAP *FLR *M36CFG *USRPRF

*CSPTBL *IGCSRT *NTBD IBM libraries

If you look at the list of object types you will notice that most of them are either legacy
objects (such as folders or documents), configuration objects (such as device descriptions or
line descriptions), or objects closely related to system security (such as user profiles or
authority lists).

If you want to keep objects in the system ASP in sync between your production and your
backup system (such as user IDs and passwords) you can use the administrative domain to
help you with this task. See 3.5, “Advanced node failure detection” on page 43.

Chapter 2. Implementing an independent auxiliary storage pool 27


2.3.2 Accessing objects in an IASP
By default, each job on the system can only access objects that are stored in the system ASP.
To get access to objects in an IASP, that IASP has to be varied on. In addition, the IASP has
to be added to the namespace of the job. This can be done in these ways:
 By using the CL command SETASPGRP
 By changing the job description that the job is using

The recommended way is to use job descriptions, as that is also applicable to prestarted jobs.

Important: Never change the QDFTJOBD job description. Doing so leaves you with an
unusable system after an IPL.

2.3.3 Considerations for specific environments


Now that we know how you generally access data in an IASP, let us look at specific
environments and the impact using an IASP has on them.

Considerations for system values


Before you implement independent disk pools, examine how you use the following system
values. System values have no access to SETASPGRP. In most cases, the programs that they
reference as their values must exist in *SYSBAS. The system values that are affected by an
implementation of independent disk pools are as follows:
 QALWUSRDMN: Allows user domain objects in libraries
This value specifies which libraries can contain user domain user (*USRxxx) objects. You
can specify up to 50 individual libraries or all libraries on the system. Specifying the name
of a library makes all libraries with that name (which might exist in separate independent
auxiliary storage pools) eligible to contain user domain user objects.
 QATNPGM: Attention program
This value specifies the name and library of the attention program. This program must
exist in the system ASP or in a basic user ASP.
 QCFGMSGQ: Configuration message queue
This system value allows you to specify the default message queue that the system uses
when sending messages for lines, controllers, and devices. The message queue must
exist in the system ASP or in a basic user ASP.
 QCTLSBSD: Controlling subsystem
The controlling subsystem is the first subsystem to start after an IPL. At least one
subsystem must be active while the system is running. This is the controlling subsystem.
Other subsystems can be started and stopped. If this subsystem description cannot be
used (for example, it is damaged), the backup subsystem description QSYSSBSD in the
library QSYS can be used. A subsystem description specified as the controlling
subsystem cannot be deleted or renamed after the system is fully operational. The
subsystem description specified here must be located in the system ASP.
 QIGCCDEFNT: Double-byte code font
This value is used when transforming an SNA character string (SCS) into an Advanced
Function Printing Data Stream (AFPDS). It is also used when creating an AFPDS spooled
file with shift in/shift out (SI/SO) characters present in the data. The IGC coded font must
exist in the system ASP or in a basic user ASP. The shipped value is different for different
countries or regions.

28 PowerHA SystemMirror for IBM i Cookbook


 QINACTMSGQ: Inactive job message queue
This value specifies the action that the system takes when an interactive job has been
inactive for an interval of time (the time interval is specified by the system value
QINACTITV). The interactive job can be ended, disconnected, or message CPI1126 can
be sent to the message queue that you specify. The message queue must exist in the
system ASP or in a basic user ASP.
If the specified message queue does not exist or is damaged when the inactive timeout
interval is reached, the messages are sent to the QSYSOPR message queue. All of the
messages in the specified message queue are cleared during an IPL. If you assign a
user's message queue as QINACTMSGQ, the user loses all messages that are in the
user's message queue during each IPL.
 QPRBFTR: Problem log filter
This value specifies the name of the filter object used by the Service Activity Manager
when processing problems. The filter must exist in the system ASP or in a basic user ASP.
 QPWDVLDPGM: Password validation program
This value provides the ability for a user-written program to perform additional validation
on passwords. The program must exist in the system ASP or in a basic user ASP.
 QRMTSIGN: Remote sign-on control
This system value specifies how the system handles remote sign-on requests. The
program option allows you to specify the name of a program and library to decide which
remote sessions to allow and which user profiles to automatically sign on from which
locations. The program must exist in the system ASP or in a basic user ASP.
 QSRTSEQ: Sort sequence
This system value specifies the default sort sequence algorithm to be used by the system.
The sort sequence table must exist in the system ASP or in a basic user ASP.
 QSTRUPPGM: Startup program
This value specifies the name of the program called from an autostart job when the
controlling subsystem is started. This program performs setup functions, such as
starting subsystems and printers. The program must exist in the system ASP or in a
basic user ASP.
 QSYSLIBL: System part of the library list
When searching for an object in the library list, the libraries in the system part are
searched before any libraries in the user part are searched. The list can contain as many
as 15 library names. The libraries must exist in the system ASP or in a basic user ASP.

Chapter 2. Implementing an independent auxiliary storage pool 29


 QUPSMSGQ: Uninterruptible power supply (UPS) message queue
This value specifies the name and library of the message queue that will receive UPS
messages. It allows you to monitor the message queue and control the power down. If the
message queue is not the system operator message queue (QSYS/QSYSOPR), all UPS
messages are also sent to the system operator message queue.
 QUSRLIBL: User part of the library list
When searching for an object in the library list, the libraries in this part are searched after
the libraries in the system part and after the product library and current library entries. The
list might contain as many as 25 library names. The libraries must exist in the system ASP
or in a basic user ASP.

Considerations for network attributes


When you set up independent disk pools for the first time or move applications to
independent disk pools, consider the keywords and parameters for the system network
attributes. If the keywords and parameters highlighted in the following sections are in use,
review them for the impact that independent disk pools might have on their use. These
parameters are on the Change Network Attributes (CHGNETA) command. Some of them are on
the Retrieve Network Attributes (RTVNETA) command.

For more information about these commands, see the CL Command Finder function in the
iSeries Information Center:
http://publib.boulder.ibm.com/eserver/ibmi.html

To access this function, type CL Command Finder in the Search field:


 Alert Filters (ALRFTR)
This parameter specifies the qualified name of the alert filter used by the alert
manager when processing alerts. The alert filter must exist in the system ASP or in a basic
user ASP.
 Message Queue (MSGQ)
This parameter specifies the qualified name of the message queue where messages
received through the SNADS network are sent for users with no message queue specified
in their user profile or whose message queue is not available. The message queue must
exist in the system ASP or in a basic user ASP.
 Distributed Data Management Access (DDMACC)
This parameter specifies how the system processes distributed data management (DDM)
and DRDA requests from remote systems for access to the data resources of the system.
The DDM and DRDA connections refer to APPC conversations or active TCP/IP or
OptiConnect connections. Changes to this parameter are immediate and apply to DRDA,
DDM, or DB2 Multisystem applications. However, jobs that are currently running on the
system do not use the new value. The DDMACC value is accessed only when a job is first
started. You must specify a special value or program name that dictates how the requests
are to be handled. If a program name is specified, the program must exist in the system
ASP or in a basic user ASP.
 PC Support Access (PCSACC)
This parameter specifies how Client Access/400 requests are handled. You must specify
a special value or program name that dictates how the requests must be handled. This
permits greater control over Client Access/400 applications. Changes to this parameter
are immediate. However, jobs currently running on the system do not use the new value.
The PCSACC value is used only when a job is first started. If a program name is specified,
the program must exist in the system ASP or in a basic user ASP.

30 PowerHA SystemMirror for IBM i Cookbook


Considerations for ODBC/JDBC
ODBC and JDBC work with prestarted jobs that run under user profile QSYS. After a request
comes into the system, that request gets connected to one of the prestarted jobs. That
request has to use a user profile and password to authenticate. The prestarted job then gets
changed to run with this user profile and the environment defined in the user profile settings.
Provided that the user profile uses a job description associated with an IASP, the ODBC or
JDBC connection can therefore access objects located in the IASP.

If for any reason using a specific job description is not possible with some ODBC or JDBC
connections, then they both also provide parameters in the connection setup to explicitly
include an IASP in their namespace (Example 2-2) for ODBC.

Example 2-2 Setting access to IASP1 with ODBC


SQLAllocHandle(...);
SQLSetEnvAttr(...);
SQLAllocHandle(...);
SQLDriverConnect(hdbc, NULL,
“DSN=myDSN;DATABASE=IASP1;UID=myUID;PWD=myPWD;”,
SQL_NTS,NULL,0,NULL,SQL_DRIVER_NOPROMPT);

This gives the ODBC connection access to IASP1.

For JDBC, connecting to IASP1 is shown in Example 2-3.

Example 2-3 Setting access to IASP1 in JDBC


DriverManager.registerDriver(new AS400JDBCDriver());
AS400JDBCDataSource ds = new AS400JDBCDataSource(“SYS1”);
ds.setUser(“xxxxxxxxxx”);
ds.setPassword(“yyyyyyyyyy”);
ds.setNaming(“sql”);
ds.setDatabaseName(“IASP1”);

Considerations for FTP


FTP also uses prestarted jobs that run under user profile QTCP. After a request comes into
the system, that request gets connected to one of the prestarted jobs. Normally, this request
has to use a user profile and password to authenticate. The prestarted job then gets changed
to run with this user profile and, provided that the user profile uses a job description
associated with an IASP, can access objects located in the IASP.

If this does not work in some cases in your environment (for example, because you provide
anonymous FTP access to your system), then you can use the following command to get
access to IASP1 within the FTP job:
quote rcmd SETASPGRP IASP1

Considerations for the Integrated File System (IFS)


IFS objects are stored in a directory structure. Access to the objects is by a path that
navigates the directory structure to reach the object. An available IASP has a directory in the
root directory that has the same name as the IASP. When the IASP is available, the contents
of the IASP are mounted to the IASP directory.

Chapter 2. Implementing an independent auxiliary storage pool 31


When an IASP is not available (or before the IASP is created), it is possible to create a
directory with the name of the IASP. If there is a directory with the same name as an IASP,
when the IASP is varied on:
 The MOUNT operation will be successful if the existing directory is empty.
 The MOUNT operation will fail if there are any objects in the existing directory. The vary-on
will not fail. The first indication of failure is likely to occur when users try to access objects
in the directory. The objects will be missing or incorrect. The only indication that the
MOUNT operation failed is message CPDB414 - file system failure with a reason code 1
(The directory to be mounted over is not empty) in the joblog of the thread that performed
the vary-on operation. If the IASP environment uses IFS, each vary-on operation should
be checked to ensure that the IFS mounted properly.

Access to IFS objects is not affected by SETASPGRP or by a job description pointing to an IASP
but has to be done using the hierarchical path structure of the IFS. Therefore, if you do not
want to change hard-coded paths to IFS objects in your applications, you can create symbolic
links from the original location to the IASP location.

Tip: Check that you do not have any IFS directories with the same name as the primary
IASP that you want to create.

Considerations for DRDA


There are certain DRDA-related objects that cannot be contained in an IASP. DDM user exit
programs must reside in libraries in the system database, as must any ARD programs.

Be aware that the process of varying on an IASP causes the RDB directory to be unavailable
for a short period of time. This can cause attempts by a DRDA application requester or
application server to use the directory to be delayed or to time out.

Local user database entries in the RDB directory are added automatically the first time that
the associated databases are varied on. They are created using the *IP protocol type and
with the remote location designated as LOOPBACK. LOOPBACK indicates that the database
is on the same server as the directory.

Considerations for database access using SQL


SQL connects to the database that is set in a job’s environment. Therefore, if your job
description connects you to an IASP, any SQL statement within that job uses the IASP as
its database.

When a static SQL program is run, an access plan is created and stored with the program. If
the SQL program is located in the system ASP, a job or thread with IASP1 in its namespace
will create an access plan for data in IASP1. When a job or thread with IASP2 in its
namespace runs the same SQL program, the existing access plan is invalidated, so a new
access plan is created and stored in the program. We therefore recommend creating
separate static SQL applications in each IASP for best performance if you use more than one
IASP in your environment.

For extended dynamic SQL, create a separate SQL package in each IASP for best performance.

Query Management Query and Query Management Procedures


You can resolve the SQL objects (tables, functions, views, types) that are referenced in a
Query Management Query (*QMQRY) object. To do this, you use the RDB specified on the
RDB parameter or the RDB specified on the CONNECT/SET CONNECTION commands. This RDB

32 PowerHA SystemMirror for IBM i Cookbook


might be an IASP. The query management objects referenced must be in the current RDB
(name space).

When output from STRQMQRY is directed to an output file, Query Management ensures that
the output file is created on the RDB (name space) that was current at the time that STRQMQRY
is executed.

Considerations for DB2 Web Query


Web query only references objects in the current RDB (namespace). A *QRYDFN object
created in *SYSBAS might reference files in an IASP and vice versa. If a *QRYDFN object
created to reference objects in an IASP runs when a different IASP is set as the current RDB
(namespace), the *QRYDFN runs successfully if the new IASP contains objects with the
same name and the file formats are compatible.

Considerations for spool files


Outqueues can be put into an IASP so that spoolfiles are available on the backup system
after a switchover or failover situation. You can only do this for outqueues that are not
attached to a physical printer device, as those outqueues have to be placed in QUSRSYS. In
addition, for outqueues in an IASP, the connection between a job and its spoolfile ends when
the job itself ends. Users can therefore no longer access their old spoolfiles using WRKSPLF,
but instead have to use WRKOUTQ.

2.3.4 Steps for application migration


With the IASP created and understanding the behavior of an IASP, you can start to move
your applications to an IASP environment. The general steps to achieve this are:
1. Restore your application libraries into the IASP. Be aware that you cannot have identical
library names in the system ASP and in the IASP. If your application was installed in the
system ASP of the system that you are using to do the migration, then you have to delete
the original libraries before you can restore them into the IASP. Objects not supported in
an IASP have to be copied to a different library in the system ASP.
2. Copy IFS data that is part of your application into the new directory inside the IASP.
Delete the original IFS data and create symbolic links that redirect access from the old
directories to the new ones in the IASP.
3. If you have application objects stored in QGPL or QUSRSYS, move them to a library
inside the IASP. QGPL and QUSRSYS cannot be moved to the IASP.
4. Change your application job description to connect to the IASP using the INLASPGRP
parameter. If you are using the QDFTJOBD job description then you have to first copy it
and change the copy. Never change QDFTJOBD to access an IASP (as it is used for the
startup program during an IPL, the IPL would fail because at this point in time the IASP is
not yet available). Make sure that you set your library environment correctly in the job
description because libraries in an IASP cannot be referenced by system values
QSYSLIBL or QUSRLIBL.
5. If you created new job description, change user profiles to use them. Make sure that you
do not change your IBM i administration user profiles to use a job description with an IASP
included. If the IASP is not varied on for any reason, you cannot sign on to the system if
your profile points to a job description with an IASP.
6. Test your application.
7. Change your application environment to work with the IASP. Think of save and restore
procedures or changes in the startup program to vary on the IASP.

Chapter 2. Implementing an independent auxiliary storage pool 33


8. Decide on procedures to synchronize objects in the system ASP from the primary to the
backup system.
9. Switch your application over to the backup system and test it there.

34 PowerHA SystemMirror for IBM i Cookbook


3

Chapter 3. IBM i clustering


In this chapter, we discuss key components of IBM i cluster technology. Before exploring the
implementation of IBM PowerHA SystemMirror for i, it is important to first understand IBM i
clustering technology and capabilities.

© Copyright IBM Corp. 2012. All rights reserved. 35


3.1 Cluster
A cluster is a collection of complete systems that work together to provide a single, unified
computing resource. The cluster is managed as a single system or operating entity
(Figure 3-1). It is designed specifically to tolerate component failures and to support the
addition or subtraction of components in a way that is transparent to users. Clusters can be
simple, consisting of only two nodes, or very complex with a large number of nodes.

CLUSTER

Node 1 Node 2

Figure 3-1 A cluster consisting of two nodes

These are the major benefits that clustering offers a business:


 Simplified administration of servers by allowing a customer to manage a group of systems
as a single system or single database
 Continuous or high availability of systems, data, and applications
 Increased scalability and flexibility by allowing a customer to seamlessly add new
components as business growth develops

Attributes normally associated with the concept of clustering include these:


 Simplified single system management
 High availability and continuous availability
 High-speed interconnect communication
 Scalability and flexibility
 Workload balancing
 Single system image
 Shared resources

Note: Small outages, tolerated just a few years ago, can now mean a significant loss of
revenue and of future opportunities for a business. The most important aspect of clustering
is high availability (that is, the ability to provide businesses with resilient resources).

36 PowerHA SystemMirror for IBM i Cookbook


A cluster, device domain, device CRG, and device description are configuration objects used
to implement independent ASPs or clusters. Figure 3-2 illustrates the inter-relationship of
each IASP and cluster configuration object.

Cluster
Collection of IBM i Systems

Device Domain
Collection of cluster nodes that share resources (switchable DASD towers)
Manages assignment of common IASP ID, disk unit and virtual addresses across domain

Device CRG (Cluster Resource Group)


Cluster Control Object for a set of IASPs

Device Description
Logical control name for varying on/off
an IASP
Drives Drives
IASP
Defines a physical set of switchable
drives
Pre-requisite: cluster

Figure 3-2 Switchable IASP object relationship

Base cluster functions


Several basic IBM i cluster functions monitor the systems within the cluster to detect and
respond to potential outages in the high-availability environment. Cluster resource services
provide a set of integrated services that maintain cluster topology, perform heartbeat
monitoring, and allow creation and administration of cluster configuration and cluster
resource groups. Cluster resource services also provides reliable messaging functions that
keep track of each node in the cluster and ensure that all nodes have consistent information
about the state of cluster resources:
 Heartbeat monitoring
Heartbeat monitoring is an IBM i cluster base function that ensures that each node is
active by sending a signal from every node in the cluster to every other node in the cluster
to convey that they are still active.
 Reliable message function
The reliable message function of cluster resource services keeps track of each node in an
IBM i cluster and ensures that all nodes have consistent information about the state of
cluster resources.

Why you want clustering


The concept of high availability in the sense of disaster recovery is an important
consideration. However, disasters are not the only reason why high availability is important.

Chapter 3. IBM i clustering 37


Disasters or unplanned outages account for only 20% of all outages. The majority of outages
consist of planned ones, such as a shutdown to perform an upgrade or complete a total
system backup. A relatively straightforward action, like the backup of databases and other
objects, accounts for 50% of all planned outages.

Clusters are a very effective solution for continuous availability requirements on an IBM i
system or logical partition, providing fast recovery for the widest range of outages possible,
with minimal cost and overhead.

Some people might think that a backup of the server is not an outage. But IBM i users are not
interested in such technicalities. If access to their data on the system is not possible, the user
is most concerned about when the system is available again so that work can continue.

IBM i clustering technology offers you state-of-the-art and easy-to-deploy mechanisms to put
your business on the path to continuous availability (Figure 3-3).

Eliminate System Data and


Outages Environmental
Balanced
 PowerHA Resiliency
Systems
 Solutions
Growth
 Reduce Frequency & Application
Duration of Outages Resiliency

Administrative
Domian

IBM i Cluster
Scalability RAS - Reliability, Availability, Serviceability
Automation
Backup and Recovery Concurrent Maintenance
PowerHA for i
Selective Fault Tolerance end-to-end
Systems Management solutions
Fault Isolation
Performance Management Predictive Failure Analysis

Virtualization Service Agent – Remote Support

Figure 3-3 IBM i clustering technology state-of-the-art

3.2 Cluster nodes


A cluster node is any IBM i system or partition that is a member of a cluster. Cluster nodes
must be interconnected on an IP network. A cluster node name is an eight-character cluster
node identifier. Each node identifier is associated with one or two IP addresses that represent
the system.

Any name can be given to a node. However, we recommend that you make the node
name the same as the system name. Cluster communications that run over IP connections
provide the communications path between cluster services on each node in the cluster. The
set of cluster nodes that is configured as part of the cluster is referred to as the cluster
membership list.

A cluster consists of a minimum of two nodes. The environment can be extended to a cluster
with a maximum of 128 nodes.

38 PowerHA SystemMirror for IBM i Cookbook


A node of a cluster can fill one of three possible roles within a recovery domain. These are the
roles and associated functions:
 Primary node
– The point of access for a resilient device.
– Contains the principal copy of any replicated resource.
– The current owner of any device resource.
– All CRG objects can fail over to a backup node.
 Backup node
– Can take over the role of primary access at failure of the current primary node.
– Contains a copy of the cluster resource.
– Copies of data are kept current via replication.
 Replicate node
– Has copies of cluster resources.
– Unable to assume the role of primary or backup (typically used for functions such as
data warehousing).

3.3 Device domain


A device domain is the first of the cluster constructs to be defined when creating a switchable
IASP. It is a logical construct within Cluster Resource Services that is used to ensure that
there are no configuration conflicts that prevent a switchover or failover.

The device domain is a subset of cluster nodes.

Chapter 3. IBM i clustering 39


The set of configuration resources associated with a collection of resilient devices can be
switched across the nodes in the device domain. Resource assignments are negotiated to
ensure that no conflicts exist. The configuration resources assigned to the device domain
must be unique within the entire device domain. Therefore, even though only one node can
use a resilient device at any given time, that device can be switched to another node and
brought online (Figure 3-4).

Device Domain

Node 1 Node 2

Cluster Resources

Device Domain

Figure 3-4 Device domain

These cluster resources are negotiated across a device domain to ensure that there are
no conflicts:
 IASP number assignments
IASPs are automatically assigned a number to correlate the name of the IASP. The user
chooses the resource name. The system manages the assigned IASP numbers, which
might not be in numerical order. The order depends on a number of factors, including the
creation date and the creation of IASPs on other nodes in the device domain.
 DASD unit number assignments
To keep from conflicting with the permanently attached disk units of each node, all IASP
unit numbers begin with a four.
 Virtual address assignments
The cluster configuration determines the virtual address space required for the IASP.
Virtual address assignments (the cluster configuration) are ensured not to conflict across
all nodes in the device domain.

Note: The collection of switched disks, the independent disk pool identification, disk
unit assignments, and virtual address assignments must be unique across the entire
device domain.

40 PowerHA SystemMirror for IBM i Cookbook


3.4 Cluster resource group
A cluster resource group is an IBM i system object that is a set or grouping of cluster
resources. The cluster resource group (and replication software) is a foundation for all types
of resilience.

Resources that are available or known across multiple nodes within the cluster are called
cluster resources. A cluster resource can conceptually be any physical or logical entity (that
is, database, file, application, device). Examples of cluster resources include IBM i objects, IP
addresses, applications, and physical resources. When a cluster resource persists across an
outage, that is any single point of failure within the cluster, it is known to be a resilient
resource. As such, the resource is resilient to outages and accessible within the cluster even
if an outage occurs to the node currently hosting the resource.

Cluster nodes that are grouped together to provide availability for one or more cluster
resources are called the recovery domain for that group of cluster resources. A recovery
domain can be a subset of the nodes in a cluster, and each cluster node might participate in
multiple recovery domains. Resources that are grouped together for the purposes of recovery
action or accessibility across a recovery domain are known as a cluster resource group
(Figure 3-5).

CRG
B

CRG CRG CRG


A A B

Recovery Domain Cluster Resource Group

Cluster Resources
(for example, switched disk with IASP)

Figure 3-5 Cluster resource group

There are four cluster resource group (CRG) object types that are used with Cluster Services
at V7R1:
 Application CRG
An application CRG enables an application (program) to be restarted on either the same
node or a different node in the cluster.
 Data CRG
A data CRG enables data resiliency so that multiple copies of data can be maintained on
more than one node in a cluster.

Chapter 3. IBM i clustering 41


 Device CRG
A device CRG enables a hardware resource to be switched between systems. The device
CRG is represented by a (device) configuration object as a device type of independent
ASP (IASP).
 Peer CRG
A peer CRG is a non-switchable cluster resource group in which each IBM i node in the
recovery domain plays an equal role in the recovery of the node. The peer cluster
resource group provides peer resiliency for groups of objects or services. It is used to
represent the cluster administrative domain. It contains monitored resource entries, for
example, user profiles, network attributes, or system values that can be synchronized
between the nodes in the CRG.

The cluster resource group defines the recovery or accessibility characteristics and behavior
for that group of resources. A CRG describes a recovery domain and supplies the name of
the cluster resource group exit program that manages cluster-related events for that group.
One such event is moving the users from one node to another node in case of a failure.

Recovery domain
A recovery domain is a subset of nodes in the cluster that are grouped together in a cluster
resource group for purposes such as performing a recovery action. Each cluster resource
group has a recovery domain that is a subset of the nodes in the cluster. Here are facts about
recovery domains:
 The nodes within a recovery domain participate in any recovery actions for the resources
of the domain.
 Different CRGs might have different recovery domains.
 As a cluster goes through operational changes (for example, nodes end, nodes start,
nodes fail), the current role of a node might change. Each node has a preferred role that is
set when the CRG is created.
 A recovery domain can be a subset of the nodes in a cluster, and each cluster node might
participate in multiple recovery domains.

CRG exit programs


In IBM i high availability environments, cluster resource group exit programs are called after a
cluster-related event for a CRG occurs and responds to the event.

An exit program is called when a CRG detects certain events, such as a new node being
added to the recovery domain, or the current primary node failing. The exit program is called
with an action code indicates what the event is. Furthermore, the exit program has the
capability to indicate whether to process the event. User-defined simply means that the IBM i
cluster technology does not provide the exit program. Typically the exit program is provided
by the application or data replication provider. The exit program is the way that a CRG
communicates cluster events to the exit program provider. The exit program can perform the
appropriate action based on the event, such as allowing a resource access point to move to
another node. The exit program is optional for a resilient device CRG but is required for the
other CRG types. When a cluster resource group exit program is used, it is called on the
occurrence of cluster-wide events.

For detailed information about the cluster resource group exit programs, including what
information is passed to them for each action code, see:
http://publib.boulder.ibm.com/infocenter/iseries/v7r1m0/topic/apis/clrgexit.htm

42 PowerHA SystemMirror for IBM i Cookbook


3.5 Advanced node failure detection
There have been sudden cluster node outages, such as a main storage dump, HMC
immediate partition power-off, or a system hardware failure, which so far resulted in a
partitioned cluster. In this case, the user is alerted with a failed cluster communication
message CPFBB22 sent to QHST and an automatic failover not started message CPFBB4F
sent to the QSYSOPR message queue on the first backup node of the CRG.

With IBM i 7.1, PowerHA SystemMirror for i now allows advanced node failure detection by
cluster nodes. This is done by registering with an HMC or Virtual I/O Server (VIOS)
management partition on IVM-managed systems. This way clustering gets notified in case of
a severe partition or system failure to trigger a cluster failover event instead of causing a
cluster partition condition.

If a node detects a failure, it notifies the other nodes in the cluster via a distress message.
This triggers a failover if the failing node is the primary node of a CRG. If instead we get a
heartbeating failure, then a partition occurs. Partitions are usually the result of network
failures, and the nodes automatically merge after the problem has been fixed. However, there
are cases in which the node fails too quickly for a distress message to go out, which can
result in a “false” partition. The user can use CHGCLUNODE to change a partitioned node to a
failed node. Through the use of advanced node failure detection, the occurrences of false
partitions are reduced.

For LPAR failure conditions it is the POWER® Hypervisor™ (PHYP) that notifies the HMC
that a LPAR failed. For system failure conditions other than a sudden system power loss, it is
the flexible service processor (FSP) that notifies the HMC of the failure. The CIM server on
the HMC or VIOS can then generate an IBM Power state change CIM event for any
registered CIM clients.

Important: For systems that can be managed with a Hardware Management Console
(HMC), the Common Information Model (CIM) server runs on the HMC. The HMC affords
the most complete node failure detection because it is not part of the system and thus can
continue to operate when a system has completely failed. When using VIOS, the CIM
server must be started in the VIOS management partition by running startnetsvc
cimserver in the VIOS partition.

Whenever a cluster node is started, for each configured cluster monitor IBM i CIM client APIs
are used to subscribe for the particular power state change CIM event. The HMC CIM server
generates such a CIM event and actively sends it to any registered CIM clients (that is, there
is no heartbeat polling involved with CIM). On the IBM i cluster nodes the CIM event listener
compares the events with available information about the nodes constituting the cluster to
determine if it is relevant for the cluster to act upon. For relevant power state change CIM
events, the cluster heartbeat timer expiration is ignored (that is, IBM i clustering immediately
triggers a failover condition in this case).

Chapter 3. IBM i clustering 43


Using advanced node failure detection requires SSH and CIMOM TCP/IP communication to
be set up between the IBM i cluster nodes and the HMC or VIOS. Also, a cluster monitor
needs to be added to the IBM i cluster nodes, for example, through the new ADDCLUMON
command (Figure 3-6), which enables communication with the CIM server on the HMC or
VIOS.

Figure 3-6 IBM i ADDCLUMON command

Note: Notification of failures by an HMC or VIOS depends on a TCP/IP application server


running on the cluster node that is to receive the notification. If the application server is not
running, the advanced node failure detection is not aware of node failures. The application
server must be started and left running anytime that the cluster node is active. Use
STRTCPSVR *CIMOM CL to start the application server.

44 PowerHA SystemMirror for IBM i Cookbook


4

Chapter 4. PowerHA architecture


This chapter provides information on the basic concepts and mechanisms used in clusters
built with IBM PowerHA SystemMirror for i.

© Copyright IBM Corp. 2012. All rights reserved. 45


4.1 PowerHA technologies
In this chapter we discuss the technologies of exchanging application data between PowerHA
cluster nodes.

Clustering solutions build on PowerHA, based on the technologies of the physical exchange
of the data included on the independent ASP (iASP). These are the mechanisms used by
PowerHA described here:
 Switched disks
 Host-based replication
 Storage-based replication
 Administrative domain

Figure 4-1 and Figure 4-2 on page 47 provide information about PowerHA SystemMirror
feature availability in relation to the hardware being used.

Internal SAS/SSD DS5000 DS6000 DS8000

POWER5/6/7 POWER6/7 POWER5/6/7 POWER5/6/7


PowerHA SystemMirror 6.1 or 7.1
FlashCopy No No Yes Yes
Metro Mirror No No Yes Yes
Global Mirror No No Yes Yes
Switched IASP Yes Yes Yes Yes
LUN Level Switching No No Yes (7.1) Yes (7.1)
Geographic Mirroring Yes Yes Yes Yes
PowerHA SystemMirror 6.1 or 7.1 plus Advanced Copy Services (ACS)
FlashCopy No Yes Yes Yes
Metro Mirror No Yes Yes Yes
Global Mirror No Yes Yes Yes
LUN Level Switching No Yes Yes (6.1) Yes (6.1)
Metro/Global Mirror No No Yes Yes
External Storage Full System Copy (crash consistent copy and cloning)
FlashCopy No Yes Yes Yes
Global Mirror No Yes Yes Yes
Metro Mirror No Yes Yes Yes
Logical Replication Add-on Software
iCluster and others Yes Yes Yes Yes

Figure 4-1 IBM i native attach storage and resiliency

46 PowerHA SystemMirror for IBM i Cookbook


DS3000 DS4000 DS5000 DS6000 DS8000 XIV SVC / V7000
BladeCenter POWER6/7 POWER6/7 POWER6/7 POWER6/7 POWER6/7 POWER6/7
BladeCenter BladeCenter BladeCenter BladeCenter BladeCenter BladeCenter
PowerHA SystemMirror 6.1 or 7.1
FlashCopy No No No No Yes1 No Yes2
Metro Mirror No No No No Yes1 No Yes2
Global Mirror No No No No Yes1 No Yes2
Switched IASP No No No No No No No
LUN Level Switch No No No No Yes1 No No
Geo’mirroring Yes Yes Yes No Yes Yes Yes
PowerHA SystemMirror 6.1 or 7.1 plus Advanced Copy Services (ACS)
FlashCopy No No Yes1 No Yes1 No No
Metro Mirror No No Yes1 No Yes1 No No
Global Mirror No No Yes1 No Yes1 No No
LUN Level Switch No No No No Yes1 No No
External Storage Full System Copy (crash consistent copy and cloning)3
FlashCopy No Yes Yes Yes Yes Yes Yes
Metro Mirror No Yes Yes Yes Yes Yes Yes
Global Mirror No Yes Yes Yes Yes Yes Yes
Logical Replication Add-on Software
iCluster, et. al. Yes Yes Yes Yes Yes Yes Yes

1 Requires NPIV capable fiber channel adapter. DS5000 NPIV support requires IBM i 7.1 TR2
2 Requires IBM i 7.1 TR3 and PRPQ 5799-HAS
3 See Redbook “IBM i and Midrange External Storage SG247668” and ”IBM i Virtualization and DS4000 Read-me First”

Figure 4-2 IBM i PowerVM® VIOS storage and resiliency

4.1.1 Switched disks


Switched disk is a solution based on the concept of sharing the same pool of disks (iASP)
between different systems.

Chapter 4. PowerHA architecture 47


To illustrate this idea we modified the generic cluster model to the one shown in Figure 4-3.

SYSBAS SYSBAS

Switched
Disk

Node1 Node2

IASP

Site A
Figure 4-3 Model of switched disk solution

These are the types of switched disk solutions:


 Switching disks between two separate systems using the commonHSL loop
 Switching disks between two LPARs in the same system

Note: POWER6® server is the last one that uses the HSL loop and the last server where
switching disks build on this technology is supported.

All solutions are the same from the logical standpoint.

Using the switched disk solution we have a single IASP that can be owned by one or the
other node in the cluster at a given time. During the switch process the switchable IASP
changes its owner and all the data on it is accessible by the other node.

These are the advantages of this mechanism:


 Simplicity
 Low cost

These are the disadvantages of this mechanism:


 HA within single POWER system: When using the HSL loop or switching between LPARs
in the same system. When using switched LUNs it is possible to switch the disks between
different systems.
 Low resiliency to data loss: You need to implement disk redundancy at the storage level.

Switched disks are not advised to be a single HA solution for a production system. However
they can be a complementary element of more sophisticated solutions (for example, a 3-node
solution with storage-based replication).

Note: The data on the switched disks is in one copy, so it should be protected on the disk
level (for example, by one of the RAID mechanisms).

48 PowerHA SystemMirror for IBM i Cookbook


4.1.2 Host-based replication (geographic mirroring)
Host-based replication is used to keep consistent copy of an iASP assigned to one of the
cluster nodes onto another iASP assigned to another node. The processes that keep the
iASP in synchronization run on the nodes that own the iASPs. In most cases the TCP/IP
network is being used for sending the changes to backup iASP.

In the case of the PowerHA System Mirror for i, the host-based replication feature is called
geographic mirroring.

Figure 4-4 shows the idea of host-based replication.

Node1 Cluster – Device Domain - CRG Node2

PRODUCTION MIRROR
copy copy

SYSBAS SYSBAS

/ Network
TCP/IP

IASP IASP
source Geographic target
Mirroring

Site A Site B
Figure 4-4 Host-based replication in PowerHA: Geographic mirroring

Note: Notice that iASPs can be built on internal or external disk drives.

Geographic mirroring guarantees that the changes to the replicated iASP will be applied to
the target copy in the original order, because in case of failure the target copy will be usable
and accessible by the node that owns it.

Note: It must be stressed that the commitment control must be in place to guarantee the
database consistency in case of unexpected failure.

Geographic mirroring introduces high-availability protection in case of the server failure, and
in case the servers are in a different location, it allows protection in case of a site outage.

The copy owned by the primary node is the production copy and the copy owned by the
backup system at the other site is the mirror copy. Users and applications can only access
the independent disk pool on the primary node, the node that owns the production copy.
Changes that are made to the production copy (source system) are guaranteed by the
geographic mirroring functionality to be made in the same order on the mirror copy
(target system).

Chapter 4. PowerHA architecture 49


Geographic mirroring allows for the production and mirrored copies to be in the same site for
high-availability protection in the event of server failure. It is also possible to separate the two
systems geographically for disaster recovery protection in the event of a site-wide outage,
provided that the communication link between the two sites is fast enough. In the case of
using synchronous delivery mode, communication speed and throughput have an impact on
application response time on the production system. This is due to the fact that the production
system waits until a write operation has at least reached main memory on the backup system
and the backup system has sent a confirmation back to the production system before a local
write to the iASP of the production system is considered finished. PowerHA for i Enterprise
additionally offers the asynchronous mode of delivery of the changes to the remote IASP. In
this case the network delivery time does not affect the application performance. However, the
changes are not guaranteed to be delivered to the backup system in case of the unexpected
failure of the primary node.

Synchronous and asynchronous modes of operation


When using geographic mirroring replication for iASP, there are two parameters that are
related to synchronism in the data replication:
 Transmission delivery
This parameter is available when using PowerHA for System i Enterprise. It allows you to
decide whether the data will be delivered to the adjacent node in synchronous or
asynchronous mode. Synchronous mode of delivery requires that when the data is
received by the backup node it sends the confirmation to the primary, and after the
confirmation is received the operation is confirmed. In asynchronous mode of delivery the
data is sent to the backup node and this completes the operation.
 Mirroring mode
Mirroring mode of operation decides whether the data needs to be written to the disk on
the backup node in order to have a completed status on the primary node. This is
synchronous delivery mode. If the mirroring mode is asynchronous, the operation is
completed on the primary node when it is received by the backup node. It does not have to
be put on disk before it is completed on the primary.

50 PowerHA SystemMirror for IBM i Cookbook


The relationship between the operations in synchronous and asynchronous modes of
operations are shown in Figure 4-5, and the order of the operations for each type of delivery
is shown in Table 4-1.

Node 1
PRODUCTION
Copy

Node 2
2 MIRROR
Main TCP/IP Network Copy
Memory 3

4 1
Main
Memory
IOA cache

IASP
source
5

IOA cache

IASP
target

Site A Site B
Figure 4-5 Geographic mirroring operations

Table 4-1 Geographic mirroring operations relationship


Transmission Mirroring mode Local operations Remote operations
delivery

*ASYNC *ASYNC 1,4 2

*SYNC *ASYNC 1,4 2,3

*SYNC *SYNC 1,4 2,5,3

More details about the geographic mirroring can be found in Chapter 5, “Geographic
Mirroring” on page 67.

Note: When using host-based replication in synchronous mode of operation, the network
latency caused by the distance between the hosts can cause performance degradation of
the applications using the replicated iASP. The mirroring mode can severely affect
performance when the target system’s disk subsystem has lower performance than the
source system.

Advantages of host-based replication


Geographic mirroring is the only physical replication that can be used for the IASPs that are
built on the internal disks subsystems. It has a relatively low cost of implementation due to the

Chapter 4. PowerHA architecture 51


fact that it is a software-based solution and does not require external storage devices to be in
place for replication.

Disadvantages
The disadvantage of this solution can be the fact that it uses the host resources, such as CPU
and memory. That might affect performance of other processes running on the system.

4.1.3 Storage-based replication


In contrast to host-based replication (geographic mirror), storage-based replication is done at
the storage subsystem devices level. The types of the storage level replication supported by
IBM PowerHA SystemMirror for i are as follows:
 Metro Mirror
 Global Mirror
 FlashCopy
 LUN-level switching

Metro Mirror (synchronous)


Metro Mirror is a replication solution based on the synchronous data replication between
external IBM Storage Systems connected to IBM i. This means that all the operations are
acknowledged to the source system when they are completed on both storage susbsystems.
Data between storage systems is replicated over the Storage Area Network (SAN), either using
FC or FC over IP. Synchronous replication in case of Metro Mirror means that the source and a
copy of the IASP is in a consistent state when an unexpected failure occurs on any of the
cluster nodes.

When using a Metro Mirror replication with PowerHA System Mirror, each node has a locally
attached external IBM Storage System, and the replicated IASP must be located in it.
Figure 4-6 shows the most common configuration. The System ASP can reside on internal or
external disks.

SITE 1

IBM Power Systems IBM Power Systems

Production Backup
LPAR LPAR

IASP IASP
SRC Metro Mirror CPY

IBM System Storage IBM System Storage

Figure 4-6 Metro Mirror architecture

52 PowerHA SystemMirror for IBM i Cookbook


Because of the synchronous mode of operation, the distance between the replication sites is
limited to metro distances, which is about 30 kilometers. For more information about Metro
Mirror see 6.2, “Metro Mirror” on page 93.

In addition to IBM TotalStorage devices you can build a Metro Mirror solution using IBM
Storwize V7000 and SAN Volume Controller. For more details about this type of
implementation see 7.2, “Metro Mirror” on page 119.

Global Mirror (asynchronous)


Global Mirror is an asynchronous replication between local and remote IBM Storage
Systems. Similar to Metro Mirror, this replication is also controlled by storage systems and
data is being replicated on the disk level. Due to asynchronous replication, this solution can
be used for replication between storage systems located a long distance from each other
because the delay introduced by the remote write is not affecting the local operations.

Global Mirror is based on these copy services functions of a Storage system:


 Global Copy
 FlashCopy
 FlashCopy consistency group

The use of FlashCopy and the FlashCopy consistency group with Global Copy allows you to
maintain a consistent copy of the IASP in the backup site. Figure 4-7 shows the general
architecture for the Global Mirror replication.

SITE 1 SITE 2

IBM Power Systems IBM Power Systems

Production Backup
LPAR LPAR

IASP IASP
SRC Global Copy CPY1
FlashCopy

IASP
CPY2

IBM System Storage IBM System Storage

Figure 4-7 Global Mirror architecture

How it works
Global Mirror, as a long-distance remote copy solution, is based on an efficient combination
of Global Copy and FlashCopy functions. It is the Storage system microcode that provides,
from the user perspective, a transparent and autonomic mechanism to intelligently utilize
Global Copy in conjunction with certain FlashCopy operations to attain consistent data at the
remote site.

Chapter 4. PowerHA architecture 53


For more details about the Global Mirror solution see 6.3, “Global Mirror” on page 98, which
describes the Global Mirror for DS8000 Copy Services.

Global Mirror can be used in solutions based on IBM Storwize V7000 and SAN Volume
Controller. See 7.3, “Global Mirror” on page 123.

LUN-level switching
LUN-level switching is a new function provided by PowerHA SystemMirror for i. It allows you
to switch a set of LUNs (a volume group) between systems. The idea of this solution is similar
to the switched disks described in 4.1.1, “Switched disks” on page 47. For more details about
this solution, see 6.4, “LUN-level switching” on page 105.

FlashCopy
FlashCopy allows you to create a point-in-time copy of the logical volume. By doing a
FlashCopy, a relationship is established between a source volume and a target volume. Both
are considered to form a FlashCopy pair. As a result of the FlashCopy, either all physical
blocks from the source volume are copied (when using the copy option) to the target volume,
or when using the nocopy option. Only those parts are copied that are changing in the source
data since the FlashCopy was established. The target volume needs to be the same size or
bigger than the source volume whenever FlashCopy is used to flash an entire volume.
Figure 4-8 shows an example of the architecture for the FlashCopy. In this example, when the
production IASP FlashCopy is established, the copy can be accessed by another LPAR. The
copy of the IASP is read and write capable, so you can use it for backup or testing purposes.

IBM Power Systems IBM System Storage

Production
LPAR IASP
SRC
FlashCopy

Backup
LPAR IASP
CPY

Figure 4-8 FlashCopy architecture example

Within PowerHA for i, the classic FlashCopy and FlashCopy SE are supported. When using
classic FlashCopy, the target volume size that is reported to the system must be allocated in
the TotalStorage device.

When using FlashCopy SE (Space Efficient) you can create virtual volumes that are space
efficient and use them as a target for the FlashCopy operation. The space efficient volumes
are volumes that have a defined virtual size, but the physical storage is not physically
allocated to them.

Typically, large applications such as databases have their data spread across several
volumes, and their volumes should all be FlashCopied at exactly the same point-in-time.
FlashCopy offers consistency groups, which allows multiple volumes to be FlashCopied at
exactly the same instance.

54 PowerHA SystemMirror for IBM i Cookbook


4.1.4 Administrative domain
A cluster administrative domain provides a mechanism for maintaining a consistent
operational environment across cluster nodes within an IBM i high availability environment. A
cluster administrative domain ensures that highly available applications and data behave as
expected when switched to or failed over to backup nodes.

There are often configuration parameters or data associated with applications and application
data, which are known collectively as the operational environment for the application.
Examples of this type of data include user profiles used for accessing the application or its
data, or system environment variables that control the behavior of the application. With a
high-availability environment, the operational environment needs to be the same on every
system where the application can run, or where the application data resides. When a change
is made to one or more configuration parameters or data on one system, the same change
needs to be made on all systems. A cluster administrative domain lets you identify resources
that need to be maintained consistently across the systems in an IBM i high availability
environment. The cluster administrative domain then monitors for changes to these resources
and synchronizes any changes across the active domain.

When a cluster administrative domain is created, the system creates a peer CRG with the
same name. The nodes that make up the cluster administrative domain are defined by the
CRGs recovery domain. Node membership of the cluster administrative domain can be
modified by adding and removing nodes from the recovery domain using these commands:
 Add Admin Domain Node Entry (ADDCADNODE).
 Remove Admin Domain Node Entry (RMVCADNODE).
 Work with Cluster (WRKCLU).

Each cluster node can be defined in only one cluster administrative domain within the cluster.

Note: To work with cluster CL commands or the Cluster Resource Services graphical
interface, you must have IBM PowerHA SystemMirror for i licensed program installed.

After the cluster administrative domain is created, it can be managed with CL commands or
the Cluster Resource Services graphical interface in IBM Systems Director Navigator for i.

The type of objects that can be managed in a cluster administrative domain, also known as
monitored resources, have been enhanced in IBM i 7.1. See Table 4-1 on page 51:

Table 4-2 Monitored resource entry type support


Object or attribute description Type

Authorization lists (*) *AUTL

Classes *CLS

Ethernet line descriptions *ETHLIN

Independent disk pools device descriptions *ASPDEV

Job descriptions *JOBD

Network attributes *NETA

Network server configuration for connection security *NWSCFG

Network server configuration for remote systems *NWSCFG

Network server configurations for service processors *NWSCFG

Chapter 4. PowerHA architecture 55


Object or attribute description Type

Network server descriptions for iSCSI connections *NWSD

Network server descriptions for integrated network servers *NWSD

Network server storage spaces *NWSSTG

Network server host adapter device descriptions *NWSHDEV

Optical device descriptions *OPTDEV

Printer device descriptions for LAN connections* *PRTDEV

Printer device descriptions for virtual connections* *PRTDEV

Subsystem descriptions *SBSD

System environment variables *ENVVAR

System values *SYSVAL

Tape device descriptions *TAPDEV

Token-ring line descriptions *TRNLIN

TCP/IP attributes *TCPA

User profiles *USRPRF


* Available from IBM i 7.1. PowerHA SystemMirror for i is required to support these new administration domain
monitored resource entries.

Monitored resources
A monitored resource is a system resource that is managed by a cluster administrative
domain. Changes made to a monitored resource are synchronized across nodes in the
cluster administrative domain and applied to the resource on each active node. Monitored
resources can be system objects such as user profiles or job descriptions. A monitored
resource can also be a system resource not represented by a system object, such as a single
system value or a system environment variable. These monitored resources are represented
in the cluster administrative domain as monitored resource entries (MREs).

A cluster administrative domain supports monitored resources with simple attributes and
compound attributes. A compound attribute differs from a simple attribute in that it contains
zero or more values, while a simple attribute contains a single value. Subsystem Descriptions
(*SBSD) and Network Server Descriptions (*NWSD) are examples of monitored resources
that contain compound attributes.

For MREs to be added, the resource must exist on the node from which the MREs are added.
If the resource does not exist on every node in the administrative domain, the monitored
resource is created. If a node is later added to the cluster administrative domain, the
monitored resource is created. MREs can only be added to the cluster administrative domain
if all nodes in the domain are active and participating in the group. MREs cannot be added in
the cluster administrative domain if the domain has a status of Start of changePartitionedEnd
of change.

You can add the MRE using ADDCADMRE or with the PowerHA GUI. The PowerHA GUI allows
you to select the MRE’s monitored parameters from a list, while in the command you need to
specify them.

To remove MRE from the administrative domain, you can use RMVCADMRE or the
PowerHA GUI.

56 PowerHA SystemMirror for IBM i Cookbook


Determine the status of the cluster administrative domain and the status of the nodes in the
domain by using Cluster Resource Services graphical interfaces using these options:
 IBM Systems Director Navigator for i.
 Display CRG Information (DSPCRGINF).
 Work with Cluster (WRKCLU) commands.

Note: To use the Cluster Resource Services graphical interface or the Display CRG
Information (DSPCRGINF) command, you must have the IBM PowerHA SystemMirror for i
licensed program installed.

4.2 ASP copy descriptions


ASP copy descriptions are used by PowerHA to manage geographic mirroring, Metro Mirror,
Global Mirror, and FlashCopy copies. The copy description defines all the parameters that
are needed to access the disk units assigned to given iASP and execute Copy Services
operations by PowerHA. Figure 4-9 shows the relations between management objects
described in this section. The IASPs that are in the device domain for the cluster have the
same ASP copy descriptions on all nodes of the cluster.

PowerHA
Cluster

Node 1 Node 2

IASP1 Device Domain IASP2

ASP Copy ASP Copy


Description Description

ASP Copy ASP Copy


Description Description

ASP
Session

IBM System Storage IBM System Storage

Figure 4-9 General view of the ASP copy descriptions and ASP sessions

Chapter 4. PowerHA architecture 57


Figure 4-10 shows an example of the ASP copy description being added to the system.

Add ASP Copy Description (ADDASPCPYD)

Type choices, press Enter.

ASP copy . . . . . . . . . . . . ASPCPY > ASPCPYD1


ASP device . . . . . . . . . . . ASPDEV > IASP1
Cluster resource group . . . . . CRG > PWRHA_CRG1
Cluster resource group site . . SITE > SITE1
Storage host: STGHOST
User name . . . . . . . . . . > DS_USER
Password . . . . . . . . . . . > DS_PASSWORD
Internet address . . . . . . . > '10.10.10.10'

Location . . . . . . . . . . . . LOCATION > NODE1

Logical unit name: LUN


TotalStorage device . . . . . > 'IBM.2107-75AY032'
Logical unit range . . . . . . > 'A010-A013'
+ for more values
Consistency group range . . .
+ for more values
Recovery domain: RCYDMN
Cluster node . . . . . . . . . *NONE
Host identifier . . . . . . .
+ for more values
Volume group . . . . . . . . .
+ for more values
+ for more values

Figure 4-10 ADDASPCPYD example

Table 4-3 describes the parameters of this command.

Table 4-3 Parameters of the ASP copy description.


Parameter name Description

ASPCPY Defines the name of the ASP copy description.


Needs to be unique in the cluster.

ASPDEV This is the name of the IASP for which the ASP
copy description is being created. One IASP can
have many ASP copy descriptions.

CRG Cluster resource group in which this ASP is being


used. This parameter is valid for replication
solutions (geographic mirroring, Metro Mirror,
and Global Mirror) and for switched disks
solution. This parameter is not used for
FlashCopy purposes.

58 PowerHA SystemMirror for IBM i Cookbook


Parameter name Description

SITE Same as in the case of CRG, this parameter is for


replication solutions and is needed for switching
the IASP between the sites. This ASP copy
description will be used when the cluster
resource group will be switched to the site
specified in this parameter.

STGHOST This group of parameters defines the address,


user ID, and password for the HMC controlling
the IBM Data Storage system.

LOCATION This specifies the node that is used for this ASP
copy description.

This is used for FlashCopy to define which node


will be using this copy of IASP. In other cases it
should be set to *DEFAULT.

LUN This group defines which storage system to


use and what LUNs to be used for this ASP
copy description. This parameter is used to
bind the IASP with storage on the IBM Data
Storage system.

RCYDMN Recovery domain used for switchable LUNs


solution.

Chapter 4. PowerHA architecture 59


4.3 ASP sessions
An ASP session is used to link two ASP copy descriptions and start the Copy Services
functions between them. The status of the ASP session describes the current status of the
replication. Figure 4-11 shows an example of the status of the ASP session.

Display ASP Session NODE1


09/15/11 15:42:49
Session . . . . . . . . . . . . . . . . . . : IASP1SSN
Type . . . . . . . . . . . . . . . . . . : *METROMIR

Copy Descriptions

ASP
device Name Role State Node
IASP1 IASP1CPYD1 SOURCE UNKNOWN NODE1
IASP1 IASP1CPYD2 TARGET ACTIVE NODE2

Bottom
Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=Automatic refresh

Figure 4-11 ASP session example

4.3.1 Start/end
To start the session you can issue straspssn from the 5250 session (Figure 4-12 on
page 61). Table 4-4 describes the parameters for this command.

Table 4-4 STRASPSSN command parameters


Parameter Description

SSN Session name

TYPE Type of the session. The possible values are:


 *GEOMIR for the geographic mirroring
session
 *METROMIR for the Metro Mirror session
 *GLOBALMIR for the Global Mirror session
 *FLASHCOPY for the FlashCopy session

ASPCPY The names of the ASP copy descriptions that the


session is started between.

FLASHTYPE When you use type of the session *FLASHCOPY


you can choose here whether is the FlashCopy
will be COPY or NOCOPY type.

60 PowerHA SystemMirror for IBM i Cookbook


Parameter Description

PERSISTENT In case you use FlashCopy, you choose whether


the relation should be persistent.

Start ASP Session (STRASPSSN)

Type choices, press Enter.

Session . . . . . . . . . . . . SSN > ASPSSN1


Session type . . . . . . . . . . TYPE > *METROMIR
ASP copy: ASPCPY
Preferred source . . . . . . . > ASPCPYD1
Preferred target . . . . . . . > ASPCPYD2
+ for more values
FlashCopy type . . . . . . . . . FLASHTYPE *NOCOPY
Persistent relationship . . . . PERSISTENT *NO

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys

Figure 4-12 STRASPSSN

The ASP session is also started as a result of making IASP highly available in the PowerHA
cluster. To stop the session use this command:
ENDASPSSN SSN(IASP1SSN)

Or use the wrkaspcpyd option 24 next to the related ASP copy description. When the ASP
session is ended, the relation between the ASP copy descriptions is removed.

4.3.2 Changing attributes


In addition to the previously mentioned starting and stopping of the ASP sessions, you can
change some of the attributes of the session, including its status. Most of the changes
described later in this section can be done using CHGASPSSN.

The Change Auxiliary Storage Pool Session (CHGASPSSN) command can be used to change an
existing geographically mirrored, Metro Mirrored, Global Mirrored, or FlashCopy session.

Chapter 4. PowerHA architecture 61


Figure 4-13 is an example. Note that all values are not possible for all session types.

Change ASP Session (CHGASPSSN)

Type choices, press Enter.

Session . . . . . . . . . . . . > Name


Option . . . . . . . . . . . . . *CHGATTR, *SUSPEND...
ASP copy:
Preferred source . . . . . . . *SAME Name, *SAME
Preferred target . . . . . . . *SAME Name, *SAME
Consistency source . . . . . . *SAME Name, *SAME, *NONE
Consistency target . . . . . . *SAME Name, *SAME, *NONE
+ for more values
Suspend timeout . . . . . . . . *SAME 60-3600, *SAME
Mirroring mode . . . . . . . . . *SAME *SAME, *SYNC, *ASYNC
Synchronization priority . . . . *SAME *SAME, *LOW, *MEDIUM, *HIGH
Tracking space . . . . . . . . . *SAME 0-100, *SAME
FlashCopy type . . . . . . . . . *SAME *SAME, *COPY
Persistent relationship . . . . *SAME *SAME, *YES, *NO
ASP device . . . . . . . . . . . *ALL Name, *ALL
+ for more values
More...
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 4-13 Example of CHGASPSSN panel

To suspend or resume geographic mirroring, Metro Mirror, or Global Mirror through this
command, you must have already configured the mirror copy disks and cluster resource
group with a recovery domain that has two sites.

To change session attributes (the *CHGATTR parameter on the CHGASPSSN command) when
using geographic mirroring, the production copy of the iASP must be varied off.

CHGASPSSN with the *Detach, *Reattach, *Suspend, and *Resume options can be
confusing. It is hard to know which node you have to run which option from, in addition to
what states the iASP copies must be in first. Table 4-5 provides a quick reference list. To
simplify the chart we are letting source also mean production copy and target also mean
mirror copy.
Table 4-5 What options can be run from where and what status the iASP copies must be in
Source IASP Target IASP
CHGASPSSN Can run from Can run from must be varied must be varied
option Environment source? Target? off off

*Detach Metro Mirror Yes No Yes Yes

*Reattach Metro Mirror No Yes No Yes

*Suspend Metro Mirror Yes Yes No Yes

*Resume Metro Mirror Yes Yes No Yes

*Detach Geographic Mirroring Yes No No Yes

*Reattach Geographic Mirroring Yes No No Yes

62 PowerHA SystemMirror for IBM i Cookbook


Source IASP Target IASP
CHGASPSSN Can run from Can run from must be varied must be varied
option Environment source? Target? off off

*Suspend Geographic Mirroring Yes No No Yes

*Resume Geographic Mirroring Yes No No Yes

CHGASPSSN *SUSPEND or *DETACH will offer the tracking *YES or *NO (Figure 4-14).
However, the parameter is ignored if this is not a geographic mirroring solution. Both Metro
Mirror and Global Mirror will track regardless of this option.

Change ASP Session (CHGASPSSN)

Type choices, press Enter.

Session . . . . . . . . . . . . > ASPSSN Name


Option . . . . . . . . . . . . . > *DETACH *CHGATTR, *SUSPEND...
Track . . . . . . . . . . . . . *YES *YES, *NO

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 4-14 CHGASPSSN *DETACH for geographic mirroring

CHGASPSSN with geographic mirror will allow a *DETACH if the iASP is varied on. Because
the iASP is in use, this means that the mirror copy will have to go through an abnormal
vary-on, and there are risks to your data.

CHGASPSSN with Metro Mirroring will not allow a *DETACH if the iASP is varied on. If you
try you will get a CPD26B9 (Figure 4-15).

Additional Message Information

Message ID . . . . . . : CPD26B9 Severity . . . . . . . : 40


Message type . . . . . : Diagnostic
Date sent . . . . . . : 05/12/08 Time sent . . . . . . : 09:44:35

Message . . . . : Device METRO must be varied off for this change.


Cause . . . . . : The requested changes cannot be made on device METRO while
it is varied on.
Recovery . . . : Vary off the device (VRYCFG command). Then try the
request again.
Figure 4-15 msgCPD26B9 example

CHGASPSSN using the *DETACH parameter removes a FlashCopy relation, but it keeps the
ASP session.

CHGASPSSN using the *REATTACH parameter recreates a FlashCopy using the


parameters in an already created ASP session.

When you are going to reattach a Metro Mirror session, a message (CPF9898) will go to the
QSYSOPR message queue that you will have to use to confirm the reattach.

Chapter 4. PowerHA architecture 63


Error messages will not appear at the bottom of the panel if you run this message through
WRKASPCPYD. You will need to check your joblog specifically.

64 PowerHA SystemMirror for IBM i Cookbook


Part 2

Part 2 Concepts and planning


In this part we provide concepts and planning information for implementing the IBM High
Availability solution with PowerHA SystemMirror for i. We also introduce you to the various
interfaces that are available in IBM PowerHA for i.

This part includes these chapters:


 Chapter 5, “Geographic Mirroring” on page 67
 Chapter 6, “DS8000 Copy Services” on page 81
 Chapter 7, “Storwize V7000 and SAN Volume Controller Copy Services” on page 113
 Chapter 8, “Planning for PowerHA” on page 131
 Chapter 9, “PowerHA user interfaces” on page 157
 Chapter 10, “Advanced Copy Services for PowerHA” on page 167

© Copyright IBM Corp. 2012. All rights reserved. 65


66 PowerHA SystemMirror for IBM i Cookbook
5

Chapter 5. Geographic Mirroring


In this chapter, we introduce geographic mirroring and how it works from a
synchronous/asynchronous point of view. We illustrate several scenarios that serve as
disaster recovery solutions by using switched disk for local high availability and geographic
mirroring between remote sites.

These are the topics that we discuss:


 Concept of geographic mirroring
 Synchronous geographic mirroring
 Asynchronous geographic mirroring
 Switched disk for local HA + Geographic Mirroring for DR scenario

© Copyright IBM Corp. 2012. All rights reserved. 67


5.1 Concept of geographic mirroring
Geographic mirroring has been available in i5/OS V5R3M0. Geographic mirroring specifically
refers to the IBM i host-based management storage mirroring solution. Now geographic
mirroring is provided as a function of PowerHA SystemMirror for i products. PowerHA
SystemMirror for i also includes replication solutions such as Metro Mirror and Global Mirror
by using IBM Storage products.

The basis of geographic mirroring is extended IBM i disk mirroring technology to multiple
systems environment. Geographic mirroring is managed by IBM i storage management
capability so that replication is performed by each memory page segment (4096 bytes).

Geographic mirroring is intended for use by clustered system environments and uses data
port services. Data port services are System Licensed Internal Code that supports the
transfer of large volumes of data between a source system and one of any specified target
systems. This is a transport mechanism that communicates over TCP/IP.

Prior to IBM i 7.1, it provides only synchronous delivery mode. Be aware that a local write
waits for the data to reach the main storage of the backup node before the write operation is
considered to be finished.

Asynchronous geographic mirroring (asynchronous delivery mode) is supported by PowerHA


SystemMirror for i Enterprise Edition (5770-HAS option 1) with IBM i 7.1 extending the
previously available synchronous geographic mirroring option, which for performance
reasons is practically limited to metro area distances up to 30 km.

Cluster
Device domain
Cluster resource group (CRG)

Node A Node B
Production Copy Mirror Copy

TCP/IP Network
SYSBAS SYSBAS

IASP IASP
source target

Data port services for


data copy
Site A Site B

Figure 5-1 Geographic mirroring between two sites

68 PowerHA SystemMirror for IBM i Cookbook


Note: The data replication process between the two sets of disks (production copy and
mirror copy), described by the blue arrow in Figure 5-1, is performed based on TCP/IP
communication between the two systems.

The copy owned by the primary node is the production copy, and the copy owned by the
backup system at the other site is the mirror copy. Users and applications can only access the
independent disk pool on the primary node, the node that owns the production copy. Changes
that are made to the production copy (source system) are guaranteed by the geographic
mirroring functionality to be made in the same order on the mirror copy (target system
or partition).

Geographic mirroring allows for the production and mirrored copies to be at the same site for
high-availability protection in the event of server failure, and it is also possible to separate the
two systems geographically for disaster recovery protection in the event of a site-wide outage,
provided that the communication link between the two sites is enough for your
synchronization timing policy.

Geographic mirroring functionality involves the use of these cluster components:


 Cluster
 Device domain
 Cluster resource group
 ASP copy description
 ASP session
 Cluster administrative domain
 Independent auxiliary storage pools (IASPs)

Configuration basis for geographic mirroring


The nodes participating in geographic mirroring must be part of the same cluster and of the
same device domain, and their role is defined in the recovery domain of the cluster resource
group (CRG). Before configuring geographic mirroring, you must specify a site name and the
TCP/IP address to be used for the mirroring process (up to four) for each node in the recovery
domain within the device CRG of your IASP. When you configure the cluster for geographic
mirroring, you have many options for defining the availability and the protection of the
independent disk pool.

Geographic mirroring happens on the page level of storage management. Therefore, the size
of individual disks and the disk protection used on the production IASP can differ from what is
used on the backup IASP. The overall size of the two IASPs should be about the same on
both systems though.

Important: If both disk pools (source and target) do not have the same disk capacity
available, when the mirrored copy reaches 100%, geographic mirroring is suspended. You
can still add data to your production IASP, but you lose your high availability. If, however,
the production copy reaches 100%, you are no longer able to add data to that IASP, and
applications trying to do so come to a stop. An IASP can never overflow to the system
ASP, as that would compromise your high-availability environment.

Data port services


Geographic mirroring provides logical page-level mirroring between independent disk pools
through the use of data port services. Data port services manages connections for multiple IP
addresses (up to four), which provides redundancy and greater bandwidth in geographic
mirroring environments. We strongly suggest that different IP interfaces, connected to
different networks, be used for data port services and cluster heartbeating. You can see more

Chapter 5. Geographic Mirroring 69


detailed considerations of the communication environment in “Communications lines” on
page 147.

Mirroring option
You can specify how to replicate from your production data in IASP to back up IASP.
Mirroring options of geographic mirroring can be specified as ASP session attributes, which
are a combination of two parameters:
 Transmission delivery
 Mirroring mode

Table 5-1 shows you the combination of these parameters that you can specify.

Table 5-1 Mirroring option


Mirroring mode

*SYNC *ASYNC

Synchronous geographic Synchronous geographic


*SYNC mirroring mirroring
Transmission (synchronous mirroring mode) (asynchronous mirroring mode)
delivery

*ASYNC N/a Asynchronous geographic


mirroring

As shown in Table 5-1, synchronous geographic mirroring has two modes:


 Synchronous mirroring mode
 Synchronous mirroring mode (also called semi-asynchronous mode)

Both mirroring modes use synchronous communication between sites. Synchronous


geographic mirroring mode is configured to specify transmission delivery as synchronous in
IBM i 7.1.

Figure 5-2 shows the DSPASPSSN command screen as configuring synchronous geographic
mirroring with synchronous mirroring mode.

Display ASP Session DEMOGEO1


09/30/11 11:20:48
Session . . . . . . . . . . . . . . . . . . : GEOMIRROR
Type . . . . . . . . . . . . . . . . . . : *GEOMIR
Transmission Delivery . . . . . . . . . . : *SYNC
Mirroring Mode . . . . . . . . . . . . . : *SYNC
Suspend timeout . . . . . . . . . . . . . : 120
Synchronization priority . . . . . . . . : *MEDIUM
Tracking space allocated . . . . . . . . : 100%

Figure 5-2 Display ASP session command: Synchronous mirroring mode

70 PowerHA SystemMirror for IBM i Cookbook


Figure 5-3 shows the DSPASPSSN command screen as configuring synchronous geographic
mirroring with asynchronous mirroring mode.

Display ASP Session DEMOGEO1


09/30/11 11:15:49
Session . . . . . . . . . . . . . . . . . . : GEOMIRROR
Type . . . . . . . . . . . . . . . . . . : *GEOMIR
Transmission Delivery . . . . . . . . . . : *SYNC
Mirroring Mode . . . . . . . . . . . . . : *ASYNC
Suspend timeout . . . . . . . . . . . . . : 120
Synchronization priority . . . . . . . . : *MEDIUM
Tracking space allocated . . . . . . . . : 100%

Figure 5-3 Display ASP session command: Asynchronous mirroring mode

Figure 5-4 shows the DSPASPSSN command screen as configuring asynchronous


geographic mode.

Display ASP Session DEMOGEO1


09/30/11 11:17:20
Session . . . . . . . . . . . . . . . . . . : GEOMIRROR
Type . . . . . . . . . . . . . . . . . . : *GEOMIR
Transmission Delivery . . . . . . . . . . : *ASYNC
Mirroring Mode . . . . . . . . . . . . . : *ASYNC
Suspend timeout . . . . . . . . . . . . . : 120
Synchronization priority . . . . . . . . : *MEDIUM
Tracking space allocated . . . . . . . . : 100%


Figure 5-4 Display ASP session command: Asynchronous geographic mode

Synchronization
When geographic mirroring is resumed after suspend or detach, the mirror copy will be
resynchronized with the production copy. The production copy can function normally during
synchronization, but performance might be negatively affected. During synchronization, the
contents of the mirror copy are unusable, and it cannot become the production copy. If the
independent disk pool is made unavailable during the synchronization process,
synchronization resumes where it left off when the independent disk pool is made available
again. Message CPI095D is sent to the QSYSOPR message queue every 15 minutes to
indicate progression of the synchronization.

These are the two types of synchronization:


 Full synchronization
Indicates that a complete synchronization takes place. Changes to the production copy
were not tracked to apply to the synchronization. A full synchronization first deletes all
data in the backup IASP and then copies the current data from the production IASP to the
backup IASP.
 Partial synchronization
Indicates that changes to the production copy were tracked while geographic mirroring
was suspended or detached. This might shorten the synchronization time considerably

Chapter 5. Geographic Mirroring 71


because a complete synchronization is unnecessary. In this case when the mirror copy is
reattached and geographic mirroring is resumed, only tracked changes will need to be
synchronized. Changes made on the production copy (since the detach has been done)
are sent to the mirror copy, and any change made on the mirror copy will be overlayed
with the original production data coming from the production copy of the IASP. Logically,
any changes made on the detached mirror copy are undone, and any tracked changes
from the production copy are applied.

Message CPI095D indicates what types of synchronization happened. Figure 5-5 shows
the message.


The synchronization is of type 2. The synchronization types and their meanings
are as follows:
1 - The synchronization being performed is a synchronization of tracked
changes.
2 - The synchronization being performed is a synchronization of all data.

Figure 5-5 Message CPI095D

Two parameters can be used to better manage IASP copies synchronization and application
performances when geographic mirroring is used:
 Synchronization priority
When you set the attributes for geographic mirroring, you can set the synchronization
priority. You can select synchronization priority as high, medium, or low. If synchronization
priority is set high, the system uses more resources for synchronization, which results in a
sooner completion time. The mirror copy is eligible to become a production copy faster, so
you are protected sooner. However, high priority can cause degradation to your
application. We recommend that you try high priority first, so you are protected as soon as
possible. If the degradation to your application performance is not tolerable, then lower the
priority. Be aware that you need to vary off the IASP to perform this change.
 Suspend timeout
In addition to synchronization priority, you can also set the suspend timeout. The
suspend timeout specifies how long your application can wait when geographic mirroring
cannot be performed. When an error, such as a failure of the communication link, prevents
geographic mirroring from occurring, the source system waits and retries for the
specified suspend timeout before suspending geographic mirroring, which allows your
application to continue.

72 PowerHA SystemMirror for IBM i Cookbook


Figure 5-6 shows you as example of display geographic mirroring attribute via DSPASPSSN.

Display ASP Session DEMOGEO1


09/30/11 11:13:03
Session . . . . . . . . . . . . . . . . . . : GEOMIRROR
Type . . . . . . . . . . . . . . . . . . : *GEOMIR
Transmission Delivery . . . . . . . . . . : *SYNC
Mirroring Mode . . . . . . . . . . . . . : *SYNC
Suspend timeout . . . . . . . . . . . . . : 120
Synchronization priority . . . . . . . . : *MEDIUM
Tracking space allocated . . . . . . . . : 100%

Copy Descriptions

ASP ASP Data


Device Copy Role State State Node
IASP1 S1CPYD PRODUCTION AVAILABLE USABLE DEMOGEO1
IASP1 S2CPYD MIRROR ACTIVE USABLE DEMOGEO2

Bottom
Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=Automatic refresh


Figure 5-6 Display geographic mirroring attribute via DSPASPSSN command

Tracking space
The tracking space was introduced with IBM i 5.4. It enables geographic mirroring to track
changed pages while in suspended status. With tracked changes we can avoid full
resynchronization after a resume in many cases, therefore minimizing exposed times where
you do not have a valid mirror copy. Tracking space gets configured when you configure
geographic mirroring or change geographic mirroring attributes via CHGASPSSN.

Tracking space is allocated inside of the independent ASPs. The more tracking space that
you specify, the more changes the system can track. The amount of space for tracking can be
defined by the user up to 1% of the total independent ASP capacity. When we specify tracking
space size, we can specify a percentage of total usable tracking space size. If you specify
100%, you have 1% of your total independent ASP capacity as tracking space size. For
example, if you have an independent ASP with 100 GB, you can have a maximum of 1 GB of
storage space as the tracking space, and if you specify the tracking space parameter as 50%,
you have 500 MB of storage space as tracking space. Be aware that this tracking space does
not contain any changed data. It just holds information about what pages in the IASP have
been changed.

Chapter 5. Geographic Mirroring 73


5.2 Synchronous geographic mirroring
In this section we discuss synchronous geographic mirroring.

5.2.1 Synchronous geographic mirroring with synchronous mirroring mode


Synchronous geographic mirroring with synchronous mirroring mode needs to specify that
both transmission delivery and mirroring mode are synchronous.

When geographic mirroring is active in synchronous mode (Figure 5-7), the write on disk
operation waits until the operation is complete to the disk (actually the data write to the IOA
cache) on both the source (acknowledgement operation #4) and target systems
(acknowledgement operation #3) before sending the acknowledgment to the storage
management function of the operating system of the production copy. See the operations
numbered 1 - 4 with green arrows in Figure 5-7.

The mirror copy is always eligible to become the production copy, because the order of writes
is preserved on the mirror copy. If you are planning to use geographic mirroring as helocal HA
solution, we recommend trying synchronous mode first. If your performance remains
acceptable, continue to use synchronous geographic mirroring.

Site A Site B
Production Copy Mirror Copy

Main 2
Memory 3

4 1 TCP/IP Network

IOA Cache

IOA Cache
IASP
source
IASP
target

Figure 5-7 Synchronous Mirroring mode

5.2.2 Synchronous geographic mirroring with asynchronous mirroring mode


When geographic mirroring is using asynchronous mirroring mode, the write on disk
operation must wait to get an acknowledgement from the production copy for the write
operation when it is completed to the disk (actually to the IOA cache - operation #4) on the

74 PowerHA SystemMirror for IBM i Cookbook


source system and is received for processing on the target system (actually in main memory -
operation #2 and acknowledgement operation #3) only. See the operations numbered 1 - 4
with green arrows shown in Figure 5-8. The physical write operation, #5 in orange in the
figure, is performed later (asynchronously) to the disk on the mirror copy (target system).

In IBM i 7.1, asynchronous mirroring mode can activate to specify that transmission delivery
is synchronous and mirroring mode is asynchronous.

Site A
Production Copy Site B
Mirror Copy

2
Main
Memory 3

4 1 Main
TCP/IP Network Memory

IOA Cache 5

IOA Cache
IASP
source

IASP
target

Figure 5-8 Asynchronous mirroring mode

In this mode, the pending updates must be completed before the mirror copy can become the
production copy. This means that while you might see a slightly better performance during
normal operation, your switchover or failover times might be slightly longer because changes
to the backup IASP might still reside in the main memory of your backup system. They must
be written to disk before the IASP can be varied on.

Important: Because this mode still waits for the write to cross to the target system, it is not
truly asynchronous. We recommend this for situations in which the source and target
systems are less than 30 km apart.

5.3 Asynchronous geographic mirroring


Asynchronous geographic mirroring is a new capability supported by PowerHA SystemMirror
for i Enterprise Edition (5770-HAS option 1) with IBM i 7.1, extending the previously available
synchronous geographic mirroring option, which for performance reasons is practically limited
to metro area distances up to 30 km.

Chapter 5. Geographic Mirroring 75


The asynchronous delivery of geographic mirroring (not to be confused with the
asynchronous mirroring mode of synchronous geographic mirroring) allows IP-based
hardware replication beyond synchronous geographic mirroring limits (Figure 5-7 on page 74
and Figure 5-8 on page 75). The write on disk operation does not wait until the operation is
delivered to target system (operation #1 and #2).

Asynchronous transmission delivery, which also requires the asynchronous mirroring mode,
works by duplicating any changed IASP disk pages in the *BASE memory pool on the source
system and sending them asynchronously while preserving the write-order to the target
system in operation #3 in Figure 5-9. Therefore, at any given time, the data on the target
system (though not up-to-date) still represents a so-called crash-consistent copy of the
source system.

Site A
Production Copy Site B
Mirror Copy

3
Main
Memory 5

2 1 Main
TCP/IP Network Memory

IOA Cache 4

IOA Cache
IASP
source

IASP
target

Figure 5-9 Asynchronous geographic mirroring

The asynchronous geographic mirroring option has potential performance impacts to system
resources, such as processor and memory because communication lines with longer latency
times might tie up additional memory resources for keeping their changed data.

76 PowerHA SystemMirror for IBM i Cookbook


Also consider the distance and amount of latency of your communication lines by using mirror
activities. If you have very write-intensive applications, they can flood the communication line
and suspend your session with he auto resume process. Then you might tune the timeout
value by using CHGASPSSN (Figure 5-10). The default value of the SSPTIMO (Suspend timeout)
parameter is 120 seconds.

Change ASP Session (CHGASPSSN)

Type choices, press Enter.

Session . . . . . . . . . . . . > GEOMIRROR Name


Option . . . . . . . . . . . . . > *CHGATTR *CHGATTR, *SUSPEND...
ASP copy:
Preferred source . . . . . . . > S1CPYD Name, *SAME
Preferred target . . . . . . . > S2CPYD Name, *SAME
+ for more values
Suspend timeout . . . . . . . . > 120 60-3600, *SAME
Transmission delivery . . . . . > *SYNC *SAME, *SYNC, *ASYNC
Mirroring mode . . . . . . . . . > *SYNC *SAME, *SYNC, *ASYNC
Synchronization priority . . . . > *MEDIUM *SAME, *LOW, *MEDIUM, *HIGH
Tracking space . . . . . . . . . > 100 0-100, *SAME
FlashCopy type . . . . . . . . . *SAME *SAME, *COPY
Persistent relationship . . . . *SAME *SAME, *YES, *NO

Bottom
F3=Exit F4=Prompt F5=Refresh F10=Additional parameters F12=Cancel
F13=How to use this display F24=More keys
Figure 5-10 Example of Change ASP session command (change session attributes)

For more information about performance of geographic mirroring see 8.3.1, “Geographic
mirroring” on page 144.

5.4 Switched disk for local HA and geographic mirroring for DR


Switched disk high-availability solutions are based on the concept of switching entire
independent auxiliary storage pools (IASPs) from one system to another.

5.4.1 Switched disks between logical partitions


In this environment the hardware is switched between two logical partitions on the same
physical system. When configuring this type of environment, you need to group all of the
switchable hardware resources between partitions into an I/O pool using the Hardware
Management Console (HMC). This allows for greater flexibility in defining what will be
switched because you no longer have to move all the hardware in an entire tower when doing
the switching. With this environment you can select which pieces of hardware are switched
and which are not.

Chapter 5. Geographic Mirroring 77


Note: Use of high-speed link (HSL) loop connections for IASP switching is no longer
supported. POWER6 servers were the last servers supported for switched IASPs
between servers.

5.4.2 Combining geographic mirroring and switched disks


Switchable IASPs offer a local HA solution to address system failures involving a particular
LPAR. Adding any kind of remote replication to an existing switched disk environment
instantly transforms it into a DR environment and provides coverage for disasters involving an
entire site. As you can see in Figure 5-11, redundant LPARs provide protection from LPAR
outage that might be due to a select software outage or a select planned outage.

Source Target

LPAR-1 LPAR-1

*SYSBAS Any type of *SYSBAS


IASP supported
replication

LPAR-1

IASP
*SYSBAS

Tape backup

Figure 5-11 Switched IASP combined with replication technologies

Adding replication to a remote site server provides additional levels of protection, including
disaster recovery or off-line tape backups. This combination can be achieved using internal
disks because PowerHA supports geographic mirroring with internal disks, and the replication
can be either synchronous or asynchronous.

Note: Geographic mirroring must first be suspended before IASP can be backed up to
tape. PowerHA tracks changes and requires time for partial resynch when backup is
complete, and hence affects your recovery point objective (RPO).

78 PowerHA SystemMirror for IBM i Cookbook


As shown in Figure 5-12, use the switched disk environment locally to achieve high
availability, whereas the system using geographic mirroring can sit in a remote location and
therefore also help to achieve disaster recovery. All three nodes are part of the recovery
domain. Node 1 is the primary node, node 2 is the backup node one, and node 3 is backup
node two. If just node 1 fails, then node 2 becomes the primary node (by switching and
attaching the IASP to that node), and node 3 becomes backup node one. Geographic
mirroring would still be active. If the entire site that is hosting both node 1 and node 2 fails,
then node 3 becomes the primary node and work can continue there.

Node 1 Cluster – Device Domain - CRG Node 3


Node 2
PRODUCTION MIRROR
copy copy

LPAR/ Node 3
Node 1

IASP Geographic IASP


source Mirroring target
TCP/IP network
LPAR/
Node 2

Switched
disk

Site A Site B

Figure 5-12 Combination of geographic mirroring and switched disk

Chapter 5. Geographic Mirroring 79


80 PowerHA SystemMirror for IBM i Cookbook
6

Chapter 6. DS8000 Copy Services


In this chapter we describe the DS8000 Copy Services features, which can be used by IBM
PowerHA SystemMirror for i. This chapter describes each feature and how it works. However,
before talking about Copy Services, a brief description of DS8000 storage concepts is useful.

© Copyright IBM Corp. 2012. All rights reserved. 81


6.1 DS8000 storage concepts
In this section we describe the major concepts of the DS8000 storage system. The goal of
this virtualization is to make sure that the host’s operating systems will continue to work
exactly like they do with the usual internal drives.

For additional information, especially when performance considerations apply, refer to IBM
System Storage DS8000: Architecture and Implementation, SG24-8886.

6.1.1 Hardware overview


Before presenting the virtualization concepts, we provide a brief section about the hardware
components. At the time of writing, there are two models available as a DS8000. They are the
DS8700 and the DS8800. These are the main hardware components:
 Base and expansion frames
– The DS8700 can have up to four expansion frames. A fully populated five-frame
DS8700 contains 1024 3.5” disk drives.
– The DS8800 can have up to two expansion frames. A fully populated three-frame
DS880 contains 1056 2.5” disks.
 Processors
Both DS8700 and DS8800 use POWER6 processors, the same as those that are installed
in the IBM Power Systems 570 servers (based on POWER6).
For redundancy purposes, there are two processors complexes installed in the base
frame. They normally are named server 0 and server 1. The total installable memory is
384 GB, each server using half of the installed memory. This memory is used to provide
IO cache capabilities:
– The DS8700 can have 2-way or 4-way POWER6 processors, running at 4.7 GHz.
– The DS8800 can have 2-way or 4-way POWER6+™ processors, running at 5.0 GHz.
Both DS8700 and DS8800 use the Peripheral Component Interconnect Express (PCIe)
infrastructure to access disk subsystems and host connections.
 I/O enclosures
I/O enclosures are installed in both the base frame and the first expansion frame with a
maximum of eight enclosures. Each enclosure can have a maximum of four Fibre Channel
4-port adapters in the DS8700 and two 8-port adapters in the DS8800. The maximum
number of ports is 128 4-Gbit/s ports. 8-Gbit/s ports are supported, but installation
restrictions apply.
 Device adapters
Installed in the I/O enclosures, device adapters connect to the installed disks drives
through Fibre Channel interfaces card (running at 2 Gbps on DS8700 and 8 Gbps on
DS8800) installed in the disks enclosures. Working by pair, one attached to each server,
they are responsible for disks drives RAID protections.
 Disks enclosures
Disks enclosures contain disks drives modules, Fibre Channel Interface Cards to connect
to the device adapters and to the disks.

82 PowerHA SystemMirror for IBM i Cookbook


Table 6-1 lists available disk formats.
Table 6-1 Available disk formats
Disks drive formats Availability and connection

300 GB 15 K rpm Fibre Channel interface on DS8700

450 GB 15 K rpm Fibre Channel interface on DS8700

600 GB 15 K rpm Fibre Channel interface on DS8700

146 GB 15 K rpm SAS interface on DS8800

450 GB 10 K rpm SAS interface on DS8800

600 GB 10 K rpm SAS interface on DS8800

2 TB 7,2 K rpm SATA interface on DS8700

300 GB SSD SAS interface on DS8800

600 GB SSD Fibre Channel interface on DS8700

 Power and batteries


Power supplies, batteries, and cooling are highly redundant.
There are two redundant power supplies in each frame. Each server (processor complex)
has two power supply units. Each disk enclosure has two power supply units. Each I/O
enclosure has two power supply units.
The battery backup (BBU) assemblies help protect data in the event of a loss of
external power. In the event of a complete loss of AC input power, the battery assemblies
are used to maintain power to the processor complexes and I/O enclosures for a sufficient
period of time, to allow the contents of NonVolatileStorage memory (modified data not yet
destaged to disk from cache) to be written to a number of disk drives internal to the
processor complexes.
 Hardware Management Console (HMC)
All base frames ship with an integrated HMC, which is used by these people:
– IBM representatives to perform all maintenance operations, like replacing a failing item
– The user to perform configuration and management through command-line interfaces
and graphical user interfaces

Chapter 6. DS8000 Copy Services 83


6.1.2 Array site
An array site is a group of eight disk drives. Those disk drives are already assigned to array
sites at the installation by the DS8000, and you do not have to deal with them. They are
selected from two disk enclosures on two different loops (Figure 6-1). The loops are the
internal links between the DS8000 RAID device adapters connected to the DS8000
processors and the disks drive modules.

Array
.. … … 24 … … .. Site

.. … … 24 … … ..
Loop 1 Loop 2
Figure 6-1 An array site

All disk drives in an array site have the same capacity and the same speed.

6.1.3 Array
DS8000 logical storage configuration starts at this point. An array is a group of eight disk
drives created from an array site, defining the disk protection that you want to use. Available
disks protections are RAID10, RAID5, and RAID6. When creating the array, specify the array
site to use and the protection to apply. At this time, the DS8000 selects a spare disk to put in
the array, if needed according to sparing algorithm. None, one, or two spare disks per array
can exist.

Note: Spare disks, while they remain spare, are not a member of any RAID parity set
even if they are a member of the array. They become a member of the RAID parity set if
they replace a failing drive. These operations are automatically handled by the DS8000
when needed.

84 PowerHA SystemMirror for IBM i Cookbook


Several types of arrays exist, depending on the existence of spares and the selected
disks protection:
4+4 RAID10 protection with no spare, four disks mirrored with four other disks.
3+3+2S RAID10 protection with two spares, three disks mirrored with three other disks.
6+P+Q RAID6 protection with no spare, two parity disks.
5+P+Q+S RAID6 protection with one spare, two parity disks.
7+P RAID5 protection with no spare, one parity disk.
6+P+S RAID5 protection with one spare, one parity disk (Figure 6-2). In Figure 6-2, the
DS8000 reserves one entire disk for spare and distributes parity space and
data space on each of the other disks (on the right side, Dx stands for data
space and P for parity space).

Array
Site D1 D7 D13 ...
D2 D8 D14 ...
D3 D9 D15 ...
D4 D10 D16 ...
D5 D11 P ...
D6 P D17 ...
Creation of
an array P D12 D18 ...

Data
Data
Data
Data RAID
Data Array Spare
Data
Parity
Spare

Figure 6-2 A RAID5 array with spare

A one-to-one relationship exists between an array and an array site.

6.1.4 Rank
The next configuration step is to dedicate the array to one of the two disk formats provided by
the DS8000. It supports, at the same time, connections from both System z® servers and
Windows/Linux/Unix and IBM i operating systems, but disk formats are not the same for both
types. System z servers make use of count key data (CKD) format, while all the others make
use of fixed block (FB) format.

The rank defines the way that the entire array is formatted for CKD usage or FB usage. At
rank creation time, the array is also divided into extents. One extent is the smallest disk space
that can be allocated to DS8000 host operating systems.

Chapter 6. DS8000 Copy Services 85


For CKD usage, one extent is equivalent to 1113 cylinders, which was the capacity of an
original 3390 model 1 disk. System z administrators are used to deal with cylinders as disks
size units.

For FB usage, one extent is equivalent to 1 GB (more precisely GiB, being equal to
230 bytes).

Figure 6-3 shows an example of a fixed block rank with 1 GB extents. In Figure 6-3, one
square represents one extent. The DS8000 distributes every extent to all disks members of
the array.

Data
Data D1 D7 D13 ...
D2 D8 D14 ...
Data
D3 D9 D15 ...
Data RAID D4 D10 D16 ...
Data Array D5 D11 P ...
Data D6 P D17 ...
Parity P D12 D18 ...
Spare

Creation
of a Rank

FB Rank
1 GB 1 GB 1 GB 1 GB
of 1 GB
FB FB FB FB extents

Figure 6-3 A fixed block rank

Note: From a DS8000 standpoint, the IBM i operating system and Windows/Unix/Linux
operating systems are members of the same community, the fixed block community. This
means that, for example, on the same rank there can coexist extents used by an IBM i
operating system and others used by a Windows operating system. You have to make
sure that you understand possible performance impacts with such a configuration.

A one-to-one relationship exists between a rank and an array.

6.1.5 Extent pools


An extent pool aggregates extents of the same usage (CKD or FB) from one or more ranks
into a single object. The extent pool provides also the affinity to one of the two DS8000
servers through its rank group parameter, which can be set to 0 for server 0 or to 1 for
server 1. Affinity means preferred server usage during normal operations. For performance
reasons, it is important to properly balance servers usage. At the minimum, two extent pools
must exist on a DS8000, one for server 0 and the other for server 1.

86 PowerHA SystemMirror for IBM i Cookbook


Figure 6-4 shows an example of two CKD extent pools and two FB extents pools:
 The CKD0 extent pool has an affinity to server 0 and makes use of two ranks.
 The CKD1 extent pool has an affinity to server 1 and makes use of a single rank.
 The FBtest extent pool has an affinity to server 0 and makes use of two ranks.
 The FBprod extent pool has an affinity to server 1 and makes use of three ranks.

Note: Server affinity applies only during normal operations. If a DS8000 server fails, the
other is able to manage extent pools previously managed by the failing server.

Extent Pool CKD0 Extent Pool CKD1

1113 1113 1113 1113 1113 1113 1113 1113


Cyl Cyl Cyl Cyl Cyl Cyl Cyl Cyl
CKD CKD CKD CKD CKD CKD CKD CKD

1113 1113 1113 1113 Extent Pool FBprod


Cyl Cyl Cyl Cyl
CKD CKD CKD CKD
Server0

Server1
1 GB 1 GB 1 GB 1 GB
FB FB FB FB

Extent Pool FBtest

1 GB 1 GB 1 GB 1 GB 1 GB 1 GB 1 GB 1 GB
FB FB FB FB FB FB FB FB

1 GB 1 GB 1 GB 1 GB 1 GB 1 GB 1 GB 1 GB
FB FB FB FB FB FB FB FB

Figure 6-4 Extent pools

Note: Performance considerations apply whether you decide to assign several ranks or a
single rank to an extent pool.

A one-to-n relationship exists between an extent pool and a rank.

6.1.6 Volumes
Until this step, the DS8000 is not yet ready for providing any storage resources to host
operating systems. The next step is to create volumes, or logical volumes, by using a specific
extent pool. This one provides the volume, at creation time, with the desired number of
extents, located on the associated ranks. The volume does not take care of the extents’
physical location. Volume creation commands are not the same for CKD extent pools and for
FB extent pools because of parameter differences.

Chapter 6. DS8000 Copy Services 87


Creating a CKD volume requires the number of cylinders. Creating a FB volume (also known
as a LUN for logical unit) requires the number of GB. A volume cannot span multiple extent
pools. After it is created, the volume gets a type/model related to the host operating system.

Figure 6-5 shows the process of creating a 2.9 GB FB volume in an extent pool spanned on
two ranks. Three extents (so 3 GB) remain available in the extent pool. Other extents are in
use by other LUNs. Creating the 2.9 GB volume succeeds but allocates the three remaining
extents, leaving 100 MB unused.

Extent Pool FBprod


Logical
3 GB LUN
Rank-a 1 GB
1 GB 1 GB 1 GB
free
3 GB LUN

1 GB 1 GB Allocate a 2.9 GB volume


Rank-b Used Used
free free

Extent Pool FBprod

1 GB
Rank-a 1 GB 1 GB 1 GB
used
3 GB LUN 2.9 GB LUN
created

1 GB 1 GB
Rank-b Used Used
used used 100 MB unused

Figure 6-5 Creation of a FB volume

IBM i volumes are fixed block volumes, so they are also composed of a number of 1 GB
extents. However, for DS8000 native or NPIV attachment, the IBM i operating system
supports only certain fixed volumes sizes that are the same as supported physical disks
drives. For example, supported volume sizes are 8.58 GB, 17.54 GB, 35.16 GB, 70.54GB,
and so on. The other specific item for IBM i volumes is the disk protection that the operating
system wants to achieve. If you want to use host-based IBM i mirroring, IBM i must see
unprotected volumes. Otherwise it will be not possible to define and configure the mirrioring.
The protection flag provided by the DS8000 for an IBM i volume is only a way to make IBM i
believe that the volume is not protected. In reality, all volumes are RAID protected by the
DS8000 storage system, as discussed in 6.1.3, “Array” on page 84.

The DS8000 provides the IBM i operating system with a specific serial number for every
volume, just like physical disks drives have.

The DS8000 model determines the first two digits of the serial number (for example, DS8100,
DS8300, DS8700 report 50). They are the same for all volumes. The next four digits are the
volume ID. The last three digits are known as the OS/400 serial number suffix, which is set up
at the DS8000 storage image level. This is the same for all volumes, and it points you to the
appropriate DS8000, if your installation does not have a unique one.

88 PowerHA SystemMirror for IBM i Cookbook


Figure 6-6 shows an example of a disk serial number. In 50-EE902B8, 50 is related to the
DS8000 model, EE90 is the volume ID inside the DS8000 configuration, and 2B8 is the
os400serial at the DS8000 storage image level.

Display Resource Detail


System: CTCIHA9X
Resource name . . . . . . . : DMP008
Text . . . . . . . . . . . . : Disk Unit
Type-model . . . . . . . . . : 2107-A82
Serial number . . . . . . . : 50-EE902B8
Part number . . . . . . . . :

Location: U8233.E8B.10001AP-V6-C60-T1-W50050763070102B8-L40EE409000000000

Logical address:
SPD bus:
System bus 255
System board 128
More...
Press Enter to continue.

F3=Exit F5=Refresh F6=Print F12=Cancel

Figure 6-6 Disk serial number seen by IBM i operating system

Chapter 6. DS8000 Copy Services 89


The following figures show the related information on the DS8000 side. Figure 6-7 shows the
storage image os400serial in bold. By default, these three characters are the last three one of
the storage image world wide node name, but they can be changed.

Caution: Any change to os400serial requires a DS8000 reboot to be applied.

dscli> showsi IBM.2107-75AY031


Date/Time: September 15, 2011 1:48:29 PM CDT
Name -
desc -
ID IBM.2107-75AY031
Storage Unit IBM.2107-75AY030
Model 9B2
WWNN 5005076307FFC2B8
Signature 232c-a1ec-939a-bb44
State Online
ESSNet Enabled
Volume Group V0
os400Serial 2B8
NVS Memory 2.0 GB
Cache Memory 53.9 GB
Processor Memory 62.7 GB
MTS IBM.2424-75AY030
numegsupported 0
ETAutoMode -
ETMonitor -
IOPMmode -
Figure 6-7 DS8000 storage image settings

90 PowerHA SystemMirror for IBM i Cookbook


Figure 6-8 shows the volume settings. The volume ID is marked in bold. We can also see
that, from the DS8000 stand point, deviceMTM is the same parameter as type-model from the
IBM i operating system stand point (2107-A82 in our case).

dscli> showfbvol EE90


Date/Time: September 15, 2011 1:48:43 PM CDT
Name HA9x_sys
ID EE90
accstate Online
datastate Normal
configstate Normal
deviceMTM 2107-A82
datatype FB 520U
addrgrp E
extpool P2
exts 17
captype iSeries
cap (2^30B) 16.3
cap (10^9B) 17.5
cap (blocks) 34275328
volgrp V41
ranks 1
dbexts -
sam Standard
repcapalloc -
eam rotateexts
reqcap (blocks) 34275328
realextents 17
virtualextents 0
migrating 0
perfgrp -
migratingfrom -
resgrp -
Figure 6-8 Fixed-block volume settings

An n-to-one relationship exists between volumes and an extents pool.

6.1.7 Volume groups


As its name states, a volume group is a set of volumes. The volumes are presented together
by the DS8000 to the host operating system through the volumes group. This one has a type
that relates to the operating system that makes use of it. The types of volumes included into a
volume must match the volume group type. For example, you cannot include volumes that do
not have an IBM i type model (such as 2107-A82) into an IBM i volume group.

An n-to-n relationship exists between volumes and volumes groups. However, if a volume
belongs to more than one volume’s group, it means that it is potentially accessed by several
hosts. In this case, the host operating system is responsible for ensuring data integrity. he
IBM i operating system does not support sharing a volume with another operating system.

Chapter 6. DS8000 Copy Services 91


6.1.8 Host connections
A host connection is an object that represents the Fibre Channel adapter port of a host on
DS8000. The main attributes of a host connection are its type, related to the volume group
type, and its world-wide port name (WWPN), which is a unique 16-hexadecimal-character
address. After creating a host connection, we are almost ready for this host to get disk
resources to work with. We only have to assign the host connection a volume group. At this
time, the host will be able to work with the volumes.

Figure 6-9 shows an example of host connection relationships to volume groups. IBM i
partition number 1 has two Fibre Channel adapters, with WWPN 1 and WWPN 2. IBM i
partition number 2 has two Fibre Channel adapters, with WWPN 3 and WWPN 4. Inside the
DS8000, WWPN 1 and WWPN 2 are assigned the same volume group, volume group 1, and
WWPN 3 and WWPN 4 are assigned the same volume group, volume group 2. With this
configuration, IBM i partition 1 makes use of volumes included in volume group 1 and IBM i
partition 2 makes use of volumes included in volume group 2. Both the partitions have a dual
path to the volumes due to dual Fibre Channel attachments and assignment to the same
volume group.

IBM Power Systems


IBM i Partitions

SAN infrastructure is not


IBM i IBM i included in this diagram.
Partition 1 Partition 2
In order to ensure SAN
redundancy, we will find
at least two SAN switches.
WWPN 3
WWPN 1

WWPN 2

WWPN 4

The Fibre Channel adapters


of each partition (and those
of the DS8000 as well) are
connected to different
switches.

IBM System Storage

Volume Volume
Group 1 Group 2
for Partition 1 for Partition 2

Figure 6-9 Example of host connection relationship to volume groups

A 1-to-n relationship exists between a volume group and host connections.

92 PowerHA SystemMirror for IBM i Cookbook


6.1.9 Logical subsystems
Another logical group of volumes is the logical subsystem. It is automatically created the first
time a volume with a new logical subsystem id represented by the first two characters of the
volume id is created. Therefore all volumes whose ID starts with the same two characters
belong to the same logical subsystem (LSS). LSS have an affinity with one of the servers just
like the extents pools. Even-numbered LSS belong to server 0 and odd-numbered LSS
belong to server 1. Therefore, when we create a volume, affinity related to its extents pool and
affinity related to its id through the LSS, must match.

LSS play a lest important role for FB volumes than they do for CKD volumes. For more
information on LSS in a CKD context, refer to IBM System Storage DS8000: Architecture and
Implementation, SG24-8886. In FB context, LSS are important for some of Copy Services
usage which are detailed in next sections.

6.2 Metro Mirror


In this section, we introduce the DS8000 Metro Mirror feature when used in IBM i
environments. For a complete information, refer to Part 4, “Metro Mirror” in IBM System
Storage DS8000: Copy Services in Open Environments, SG24-6788.

6.2.1 Metro Mirror overview


Metro Mirror provides real-time copy of volumes between two DS8000. It is a synchronous
copy where write operations are completed on both DS8000 before write acknowledgements
are sent back to the source operating system. It ensures that no data is lost in the event of a
failure. Due to the synchronous copy, distance considerations between the local and the
remote DS8000 apply. No more than 300 km are allowed but, even below, write performance
becomes more and more affected when this distance increases.

Chapter 6. DS8000 Copy Services 93


Figure 6-10 illustrates this write sequence when a host requests a write operation:
1. Data is stored in the source DS8000 cache and Non Volatile Storage, to be later destaged
to the source volume.
2. Same data is sent to target DS8000, where it is stored in the target DS8000 cache and
Non Volatile Storage, to be later destaged to the target volume.
3. Write acknowledgement is sent by target DS8000 to source DS8000.
4. Write acknowledgement is sent by source DS8000 to the host.

4 Write acknowledge
Server
write
1
2
Write to secondary

LUN or LUN or
volume 3 volume
Write complete
Primary acknowledgment Secondary
(source) (target)

Figure 6-10 Metro Mirror write sequence

Note: For all IBM i volumes involved in a Metro Mirror relationship, both source and
target volumes must have the same type-model. They must have the same capacity
and the same protection flag.

6.2.2 Metro Mirror operations


This is a short review of basic operations that we can perform with Metro Mirror.

Establishing a Metro Mirror pair


This operation establishes the copy relationship between the source (or local) volume and the
target (or remote) volume. Normally, those two volumes reside on two DS8000s, but it is
possible to use the same DS8000 for tests purposes. During the initial copy, the target
volume is in a target copy pending state until all tracks are copied. At the end of the initial
copy, the target volumes is in a target full duplex state.

94 PowerHA SystemMirror for IBM i Cookbook


Some options are possible at establishment time:
No copy No data is copied from the source volume to the
target volume. You have to make sure that data is
really synchronized.
Target read With this option, the target volumes can be read by a host
while replication is active. This option is not supported by
IBM i, which must be able to write to any disk at any time.
Suspend after synchronization As soon as the target volume goes into the target full
duplex state, the replication is suspended.
Reset reserve on target The Metro Mirror replication starts even if the target
volume is reserved by a host.

Note: A PPRC path must exist from the source to the target site before creating the Metro
Mirror pair.

Suspending a Metro Mirror pair


This operation stops sending data from the source volume to the target volume. However, the
source DS8000 keeps a record of updated tracks on the source volume.

Resuming a Metro Mirror pair


This operation releases sending data from the source volume to the target volume. It makes
use of updated track records during suspend time, to send only updated data.

Terminating a Metro Mirror pair


This operation ends the replication between the source volume and the target volume and
removes this relationship. If you ever need the replication again, you have to restart it from
the beginning.

Note: After terminating the Metro Mirror pair, you can delete the related PPRC path,
depending of path usage for other pairs.

Failover and failback


Failover is related to the actions that you take to activate target volumes in case of a switch
from a production site to a backup site. From the DS8000 standpoint, actions are the same
for a planned switch and for an unplanned switch. Failover operations start after a failure or
maintenance on the initial production site. After a failover, production can run on initial
target volumes.

Failback is related to the actions that you take to come back to the nominal situation, with
production running on initial source volumes. Failback operations can start after the initial
production site failure is fixed or maintenance is finished.

On the backup site, failover function performs three steps in a single task at the volume level:
1. Terminate the original Metro Mirror relationship.
2. Establish a new relationship in the opposite direction, from the initial target to the initial
source (from the backup site to the production site).
3. Before any update to the new source volume, suspend the new Metro Mirror relationship.

After these steps, the state of the original source volume is preserved (depending on the
failure, nobody might be able to connect to the original source DS8000), and the state of

Chapter 6. DS8000 Copy Services 95


original target volumes becomes source suspended. All updates on new source volumes
are tracked.

When we are ready to switch back to the nominal situation, we can start failback. Still on the
backup site, the failback function performs various actions depending on the original source
volume state. Failback also works at the volume level:
 If the original source volume is no longer in any Metro Mirror relationship, a new Metro
Mirror relationship between the original target volume and the original source volume is
established, and all data is copied back from the backup volume to the production volume.
 If the original source volume is still a member of the appropriate Metro Mirror relationship
and has no updated tracks, only updated data on the original target volume is copied back
to the original source volume.
 If the original source volume is still a member of the appropriate Metro Mirror relationship
and has some updated tracks, updated data on the original target volume and the same
tracks as the ones that were updated on the original source volume are copied back to the
original source volume.

After the failback, the Metro Mirror direction is from the original target volume to the original
source volume, and both volumes are synchronized.

The last operation to run to recover the original environment is an another failover and
failback to perform, this time on the original production site. Figure 6-11 summarizes all these
tasks.

Primary Site (A) Secondary Site (B)

Normal operation Updates are transferred


Site A production site Source = A Target = B
Site B recovery site full duplex full duplex

When planned/unplanned outage at A: Establish with PPRC Failover


At site B: Metro Mirror Failover A full duplex Source = B
Restart application processing suspended

When site A recovered and operable. Establish with PPRC Failback


At site B: Target = A Source = B
1. Metro Mirror Failback
copy pending copy pending
2. Stop application processing

With pairs in full duplex. Establish with PPRC Failover


At site A: Source = A B full duplex
1. Metro Mirror Failover suspended
2. Start application processing
3. Metro Mirror Failback Establish with PPRC Failback
Source = A Target = B
full duplex full duplex

Figure 6-11 Metro Mirror failover and failback sequence

96 PowerHA SystemMirror for IBM i Cookbook


Note: For a planned switchover from site A to site B, and to keep data consistency at
site B, the application at site A has to be quiesced before the Metro Mirror failover
operation at site B.

6.2.3 PPRC paths and links


The peer-to-peer remote copy (PPRC) path is a logical construct on the DS8000 that defines
which physical links are used by a Metro Mirror or Global Copy relationship from a DS8000 to
another. This path is built at the logical subsystem (LSS) level. It means that Metro Mirror
relationships for all volumes in the same LSS will use the same physical links. At the DS8000
standpoint, the physical link is from one of its Fibre Channel adapters to one of the target
DS8000’s Fibre Channel adapters and can include supported channel extenders or routers.

The path creates an unidirectional association, on the source DS8000, between:


 The target DS8000 world wide node name
 The source LSS
 The target LSS
 The source IO port
 The target IO port

For bandwidth and redundancy purposes, a path can have up to eight links. The DS8000
handles failures and balances the workload across available paths. As shown on Figure 6-12,
a link can handle several paths.

LSS 0 LSS 0
LSS 1 LSS 1
Physical
LSS 2 Fibre Channel link LSS 2
LSS 3 LSS 3
.. ..
LSS 08 256 logical paths LSS 08
.. per FCP link ..
LSS nn LSS nn

Figure 6-12 PPRC paths and links

Tip: Do not forget to create a path in both directions, from production to backup DS8000
and from backup to production DS8000. Otherwise, failback operation is not possible.
Paths from backup to production can use any of the available links. There is no constraint
to use the same paths from production to backup.

Note: Although a DS8000 IO port can handle Metro Mirror traffic and host connections at
the same time, for performance reasons, we recommend not mixing them. We prefer
dedicating I/O ports to replication traffic or host connections.

Chapter 6. DS8000 Copy Services 97


6.2.4 Metro Mirror and IBM PowerHA SystemMirror for i
Within the IBM PowerHA SystemMirror for i product, Metro Mirror is used to replicate data
from an IASP to another IASP in a remote site.

Classical installation consists of a local IBM i partition and a remote IBM i partition. Both are
in a cluster, and each partition makes use of a DS8000 external storage for IASP disks. There
is no mandatory rule about SYSBAS disks. They can be internal drives or external storage.

Note: Metro Mirror relationships apply only to IASP volumes when used with IBM
PowerHA SystemMirror for i.

Figure 6-13 shows a setup overview of a Metro Mirror installation for an IASP. Both IBM i
partitions are members of a cluster, and Metro Mirror replication is in place for IASP volumes.

IBM i Cluster

DS
Metro
Mirror
IASP IASP

Local site Remote site

Figure 6-13 Metro Mirror implementation

When a switch occurs, either for a planned outage or an unplanned one, the target IASP
volumes become available due to Metro Mirror failover operations, and the remote partition
can use them for production activity.

When coming back to a normal situation is possible, failback operations can start, and the
original source partition can start production activity.

As you see in 13.1.1, “Configuring IBM i DS8000 Metro Mirror (GUI and CL commands)” on
page 264, IBM PowerHA SystemMirror for i handles all setup, switch, and switch back tasks
but one. It is not able to establish a Metro Mirror relationship between source and target
volumes. This DS8000 setup is done either through DSCLI or the Storage Manager GUI.

6.3 Global Mirror


In this section, we introduce the DS8000 Global Mirror feature when used in IBM i
environments. For a complete information, refer to Part 6, “Global Mirror,” in IBM System
Storage DS8000: Copy Services in Open Environments, SG24-6788.

98 PowerHA SystemMirror for IBM i Cookbook


6.3.1 Global Mirror overview
Global Mirror takes place when the distance from the source site to the target site is higher
than that supported for Metro Mirror replication or when synchronous replication is not an
option considering application performance impacts. However, DS8000 asynchronous
replication is not able to handle data consistency on the target site. To ensure consistency,
Global Mirror uses Flash Copy. For more information about the FlashCopy technique, see 6.5,
“FlashCopy” on page 107.

Basically, Global Mirror combines two DS8000 techniques, Global Copy and Flash Copy:
 Global Copy itself is not covered in this book, but it is almost exactly the same technique
as Metro Mirror, with two main differences:
– The replication is asynchronous. There is some delay before writes are effective on the
target site. RPO is not zero in this case. It mainly depends on network bandwidth and
latency between source and target sites. Figure 6-14 illustrates how Global Copy
operates when a host requires a write operation:
i. Data is stored in the source DS8000 cache and Non Volatile Storage, to be later
destaged to the source volume.
ii. Write acknowledgement is sent by the source DS8000 to the host.
iii. At a later time (that is, in an asynchronous manner), the source DS8000 sends the
necessary data so that the updates are reflected on the target volumes. The
updates are grouped in batches for efficient transmission.
iv. The target DS8000 returns write complete to the source DS8000 when the updates
are written to the target DS8000 cache and NVS.

b Write acknowledge
Server
write
a

LUN or LUN or
volume d volume
Write to secondary
Primary Secondary
(non-synchronously)
(source) (target)

Figure 6-14 Global Copy write sequence

– There is no guarantee that the write sequence on the target site is the same as it is on
the source site. Therefore, data is not consistent on the target site. It becomes
consistent when the application or host using the source volumes is quiesced. So data
migration is the main usage of Global Copy.

Chapter 6. DS8000 Copy Services 99


 FlashCopy takes place to help maintain consistency volumes, as described in Figure 6-15.
These are specific FlashCopy attributes required for Global Mirror:
– Inhibit target write: Protect FlashCopy target volume C from being updated by anyone
other than Global Mirror.
– Enable change recording: Apply changes only from the source volume to the target
volume that occurred to the source volume in between FlashCopy establish
operations, except for the first time when FlashCopy is initially established.
– Make relationship persistent: Keep the FlashCopy relationship until explicitly or
implicitly terminated.
– Nocopy: Do not initiate background copy from source to target, but keep the set of
FlashCopy bitmaps required for tracking the source and target volumes. These
bitmaps are established the first time that a FlashCopy relationship is created with the
nocopy attribute. Before a track in the source volume B is modified, between
Consistency Group creations, the track is copied to target volume C to preserve the
previous point-in-time copy. This includes updates to the corresponding bitmaps to
reflect the new location of the track that belongs to the point-in-time copy. Note that
each Global Copy write to its target volume within the window of two adjacent
consistency groups can cause FlashCopy I/O operations.

Host

Write I/O

F la
A B sh
Co
Source Target py
Global Copy
Copy Pending Copy Pending
C
Tertiary

O S T
O SBM: Source bitmap B B
S TBM: Target bitmap M M
OOS: Out-of-sync bitmap
Local site Remote site
Figure 6-15 Global Mirror implementation

100 PowerHA SystemMirror for IBM i Cookbook


 Consistency group creation requires three steps automatically processed by DS8000
(Figure 6-16). After step 3 is complete, C volumes represents the consistency group and
on these volumes, data are consistent.
When consistency group creation is triggered, always by the source site, three
steps occur:
a. Serialize all Global Copy source volumes. This imposes a brief hold on all incoming
write I/Os to all involved Global Copy source volumes. After all source volumes are
serialized, the pause on the incoming write I/O is released and all further write I/Os are
now noted in the change recording bitmap. They are not replicated until step 3 is done,
but application write I/Os can immediately continue.
b. Drain includes the process to replicate all remaining data that is indicated in the
out-of-sync bitmap and still not replicated. After all out-of-sync bitmaps are empty, the
third step is triggered by the microcode from the local site.
c. Now the B volumes contain all data as a quasi point-in-time copy, and are consistent
due to the serialization process in step 1 and the completed replication or drain
process in step 2. Step 3 is now a FlashCopy that is triggered by the local system’s
microcode as an inband FlashCopy command to volume B as the FlashCopy source,
and volume C as the FlashCopy target volume. Note that this FlashCopy is a
two-phase process: First, the FlashCopy command to all involved FlashCopy pairs.
Then the master collects the feedback and all incoming FlashCopy completion
messages. When all FlashCopy operations are successfully completed, the master
concludes that a new Consistency Group has been successfully created.
FlashCopy applies here only to data changed since the last FlashCopy operation. This
is because the enable change recording property was set at the time when the
FlashCopy relationship was established. The FlashCopy relationship does not end due
to the nocopy property, which is also assigned at FlashCopy establish time. Note that
the nocopy attribute results in that the B volumes are not fully replicated to the C
volumes by a background process. Bitmaps are maintained and updated instead.

Done Done Done


Start Start Start

1 Serialize all 2 3
Perform
Global Copy Drain data from local to remote site
FlashCopy
source volumes

O
C A1 B1
O
R C1
Source S Target
Tertiary

O
C A2 B2
O
R C2
Source S Target
Tertiary
Local Remote
Figure 6-16 Global Mirror consistency group formation

There are several parameters used for Global Mirror tuning. See the section “Consistency
Group interval time” in IBM System Storage DS8000: Copy Services in Open Environments,
SG24-6788. It determines how long to wait before starting with the formation of a new

Chapter 6. DS8000 Copy Services 101


consistency group (for example, performing the three steps described above). The default is
zero seconds. This means that consistency group creation happens constantly. As soon as a
consistency group is created, the creation of a new one starts immediately.

6.3.2 Global Mirror operations


This is a brief review of basic operations that we can perform with Global Mirror.

Establishing Global Copy and FlashCopy relationships


These operations are needed before Global Mirror activation.

Global Copy
This operation establishes the copy relationship between the source (or local) volume and the
target (or remote) volume. Normally, those two volumes reside on two DS8000s, but it is
possible to use the same DS8000 for tests purposes. During the copy, the target volume gets
a target copy pending state. For Global Mirror purposes, the Global Copy command with
default parameters values can be used.

Note: A PPRC path must exist from the source to the target site before creating the Global
Copy pair.

FlashCopy
Before creating the FlashCopy relationship on the target site, we recommend waiting until the
end of the Global Copy initial copy. Because of the asynchronous nature of Global Copy, you
cannot rely on the status, which will always be copy pending. However, as soon as the initial
copy step is finished you will see the out of sync tracks indicator close to zero.

Specific FlashCopy settings (Target inhibit, Record, No copy, and Persistent (which is
automatically set with Record)) are to be specified.

Creating a Global Mirror session


The session is an object that contains a reference to LSS, which is part of a Global Mirror
relationship through its volumes. The session creation operation must be done for each LSS
using an unique session ID for the entire storage image.

Adding Global Copy volumes to the Global Mirror session


Global Copy source volumes can be added at any time to the Global Mirror session.
However, as we always reference the LSS with the session, added volumes must match LSS.

Starting a Global Mirror for a session


Starting a Global Mirror requests the source DS8000 to start creating the consistency groups.

Ending a Global Mirror for a session


Ending a Global Mirror requests the source DS8000 to stop creating the consistency groups.
This operation tries to keep the last consistency group available.

Pausing and resuming a Global Mirror for a session


Pausing and resuming a Global Mirror can be used in complex situations, for example, if you
want to run tests on target sites and keep normal business on source sites. This kind of
scenario involves using a fourth set of volumes (called D volumes) on the target site.

102 PowerHA SystemMirror for IBM i Cookbook


Monitoring a Global Mirror for a session
Commands and GUI panels exist to help us monitor the Global Mirror status. For example,
showgmiroos shows how many tracks are not synchronized from the source to the target.

Terminating a Global Mirror environment


This operation requires several steps to be requested in the proper order:
1. End Global Mirror.
2. Remove Global Copy volumes from the session.
3. Remove the Global Mirror session.
4. Terminate the FlashCopy pairs.
5. Terminate the Global Copy pairs.
6. Remove the paths.

Failover
Failover operations for Global Mirror involve more steps than for Metro Mirror. The very first
one is the same as it is to reverse the Global Copy direction, so to make the target volumes
(B volumes) as source suspended. But, those volumes are not consistent. We have to play
with FlashCopy target volumes (C volumes), which are consistent. Therefore, these are the
overall operations for a failover:
1. Global Copy failover: The situation becomes as shown in Figure 6-17.

Server for
GM recovery

Failover B to A
FlashCopy

Global Copy

B
B
Source
Source
copypending Suspended C
Tertiary

Local site Remote site

Figure 6-17 Global Mirror Global Copy failover

Chapter 6. DS8000 Copy Services 103


2. Set consistent data on B volumes.
To make B volumes consistent, we use FlashCopy reverse, which copies content back to B
from C volumes. This is a background copy that updates all changes on B since the last
consistency group formation with the previous data they had stored on C volumes. This
operation terminates the FlashCopy relationship between B and C volumes. The situation
becomes as shown in Figure 6-18, but at the end there are no longer FlashCopy relationships.

Server for
GM recovery

FRR FlashCopy

B B
GC Source
Source FC Source
copypending C
Suspended Tertiary
FC Target

Local site Remote site

Figure 6-18 Global Mirror reverse FlashCopy

3. Re-establish a FlashCopy relationship between volumes B and C.


Because of the termination of the FlashCopy relationship, we have to recreate it with the
same settings that we used at the initial creation. At the end of this step we are back in the
same situation as after the first step, but with consistent data on B volumes, and we can
start host connections to these volumes.

Failback
When the source site becomes available, in a Metro Mirror environment, the first step is to run
a failback operation from the target B volumes to synchronize the content of volume A with
the content of volume B.

The second step is to make volume A source back and to start synchronizing B volumes with
A volumes.

After Global Copy and FlashCopy are working again correctly, the last operation is to restart
the Global Mirror session.

Note: Each time that a failover operation is done, applications can be quiesced before to
have consistent data with no difference between the source and target data.

6.3.3 Global Mirror and IBM PowerHA SystemMirror for i


Within IBM PowerHA SystemMirror for i product, Global Mirror is used to replicate data from
an IASP to another in a remote site, when the distance does not allow synchronous
replication with Metro Mirror.

As with Metro Mirror replication, classical installation consists of a local IBM i partition and a
remote IBM i partition. Both are in a cluster, and each partition makes use of a DS8000

104 PowerHA SystemMirror for IBM i Cookbook


external storage for IASP disks. There is no mandatory rule about SYSBAS disks. They can
be internal drives or use external storage.

Note: Global Mirror relationships apply only to IASP volumes when used with IBM
PowerHA SystemMirror for i.

Figure 6-19 shows a setup overview of a Global Mirror installation for an IASP. Both IBM i
partitions are members of a cluster, and Global Mirror replication is in place for IASP volumes.

IBM i Cluster

DS
Global
Mirror
IASP IASP
Flash-
Copy
IASP in GM

Local site Remote site

Figure 6-19 Global Mirror implementation

When a switch occurs, either in case of a planned outage or an unplanned one, the target
IASP volumes become available due to Global Mirror failover, and the remote partition can
use them for production activity.

When coming back to a normal situation is possible, failback operations can start, and the
original source partition can start back production activity.

As you see in 13.1.3, “Configuring IBM i DS8000 Global Mirror (CL commands)” on page 311,
IBM PowerHA SystemMirror for i handles all setup, switch, and switchback tasks but one. It is
not able to establish the Global Mirror configuration between source and target volumes. The
DS8000 setup is done either through DSCLI or Storage Manager GUI.

6.4 LUN-level switching


LUN-level switching is a new function provided by IBM PowerHA SystemMirror for i in
IBM i 7.1 that allows for a local high availability solution with IBM System Storage DS8000 or
DS6000 series similar to what used to be available as switched disks for IBM i internal storage.

Chapter 6. DS8000 Copy Services 105


With LUN-level switching single-copy, IASPs managed by a cluster resource group device
domain located in IBM System Storage DS8000 or DS6000 series can be switched between
IBM i systems in a cluster (Figure 6-20).

IBM i Cluster

DS8000

*SYSBAS *SYSBAS

Disk IOA Disk IOA

IASP

Production HA

Figure 6-20 LUN-level switching between servers

A typical implementation scenario for LUN-level switching is where multi-site replication using
Metro Mirror or Global Mirror is used for disaster recovery and protection against possible
storage subsystem outages, while additional LUN-level switching at the production site is
used for local high-availability protection, eliminating the requirement for a site-switch in case
of potential IBM i server outages.

106 PowerHA SystemMirror for IBM i Cookbook


For implementation of LUN-level switching, an ASP copy description needs to be created for
each switchable IASP using ADDASPCPYD, which has been enhanced with recovery domain
information for LUN-level switching (Figure 6-21).

Add ASP Copy Description (ADDASPCPYD)

Type choices, press Enter.

Logical unit name: LUN


TotalStorage device . . . . . *NONE
Logical unit range . . . . . .
+ for more values
Consistency group range . . .
+ for more values
Recovery domain: RCYDMN
Cluster node . . . . . . . . . *NONE
Host identifier . . . . . . .
+ for more values
Volume group . . . . . . . . .
+ for more values
+ for more values

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 6-21 IBM i ADDASPCPYD enhancement for LUN-level switching

An ASP session is not required for LUN-level switching, as there is no replication for the
IASP involved.

You can see a more detailed scenario in 13.1.4, “Configuring IBM i DS8000 LUN-level
switching” on page 317.

Note: Setting up an ASP copy description for LUN-level switching is only supported from
the green-screen interface.

For LUN-level switching, the backup node host connection on the DS8000 or DS6000
storage system must not have a volume group (VG) assigned. PowerHA automatically
unassigns the VG from the production node and assigns it to the backup node at site
switches or failovers.

6.5 FlashCopy
In this section we introduce the DS8000 FlashCopy feature when used in IBM i environments.
For complete information, see Part 3, “FlashCopy,” in IBM System Storage DS8000: Copy
Services in Open Environments, SG24-6788.

Chapter 6. DS8000 Copy Services 107


FlashCopy overview
FlashCopy allows you to create a point-in-time copy of the logical volume. By doing a
FlashCopy, a relationship is established between a source volume and a target volume. Both
are considered to form a FlashCopy pair. As a result of the FlashCopy, either all physical
blocks from the source volume are copied or, when using the nocopy option, only those parts
are really copied that have changed in the source data since the FlashCopy has been
established. The target volume needs to be the same size or bigger than the source volume
whenever FlashCopy is used to flash an entire volume.

Typically, large applications such as databases have their data spread across several
volumes, and their volumes should all be FlashCopied at exactly the same point-in-time.
FlashCopy offers consistency groups, which allows multiple volumes to be FlashCopied at
exactly the same instance.

Establishing a FlashCopy relationship


When the FlashCopy is started, the relationship between source and target is established
within seconds by creating a pointer table, including a bitmap for the target.

If all bits for the bitmap of the target are set to their initial values, it means that no data block
has been copied so far. The data in the target is not modified during setup of the bitmaps. At
this first step, the bitmap and the data look as illustrated in Figure 6-22.

The target volume in the following figures can be a normal volume or a virtual volume (space
efficient volume). In both cases the logic is the same.

FlashCopy established at time t0 (time-zero)

t0 tx ty tz time

bitmap

1 1 1 1 1 1

source volume target volume

data data
t0 t0 t0 t0 t0 t0

Figure 6-22 FlashCopy at time t0

108 PowerHA SystemMirror for IBM i Cookbook


After the relationship has been established, it is possible to perform read and write I/Os on
both the source and the target. Assuming that the target is used for reads only while
production is ongoing, the process will work as illustrated in Figure 6-23.

Writing to the source volume and


reading from the source and the target volume

t0 tx ty tz time
read 1 read 2

Time-zero data not yet


available in target volume;
read it from source volume.
bitmap
read write x write y write z
1 0 1 0 1 1

source volume target volume

data data
t0 tx t0 tz t0 t0 t0 t0

Before physical write to the source volume copy time-zero


data from the source volume to the target volume

Figure 6-23 Reads from source and target volumes and writes to source volume

Reading from the source


The data is read immediately (Figure 6-23).

Writing to the source


Whenever data is written to the source volume while the FlashCopy relationship exists, the
Storage system makes sure that the time-zero-data is copied to the target volume prior to
overwriting it in the source volume. When the target volume is a space-efficient volume, the
data is written to a repository.

To identify whether the data of the physical track on the source volume needs to be copied to
the target volume, the bitmap is analyzed. If it identifies that the time-zero data is not available
on the target volume, then the data will be copied from source to target. If it states that the
time-zero data has already been copied to the target volume then no further action is taken.

It is possible to use the target volume immediately for reading data and also for writing data.

Reading from the target


Whenever a read request goes to the target while the FlashCopy relationship exists, the
bitmap is used to identify whether the data has to be retrieved from the source or from the
target. If the bitmap states that the time-zero data has not yet been copied to the target, then
the physical read is directed to the source. If the time-zero data has already been copied to
the target, then the read will be performed immediately against the target (Figure 6-23).

Chapter 6. DS8000 Copy Services 109


Writing to the target
Whenever data is written to the target volume while the FlashCopy relationship exists, the
storage subsystem makes sure that the bitmap is updated. This way the time-zero data from
the source volume never overwrites updates done directly to the target volume. Figure 6-24
illustrates he concept of writes to the target volume.

Writing to the target volume

t0 ta tb time

write a write b

bitmap

0 0 1 0 1 1

source volume target volume

data data
t0 tx t0 ty t0 t0 ta t0 tb

Figure 6-24 Writes to target volume

Terminating the FlashCopy relationship


The FlashCopy relationship is automatically ended when all tracks have been copied from the
source volume to the target volume, if the FlashCopy option was to copy the data. The
relationship can also be explicitly withdrawn by issuing the corresponding commands.

A FlashCopy space-efficient relationship ends when it is withdrawn. When the relationship is


withdrawn there is an option to release the allocated space of the space efficient volume.

Full volume copy


When the copy option is invoked and the establish process completes, a background process
is started that copies all data from the source to the target. If not explicitly defined as
persistent, the FlashCopy relationship ends as soon as all data is copied.

Only the classical FlashCopy allows a full copy. FlashCopy SE has no such function. But
remember, both features can coexist.

Nocopy option
If FlashCopy is established using the nocopy option, then the result will be as shown in
Figure 6-22 on page 108, Figure 6-23 on page 109, and Figure 6-24.

The relationship will last until it is explicitly withdrawn or until all data in the source volume has
been modified. Blocks for which no write occurred on the source or on the target will stay as
they were at the time when the FlashCopy was established. If the persistent FlashCopy
option was specified, the FlashCopy relationship must be withdrawn explicitly.

110 PowerHA SystemMirror for IBM i Cookbook


6.6 FlashCopy SE
PowerHA for SystemMirror for i with IBM i 7.1 newly supports space-efficient FlashCopy of
the IBM System Storage DS8000 series.

The IBM System Storage DS8000 series FlashCopy SE licensed feature allows creation of
space-efficient FlashCopy target volumes that can help to reduce the required physical
storage space for the FlashCopy target volumes. These volumes are typically needed only for
a limited time (such as for the duration of a backup to tape).

A space-efficient FlashCopy target volume has a virtual storage capacity reported to the host
matching the physical capacity of the fully provisioned FlashCopy source volume, but no
physical storage space is ever allocated. Physical storage space for space-efficient
FlashCopy target volumes is allocated in 64 KB track granularity. This is done on demand for
host write operations from a configured repository volume shared by all space-efficient
FlashCopy target volumes within the same DS8000 extent pool (Figure 6-25).

Non-provisioned
space-efficient volumes
(no space ever allocated)

FlashCopy source Space-efficient FlashCopy source


Space- efficient
(fully-provisioned) FlashCopy (fully-provisioned)
FlashCopy
target target

Repository Volume
(over-provisioned,
e.g., 500 GB virtual &
100 GB real capacity)

Figure 6-25 DS8000 space-efficient FlashCopy

From a user perspective, the PowerHA setup (not the DS8000 FlashCopy setup) for
space-efficient FlashCopy is identical to the setup for traditional FlashCopy with the no-copy
option. The reason for this is that PowerHA SystemMirror for i internally interrogates the
DS8000 to determine the type of FlashCopy relationship and makes sure that it uses the
corresponding correct DSCLI command syntax. The syntax check is done for either
traditional FlashCopy or FlashCopy SE when issuing mkflash and rmflash.

For further information about using IBM System Storage DS8000 FlashCopy SE with IBM i,
see IBM Redbooks publication IBM System Storage Copy Services and IBM i: A Guide to
Planning and Implementation, SG24-7103.

Chapter 6. DS8000 Copy Services 111


112 PowerHA SystemMirror for IBM i Cookbook
7

Chapter 7. Storwize V7000 and SAN Volume


Controller Copy Services
This chapter provides an overview of the IBM Storwize V7000 and SAN Volume
Controller storage concepts and their Copy Services functions Metro Mirror, Global Mirror,
and FlashCopy.

© Copyright IBM Corp. 2012. All rights reserved. 113


7.1 Storwize V7000/SAN Volume Controller storage concepts
IBM Storwize V7000 and IBM System Storage SAN Volume Controller both provide in-band
block-level storage virtualization. The key concept of storage virtualization is to decouple the
logical storage representation as seen by the application servers from the underlying physical
storage by abstracting the physical location of the data. This abstraction is achieved by the
SVC/V7000 using a storage virtualization-layer, which also enables application server
transparent advanced functions for virtualized volumes like volume mirroring, volume
migration, FlashCopy creation, or thin provisioning. These advanced functions do not rely in
any way on the functionality provided by the underlying disk storage system.

7.1.1 Hardware overview


The SAN Volume Controller (SVC) is a highly scalable storage virtualization appliance that
provides the capability to virtualize external SAN-attached storage. Apart from optional
internal Solid State Drives (SSDs), the SVC has no integrated internal storage. An SVC
appliance consists of at least two and up to eight SVC nodes, each with their own dedicated
uninterruptible power supply (UPS) building an SVC cluster. Within the cluster there are
always two nodes, each paired into an I/O group that describes a node-pair with redundant
write cache responsible for processing the I/O requests for the volumes associated with that
particular I/O group. The latest hardware generation of the SVC, IBM device type 2145,
consists of model CG8 nodes based on IBM System x3550 M3 quad-core servers, each
providing 24 GB cache, four 8 Gb FC ports, and optionally either up to four SSDs or a
dual-port 10 Gb Ethernet card.

The Storwize V7000 is a modular storage system that includes the capability to virtualize
external SAN-attached storage in addition to its own internal storage. The V7000 is built upon
the IBM SAN Volume Controller (SVC) technology base using corresponding storage
concepts and management interfaces with generally the same command set. From a
hardware perspective, the V7000 IBM device type 2076 consists of two dual-active
controllers with 8 GB of cache each, eight 8 Gb FC ports plus four optical 10 Gb FCoE ports,
up to 24 disks in the controller enclosure, and up to nine SAS-chain attached expansion
enclosures with up to 24 disks each, for a supported maximum of 240 internal small form
factor (SFF) disk drives. With the release of V6.2 microcode, a V7000 clustering option was
introduced to increase the V7000 scalability by allowing you to combine two V7000 systems
into a V7000 cluster with two I/O groups.

114 PowerHA SystemMirror for IBM i Cookbook


7.1.2 Storage virtualization
SVC and V7000 share the same storage virtualization concept as that shown in Figure 7-1.
They implement an indirection, or virtualization, layer in a Fibre Channel fabric. The
virtualization layer mediates between the physical storage in RAID controllers presented as
SCSI LUNs known as managed disks or MDisks to the SVC/V7000 and virtual disks (VDisks)
or volumes presented to the application servers or hosts. The SVC/V7000 pools the managed
physical storage and provides parts of that storage pool or MDisk group to application
servers, in a way that makes the servers think that they are just using disks in the normal way.

IBM IBM Application


Servers

VD 7
VD 1
VD 2
VD 3

VD 4

VD 5

VD 6
Volumes / Virtual disks
Virtual-to-physical Mapping SVC / V7000
High Perf Low Cost Storage Virtualization
Storage pools / MDisk groups
MD 1

MD 2

MD 3

MD 4

MD 5

MD 6

MD 7

MD 8
Managed disks
LUN 1

LUN 2

LUN 3

LUN 4

LUN 1

LUN 2

LUN 3

LUN 4
SCSI LUNs

RAID RAID
controller 1 controller 2 SAN Storage

Figure 7-1 SVC and V7000 storage virtualization

All virtual disks seen by the application servers are mapped to the physical storage in the pools
using a virtualization map that is fully transparent to the servers. Using this virtualization map,
the SVC/V7000 can duplicate the server data to make instant copies or copies at a remote
site, mirror data onto multiple physical disks to improve availability, migrate data concurrently
between physical storage systems, or over allocate its physical storage to provide
thin-provisioned virtual disks to the application servers and save physical storage cost.

Note: IBM i supports SVC/V7000 thin-provisioned or space-efficient volumes only for


FlashCopy target volumes.

The SCSI LUNs created on the physical storage systems typically configured as a single LUN
per RAID array are discovered as MDisks by the SVC/V7000 and assigned by the storage
administrator to storage pools/MDisk groups usually based on common performance and
availability device characteristics. Up to 128 storage pools with up to 128 MDisks each and up
to 4096 MDisks per cluster from up to 64 storage systems are supported. At creation of the
storage pool or MDisk group, an extent size ranging from 16 MB - 8 GB has to be specified,
which is used by the SVC/V7000 to address the storage pool and determines the maximum

Chapter 7. Storwize V7000 and SAN Volume Controller Copy Services 115
managed storage capacity in the cluster (from 64 TB to 32 PB) but is not relevant for the
cluster’s storage performance itself. Virtual disks (VDisks) are created from the extents of a
storage pool with up to 2048 VDisks supported per I/O group, that is, cluster node pair.

VDisks are created as either image mode, sequential mode, or striped mode VDisks
(Figure 7-2). By default, VDisks are created in striped mode, meaning that the extents for the
VDisks are allocated in a round-robin fashion from the available MDisks within the MDisk
group. That is, the strip size is identical to the extent size. Sequential mode VDisks have their
extents allocated sequentially from the available MDisks, meaning that another MDisk is used
for VDisk extent allocation only if the previous MDisk is already fully allocated. Image mode
VDisks represent a one-to-one mapping between MDisks and VDisks and are typically used
for migrating storage to (or from) the SVC/V7000.

Image Mode:
A Virtual Disk = Physical LUN
Sequential Mode:
B Virtual Disk mapped sequentially to
a portion of a single managed disk
C
Striped Mode:
I/O Group A I/O Group B Virtual Disk striped
across multiple managed disks

MDG1 MDG3

MDG2

Figure 7-2 SVC and V7000 VDisk modes

7.1.3 I/O processing


As previously mentioned, the SVC/V7000 processes host I/O for a particular volume (VDisk)
only within an I/O group consisting of two cluster nodes, which also serve as a backup for
each other in case one cluster node becomes unoperational, for example, during microcode
updates or failure conditions. Under normal conditions, I/O for a specific volume is always
processed by the same node designated as the preferred node. When VDisks are created for
an I/O group, they are automatically evenly distributed in a round-robin fashion between both
nodes of the I/O group.

For host write I/O processing, both nodes of an I/O group are involved (Figure 7-3 on
page 117) such that the host sends its write I/O request down the preferred path to the
preferred node, and the SVC/V7000 preferred node stores the write data in its UPS
backed-up write cache and propagates a second copy of the write data via the SAN fabric to
the alternate (non-preferred) node in the I/O group. After the write data has been stored in
both nodes’ write cache, it is acknowledged by an I/O complete message to the host and
asynchronously destaged by the preferred node to the physical disks. In case one of the
nodes of an I/O group becomes unoperational, the remaining node goes into write-through
mode, meaning that because the write cache redundancy has been lost, the host write I/O is
not acknowledged until it has been written to the physical disks with corresponding host write
I/O performance implications.

116 PowerHA SystemMirror for IBM i Cookbook


Write I/O request

1 I/O Group 1
V1 V2

Preferred Alternative
node node
2
UPS1 SVC node 1 2 SVC node 2 UPS2

Cache Cache

MD1 MD2 MD3

Figure 7-3 SVC and V7000 I/O processing

Because both nodes of an I/O group are responsible for I/O processing, it is important that the
host systems are always zoned to both nodes of an I/O group and have the SDDPCM
multi-pathing software for SVC/V7000 installed.

One of the advanced functions enabled by the SVC/V7000 virtualization-layer is volume


(VDisk) mirroring for which two copies1 of extents are kept for a volume (VDisk), but the
volume itself is still represented only once to the application servers. For availability reasons,
the two copies are typically allocated on MDisks in different MDisk groups, which are ideally
from two different storage systems. As volume mirroring is implemented in the SVC/V7000
I/O stack below the cache and Copy Services, neither the host application servers nor Copy
Services functions are aware that the volume is mirrored so that any Copy Services or
migration functions work the same for mirrored as for non-mirrored volumes. Important from
an I/O processing point of view is that under normal conditions (that is, both volume copies
are available), host I/O writes are directed to both copies, but read I/O is directed only to the
primary volume, which is by default the first created copy. The location of the primary volume
can also be changed by the user to either account for load-balancing or possibly different
performance characteristics for the storage of each copy.

Regarding the I/O to the disk storage system, the SVC/V7000 uses an integrated multi-path
driver within its microcode to direct the I/O for a specific LUN to a single storage system port
using a round-robin algorithm to distribute the workload from different LUNs to different
storage system ports. All LUNs that are part of an MDisk group or storage pool (that is, that
are managed MDisks) must be available. Otherwise, the complete MDisk group is taken
offline, except for an unavailable image mode MDisk, which does not impact the availability of
other MDisks.

Due to these I/O processing characteristics of the SVC/V7000, certain considerations are
required to ensure a proper SAN switch zoning without compromising redundancy or
performance. See 8.1.5, “Storage considerations” on page 139, for further information about
IBM PowerVM Virtual I/O Server requirements for SVC/V7000 attachment.

1
In SVC/V7000 terminology, a copy refers to one instance of an entity like a volume, that is, a mirrored volume has
two copies.

Chapter 7. Storwize V7000 and SAN Volume Controller Copy Services 117
7.1.4 Copy Services
The SVC/V7000 offers a common platform and single point of control not only for regular
provisioning and management of heterogeneous storage but also for advanced functions,
such as Copy Services, that are enabled by the SVC/V7000 virtualization-layer also between
storage systems of different architectures or from different vendors. Copy Services functions
available for the SVC/V7000 are Metro Mirror for synchronous remote replication, Global
Mirror for asynchronous remote replication, and FlashCopy for point-in-time volume copies
that are each discussed in more detail in the following sections.

The SVC/V7000 Copy Services functions are licensed on a used versus installed capacity
basis with a single combined license for Metro Mirror and Global Mirror. FlashCopy is
included in the V7000 base license (5639-VM1), whereas a separate license for FlashCopy is
required for the SVC based on the virtual capacity only of the source volumes.

Both intra-cluster (that is, within the same cluster (and within the same I/O group only, rather
than used for testing or migration purposes)) and inter-cluster remote copy relationships (that
is, between two SVC or V7000 systems) are supported by the SVC/V7000.

For inter-cluster remote Copy Services, two SVC or V7000 clusters are connected to each
other over a Fibre Channel fabric. The maximum supported distance for remote Copy
Services is determined by the maximum supported round-trip latency of 80 ms. Up to 10 km
Fibre Channel distance is supported with long-wave SFPs. Extended distance support is
provided by using extenders, routers, or DWDM/CWDM hardware with up to seven hop
counts supported.

The source and target volumes for Copy Services relationships need to be of the same
virtual size.

Interoperability between SVC and V7000 Copy Services is available with the latest
SVC/V7000 V6.3 microcode and is also supported by 5799-HAS PRPQ PowerHA
SystemMirror for i. Allowing a V7000 system to participate in a SVC remote copy cluster
partnership requires its cluster layer property to be changed from storage to appliance.

For further details about SVC and V7000 refer to these IBM Redbooks publications:
 Implementing the IBM System Storage SAN Volume Controller V6.1, SG24-7933
 Implementing the IBM Storwize V7000, SG24-7938

118 PowerHA SystemMirror for IBM i Cookbook


7.2 Metro Mirror
Metro Mirror is a synchronous remote copy relationship between two SVC/V7000 volumes
(VDisks) of equal (virtual) size. When a remote copy relationship (either Metro Mirror or
Global Mirror) is established, the preferred primary volume is designated as the master
volume and the preferred secondary volume as the auxiliary volume. As long as the
secondary volume is available, every host write I/O sent to the Metro Mirror primary volume is
acknowledged back to the host only after it has been committed to the write cache of the
primary SVC/V7000 and the secondary SVC/V7000 system (Figure 7-4).

Host

1 Write 4 Ack
3 Ack Write
Cache Cache
2 Mirror Write

Master Auxiliary
volume Metro Mirror Relationship volume
(primary (secondary
role) role)

Figure 7-4 Metro Mirror SVC/V7000 write I/O processing

The role of a master or auxiliary volume is either primary or secondary, depending on the
direction or failover state of the current remote copy relationship. Up to 2048 remote copy
relationships are supported in a two-node SVC/V7000 cluster.

With SVC/V7000, establishing a Copy Services relationship is done in two phases of creating
the relationship first before starting it in a second step. This is different from, for example, IBM
System Storage DS8000 Copy Services, for which establishing a Copy Services relationship
is done in a single step by creating the out-of-sync bitmaps and starting the relationship
automatically at its creation.

When creating the Metro Mirror relationship, the user can specify whether the auxiliary
volume is already in sync with the master volume, and the background copy process is then
skipped. The in-sync option (path 1a in Figure 7-5 on page 120) is intended to be used when
the volumes were created with the format option should not be used for IBM i volumes
because IBM i specially formats the volumes itself when they are configured (that is, added to
an IBM i ASP). Hence, using the SVC/V7000 format option at volume creation and therefore
the in-sync option when creating a Metro Mirror relationship does not make sense.

7.2.1 Bandwidth thresholds


At initial synchronization, data is copied in data chunks, called grains, of 256 KB by a
background copy process from the primary to secondary remote copy volume with a default
bandwidth limit of 50 MBps between both SVC/V7000 clusters. This partnership bandwidth
limit is evenly divided by the nodes in the cluster and is an attribute of the remote copy
partnership between both SVC/V7000 systems. The bandwidth should be less than (when
still accounting for host write I/O updates during synchronization) or equal to the available
replication link bandwidth between both systems, when no relevant host write update are
expected during synchronization. Also, this overall cluster partnership bandwidth should be
chosen deliberately to not exceed the capabilities of the primary and secondary storage

Chapter 7. Storwize V7000 and SAN Volume Controller Copy Services 119
systems to prevent performance impacts for the foreground host I/O. Additionally, there
is also a relationship bandwidth limit for the maximum background copy rate for each
remote copy relationship hat defaults to 25 MBps and is an attribute of the SVC/V7000
cluster configuration.

7.2.2 Remote copy relationship states


A Metro Mirror or Global Mirror remote copy volume relationship can be in one of these states:
 Consistent stopped
 Inconsistent stopped
 Consistent synchronized
 Inconsistent copying
 Idling

Figure 7-5 shows an overview of these states as they apply to a connected remote copy
relationship and the conditions that cause a state transition.

Create
1a Remote Mirror 1b
) Relationship (o u
s y nc t of
syn
(in c)

Consistent Inconsistent
Stopped Stopped

Fo
r
Stop ( o u ce d Stop
t o Sta
fs r
2a Start or yn t 2b Start or
Error c) Error
Stop
4b enable
access 3
Consistent Inconsistent
Synchronized Background copy complete Copying

Start
5a
(in sync)
art
Stop
d St c)
e n
4a enable rc sy
access Fo t of
u
(o

Idling

Figure 7-5 SVC/V7000 remote copy volume states and transitions

120 PowerHA SystemMirror for IBM i Cookbook


The remote copy states can be described as follows:
Inconsistent stopped State after creating a remote copy relationship (without using
the in sync option) or after a failure condition that occurred
while the relationship was in inconsistent copying state.
Secondary volume data is not consistent with primary volume
data and due to the risk of (undetected) data inconsistency
should not be accessed by an application server.
Inconsistent copying State after starting an inconsistent stopped or idling
relationship with changes to be synchronized with a
background copy process running to copy data from the
primary to the secondary volume. The primary is accessible for
read and write I/O, but the secondary is offline (that is, not
accessible for either read or write I/O), while the background
copy is running and the relationship is not consistent.
Consistent synchronized State of an inconsistent copying relationship after completion
of the background copy process or of a restarted consistent
stopped relationship. The primary volume is accessible for read
and write I/O, but the secondary volume is accessible only for
read I/O. A switch of the remote copy direction does not
change this state.
Stopping the relationship takes it to the consistent stopped
state.
Stopping the relationship with the -access parameter takes it to
the idling state.
Switching the relationship leaves it in the consistent
synchronized state but reverses the primary and secondary
roles.
Consistent stopped State after stopping a consistent synchronized relationship or
after it encountered an error that forced a consistency freeze.
The secondary contains a consistent image but might be
out-of-date with respect to the primary, which might have
received write updates from the host after the relationship
entered this state. Restarting on a non-synchronized
relationship that has had changes requires the -force CLI
command parameter.
Idling State after stopping a consistent synchronized relationship
with enabling write access to the secondary volume. Both
master and auxiliary volumes operate in the primary role so
that both master and auxiliary volumes are accessible for read
and write I/O.
Changes are tracked for both the master and auxiliary volumes
so that when starting the remote copy relationship again in the
desired direction, specified by the required -primary CLI
command parameter, only a partial synchronization for the
changed grains is needed. Restarting on a non-synchronized
relationship that has had changes requires the -force CLI
command parameter.

In addition to these states that are valid for a connected remote copy relationship (that is, one
where the primary system is still able to communicate with the secondary system), there is
also a disconnected state of remote copy relationships where the primary system can no
longer communicate with the secondary system. When the clusters can communicate again,
the relationships automatically become connected again.

Chapter 7. Storwize V7000 and SAN Volume Controller Copy Services 121
If the relationship or consistency group becomes disconnected, the primary volumes
transition to inconsistent disconnected. The master side transitions to idling disconnected.

The SVC/V7000 logs informational events like remote copy relationship changes, loss of
synchronization, or remote cluster communication errors in an error log for which SNMP
traps, e-mail notification, or syslog server messages can be configured to trigger either
automation or alert the user for manual intervention.

7.2.3 Consistency groups


Consistency is a concept of the storage system ensuring write I/O processing in exactly the
same order in which write updates are received by the host and maintaining this order even
with Copy Services relationships. It applies to a single relationship, but can also be applied to
a set of relationships spanning multiple volumes by using consistency groups.

For Metro Mirror and Global Mirror remote copy relationships, maintaining the correct write
order processing requires that in case of an error event causing loss of replication for only a
subset of remote copy relationships of an application servers, remote write update processing
for the non-affected remote copy relationships is automatically stopped to ensure application
server data consistency at the secondary site.

For FlashCopy point-in-time volume copies, maintaining consistency and correct write order
processing requires that write I/O to all volumes of an application server in a consistency
group is temporarily put on hold until all FlashCopy volume relationships for the consistency
group have been started. The storage system depends on the concept of dependent writes
having been implemented in the application logic to ensure consistency across multiple
volumes in a consistency group (for example, that a journal is updated with the intended
database update before the database itself is actually updated). This application logic write
dependency ensures that when a SCSI queue full status is set as part of the consistency
group formation for a volume, further dependent application writes are put on hold by the
application so that the storage system can proceed setting SCSI queue full status for all
remaining volumes and therefore guarantee dependent write data consistency for all volumes
in the consistency group. This write dependency concept still applies for IBM i with its
single-level storage architecture, as IBM i SLIC storage management would hold off all I/O to
a disk unit in an SCSI queue full condition but would not stop the I/O to other disk units that
are still available for I/O operations.

Up to 127 FlashCopy consistency groups with up to 512 FlashCopy volume relationships in a


consistency group are supported in a SVC/V7000 system. For Metro Mirror and Global Mirror,
up to 256 remote mirror consistency groups are supported with no limit imposed for the
number of either Metro Mirror or Global Mirror remote copy relationships other than the limit
of 2048 volumes supported per I/O node pair.

PowerHA SystemMirror for i with the 5799-HAS PRPQ inherently uses consistency groups
for SVC/V7000 FlashCopy relationships and requires them to be configured for Metro Mirror
or Global Mirror relationships. Due to the IBM i single-level storage architecture, which stripes
the data across all disk units of an ASP, consistency groups should be defined on an IASP
group level.

Note: Stand-alone volume copy relationships and consistency groups share a common
configuration and state model. That is, all volume copy relationships in a consistency
group that is not empty have the same state as the consistency group.

122 PowerHA SystemMirror for IBM i Cookbook


7.3 Global Mirror
Global Mirror is an asynchronous remote copy relationship between two SVC/V7000 volumes
(VDisks) of equal (virtual) size. When a remote copy relationship (either Metro Mirror or
Global Mirror) is established, the preferred primary volume is designated as the master
volume and the preferred secondary volume as the auxiliary volume. Every host write I/O
sent to the Global Mirror primary volume is acknowledged back to the host after it has been
committed to the write cache of both nodes for the corresponding I/O group of the primary
SVC/V7000 system. At some later point in time (that is, asynchronously), this write update is
sent by the primary SVC/V7000 to the secondary SVC/V7000 system (Figure 7-6). Global
Mirror therefore provides the capability to perform remote copy over long distances, up to the
maximum supported round-trip latency of 80 ms, exceeding the performance-related
limitations of synchronous remote copy without host write I/O performance impacts caused by
remote replication delays.

Host

1 Write 2 Ack
4 Ack Write
Cache Cache
3 Mirror Write

Master Auxiliary
volume Global Mirror Relationship volume
(primary (secondary
role) role)

Figure 7-6 Global Mirror SVC/V7000 write I/O processing

Though the data is sent asynchronously from the primary to the secondary SVC/V7000, the
write ordering is maintained by sequence numbers assigned to acknowledged host write I/Os
and with the secondary applying writes in order by their sequence number. Consistency of
the data at the remote site is maintained at all times. However, during a failure condition the
data at the remote site might be missing recent updates that have not been sent or that were
in-flight when a replication failure occurred, so using journaling to allow for proper crash
consistent data recovery is of key importance, just like we generally recommend ensuring
crash consistent data recovery even when not using remote replication to be able to recover
from a loss of transient modified data in IBM i main store after a potential sudden server crash
due to a severe failure condition.

Global Mirror volume relationship states and transitions are identical to those for Metro Mirror,
as described previously in 7.2.2, “Remote copy relationship states” on page 120.

A log file is utilized by the SVC/V7000 for Global Mirror to maintain write ordering and help
prevent host write I/O performance impacts when the host writes to a disk sector that is either
currently in the process of being transmitted or due to bandwidth limits is still waiting to be
transmitted to the remote site. The SVC/V7000 also makes use of shared sequence numbers
to aggregate multiple concurrent (and dependent) write I/Os to minimize its Global Mirror
processing overhead.

Global Mirror link tolerance


Global Mirror uses a special link tolerance parameter defined at the cluster level that specifies the
duration with a default of 300 seconds, for which inadequate intercluster link performance with

Chapter 7. Storwize V7000 and SAN Volume Controller Copy Services 123
write response times above 5 ms is tolerated. If this tolerated duration of degraded performance
where the SVC/V7000 needs to hold off writes to the primary volumes with the effect of
synchronous replication-like degraded performance is exceeded, it stops the most busy active
Global Mirror relationship consistency group to help protect the application server’s write I/O
performance and logs an event with error code 1920. The link tolerance can be disabled by the
user setting its value to 0. However, this provides no protection for the application server’s write
I/O performance anymore in cases where there is congestion on either the replication link or the
secondary storage system. While the link tolerance setting allows you to define a period of
accepted performance degradation, it is still important to properly size the remote copy replication
bandwidth for the peak write I/O throughput and possible re-sync workload, and to help prevent
longer production workload performance impacts.

The concept of consistency groups to guarantee write-dependent data consistency applies to


Global Mirror the same as previously described for Metro Mirror. Consistency groups are
required to be configured for PowerHA SystemMirror for i with the 5799-HAS PRPQ.

7.4 FlashCopy
The SVC/V7000 FlashCopy function provides the capability to perform a point-in-time copy of
one or more volumes (VDisks). In contrast to the remote copy functions of Metro Mirror and
Global Mirror, which are intended primarily for disaster recovery and high-availability
purposes, FlashCopy is typically used for online backup or creating a clone of a system or
IASP for development, testing, reporting, or data mining purposes.

FlashCopy is supported also across heterogeneous storage systems attached to the


SVC/V7000, but only within the same SVC/V7000 system. Up to 4096 FlashCopy
relationships are supported per SVC/V7000 system, and up to 256 copies are supported per
FlashCopy source volume. With SVC/V7000 V6.2 and later, a FlashCopy target volume can
also be a non-active remote copy primary volume, which eases restores from a previous
FlashCopy in a remote copy environment by using the FlashCopy reverse function.

PowerHA SystemMirror i with the 5799-HAS PRPQ supports these SVC/V7000 FlashCopy
functions, which are discussed in more detail below:
 FlashCopy no-copy and background copy
 Thin-provisioned (space-efficient) FlashCopy targets
 Incremental FlashCopy
 Reverse FlashCopy
 Multi-target FlashCopy (by using separate ASP copy descriptions for each target)

124 PowerHA SystemMirror for IBM i Cookbook


7.4.1 I/O indirection
The FlashCopy indirection layer, which is logically located below the SVC/V7000 cache, acts
as an I/O traffic director for active FlashCopy relationships. To preserve the point-in-time copy
nature of a FlashCopy relationship, the host I/O is intercepted and handled according to
whether it is directed at the source volume or at the target volume, depending on the nature of
the I/O read or write and whether the corresponding grain has already been copied.
Figure 7-7 and Figure 7-8 illustrate the different processing of read and write I/O for active
FlashCopy relationships by the indirection layer.

Reads
FlashCopy Bitmap
Source Target 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 0 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 0 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1

• Reads from source processed normally


• Reads from target to grains that have been
copied from source are read from the target
• Reads from target to grains that have NOT
been copied from source are actually read
from source
Figure 7-7 SVC/V7000 FlashCopy read processing

Writes
FlashCopy Bitmap
Source Target 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 0 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 0 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1

• Attempts to write data already copied proceed as normal


• Attempts to write a source grain not already copied
intercepted and source grain copied to target before
update occurs ("copy on write")
• Writes to the target volume proceed and the FlashCopy
bitmap is updated to prevent the source grain from being
copied to the target volume

Figure 7-8 SVC/V7000 FlashCopy write processing

While a fixed grain size of 256 KB is used for remote mirror volume relationships for
FlashCopy, the user can choose from the default grain size of 256 KB or alternatively from the
smaller grain size of 64 KB as the granularity for tracking and managing out-of-sync data of a
FlashCopy relationship.

Chapter 7. Storwize V7000 and SAN Volume Controller Copy Services 125
The concept of consistency groups to ensure that dependent write data consistency across
multiple volume copy relationships applies to FlashCopy (see 7.2.3, “Consistency groups” on
page 122).

7.4.2 Background copy


A FlashCopy relationship can either be a no-copy or a background copy relationship. With a
background copy relationship, any grain from the source volume is copied to the target
volume. By default (that is, if not specifying the autodelete option), the relationship is retained
even after all grains have been copied. For a no-copy relationship, only grains that have been
modified on the source after starting the FlashCopy relationship are copied from the source
volume to the target volume prior to the source grain being allowed to be updated (copy on
write processing), provided that the corresponding grain on the target volume has not been
updated already by the host accessing the target volumes.

An option for FlashCopy is creating an incremental FlashCopy relationship, which uses


background copy to copy all of the data from the source to the target for the first FlashCopy
and then only the changes hat occurred since the previous FlashCopy for all subsequent
FlashCopies being started for the relationship.

When creating a FlashCopy relationship, the user can specify a desired copy rate for the
background copy process, which can either be 0 (meaning that a FlashCopy no-copy
relationship without a background copy is established) or any value from 1 - 100, which
translates to the desired background copy throughputs (Table 7-1).

Table 7-1 FlashCopy background copy rates


Copy rate value Data copied 256 KB grains/s 64 KB grains/s

1 - 10 128 KBps 0.5 2

11 - 20 256 KBps 1 4

21 - 30 512 KBps 2 8

31 - 40 1 MBps 4 16

41 - 50a 2 MBps 8 32

51 - 60 4 MBps 16 64

61 - 70 8 MBps 32 128

71 - 80 16 MBps 64 256

81 - 90 32 MBps 128 512

91 - 100 64 MBps 256 1024


a. Default value

126 PowerHA SystemMirror for IBM i Cookbook


A FlashCopy relationship is established on a SVC/V7000 system in three steps:
1. Creating a FlashCopy relationship
This triggers the internal creation of a FlashCopy out-of-sync bitmap used by the
SVC/V7000 for tracking the grains needing to be copied.
2. Preparing a FlashCopy relationship or consistency group
This achieves consistency for the volumes by destaging the source volume’s modified
data to disk, putting it in write-through mode, discarding the target volume’s cache data,
and rejecting any I/O to the target volume.
3. Starting a FlashCopy relationship or consistency group
This briefly pauses the I/O to the source volumes until all reads and writes below the
SVC/V7000 cache layer have completed and starts the actual FlashCopy relationship.
That is, the logical dependency between the source and target volume is established in
the SVC/V7000 indirection layer.

7.4.3 FlashCopy relationship states


A FlashCopy volume relationship can be in any of the following states:
 Idle
 Copied
 Copying
 Stopped
 Stopping
 Suspended
 Preparing
 Prepared

The FlashCopy states are described here:


Idle or copied Mapping between source and target volume exists, but the source and the
target behave as independent volumes.
Copying Background copy process is copying grains from the source to the target.
Both the source and the target are available for read and write I/O, but the
target depends on the source for grains not yet copied yet.
Stopped FlashCopy relationship was stopped either by the user or by an I/O error.
The source volume is still assessable for read and write I/O, but the target
volume is taken offline as data integrity is not provided. From a stopped
state, the relationship can either be started again with the previous
point-in-time image or lost or deleted if it is not needed anymore.
Stopping Relationship is in the process of transferring data to a depend relationship.
The source volume remains accessible for I/O, but the target volume
remains online if the background copy process completed or is put offline if
the background copy process has not completed while the relationship was
in copying state. Depending on whether the background copy completed,
the relationship moves either to the idle/copied state or the stopped state.
Suspended State of a relationship when access to metadata used by the copy process
got lost. Both the source and the target are taken offline and the
background copy is put on hold. When metadata becomes available again,
the relationship returns to the copying state or the stopping state.
Preparing Source volume is placed in write-through mode and modified data of the
source volume is destaged from the SVC/V7000 cache to create a

Chapter 7. Storwize V7000 and SAN Volume Controller Copy Services 127
consistent state of the source volume on disk in preparation for starting the
relationship. Any read or write data associated with the target volume is
discarded from the cache.
Prepared Relationship is ready to be started with the target volume having been
placed in offline state. Write performance for the source volume can be
degraded, as it is in write-through mode.

7.4.4 Thin-provisioned FlashCopy


In addition to regular, full-provisioned volumes (VDisks), which at their creation have the full
physical storage capacity allocated corresponding to their volume capacity, the SVC/V7000
also supports thin-provisioned or space-efficient volumes, which are created with a virtual
capacity reported to the host higher than the actual physical capacity pre-allocated to the
volume. If the thin-provisioned volume is created with the autoexpand option, additional
extents up to the virtual capacity of the volume are automatically allocated from the storage
pool (MDisks group) as needed when the currently allocated physical capacity gets exhausted.

Note: Using thin-provisioned SVC/V7000 volumes for IBM i is only supported for
FlashCopy target volumes, not for volumes directly assigned to an IBM i host.

Using thin-provisioned volumes also does not make sense for remote copy secondary
volumes, which become fully allocated at initial synchronization. Similarly, using
thin-provisioned volumes for FlashCopy targets only makes sense for FlashCopy no-copy
relationships that are used for a limited duration and therefore have limited changes only.

For optimal performance of thin-provisioned FlashCopy, the grain size for the thin-provisioned
volume used as the FlashCopy target should match the grain size of the FlashCopy
relationship. Additionally, for space-efficiency reasons, to help minimize physical storage
allocations on the thin-provisioned target, consider also using the small grain size of 64 KB.

128 PowerHA SystemMirror for IBM i Cookbook


7.4.5 Multi-target and reverse FlashCopy
As previously mentioned, a single FlashCopy source volume supports up to 256 target
volumes. Creating and maintaining multiple targets from a FlashCopy source volume from
different times might be useful, for example, to have multiple points to restore from by using
the reverse FlashCopy function (Figure 7-9).

Original t0
Source Target
multi-target Volume B
Volume A
relationships
t0+1 Older target
depends upon
newer target

Target
Volume C

Later recovery
2 Reverse
by reverse FlashCopy
FlashCopy operation
Source Target
Volume A Volume B

1 Optional copy of
original source OR

Target
Volume C
Target
Volume D

Figure 7-9 SVC/V7000 multi-target and reverse FlashCopy

A key advantage of the SVC/V7000 reverse FlashCopy function is that it does not destroy the
original source and target relationship, so any processes using the target (such as tape
backup jobs) can continue to run uninterrupted. It does not require a possible background
copy process to have completed, and regardless of whether the initial FlashCopy relationship
is incremental, the reverse FlashCopy function only copies data from the original target to the
source for grains that were modified on the source or target.

Consistency groups cannot contain more than one FlashCopy relationship with the same
target volume, so they need to be reversed by creating a set of new reverse FlashCopy
relationships and adding them to the new reverse consistency group.

The SVC/V7000 performs the copy on write processing for multi-target relationships in a way
that data is not copied to all targets, but only once from the source volume to the newest target
so that older targets refer to newer targets first before referring to the source. Newer targets
together with the source can therefore be regarded as composite source for older targets.

Chapter 7. Storwize V7000 and SAN Volume Controller Copy Services 129
130 PowerHA SystemMirror for IBM i Cookbook
8

Chapter 8. Planning for PowerHA


Good planning is essential for the successful setup and use of your server and storage
subsystems. It ensures that you have met all of the prerequisites for your server and storage
subsystems and everything that you need to gain advantage from best practices for
functionality, redundancy, performance, and availability.

In this chapter, we discuss important planning considerations and all the prerequisites that
you need to use IBM PowerHA SystemMirror for i in your environment.

© Copyright IBM Corp. 2012. All rights reserved. 131


8.1 Requirements for PowerHA
In this section we discuss the licensing information for IBM PowerHA SystemMirror for i
and PRPQ 5799-HAS and the requirements for Power System, Virtual I/O Server, and
Storage System.

8.1.1 Licensing considerations


Before installing the IBM PowerHA SystemMirror for i license product (5770-HAS), check
whether the following is in place:
 IBM i 7.1 is installed on all system nodes (servers or logical partitions) that will be part of
your high-availability or disaster-recovery solution.
 HA Switchable Resources (Option 41 of 5770-SS1) are installed on all system nodes
(server or logical partitions) that will be part of your high-availability or disaster-recovery
solution. As HA Switchable Resources is included in 5770-HAS PowerHA SystemMirror
for i 7.1, you do not have to order it separately, but it still has to be installed separately.

The licensed program 5770-HAS must be licensed for all processor cores in the partitions
(nodes) that you want to be part of your cluster.

There are two editions of PowerHA System Mirror for i, the Standard edition and the
Enterprise edition (Table 8-1). The standard edition, which is PID 5770-HAS, is targeted at a
data center HA solution. The Enterprise edition, which is a feature of 5770-HAS, adds support
for a multi-site HA and DR solution. There is no charge to upgrade from PowerHA 6.1 to 7.1
Standard or Enterprise Edition.

Table 8-1 PowerHA SystemMirror for i Editions


IBM i HA/DR clustering Standard edition Enterprise edition

Centralized cluster management  

Cluster resource management  

Centralized cluster configuration  

Automated cluster validation  

Cluster admin domain  

Cluster device domain  

Integrated heartbeat  

Application monitoring  

IBM i event/error management  

Automated planned fail over  

Managed unplanned fail over  

Centralized Flash Copy  

LUN-level switching  

Geomirror Sync mode  

Geomirror Async mode 

132 PowerHA SystemMirror for IBM i Cookbook


IBM i HA/DR clustering Standard edition Enterprise edition

Multi-Site HA/DR management 

DS8000/SVC/V7000 Metro Mirror 

DS8000/SVC/V7000 Global Mirror 

The pricing of PowerHA SystemMirror 7.1 for i is based on a per-core basis and is broken
down among small-tier, medium-tier, and large-tier Power server for each edition.

You can order a Power System Capacity BackUp (CBU) model for your disaster recovery
and high availability when ordering a Power System or doing a model upgrade. The terms for
the CBU for i allow a primary system processor's optional i license entitlements and 5250
Enterprise Enablement license entitlements (or optional user entitlements in the case of the
Power 520 and 720 CBU) to be temporarily transferred to the CBU Edition when the primary
system processor cores are not in concurrent use. For an IBM i environment, you must
register the order for a new CBU at IBM Capacity Backup for Power Systems site:
http://www-03.ibm.com/systems/power/hardware/cbu/

A minimum of one IBM i processor entitlement or enterprise enablement is required on each


primary machine and CBU machine at all times, including during the temporary move period.
For machines that have IBM i priced per user, the minimum quantity of IBM i user
entitlements for each machine is required on each primary machine and CBU machine at all
times, including during the temporary move period.

The same applies to PowerHA SystemMirror entitlements. A minimum of one entitlement on


the primary machine and one on the CBU machine is required at all times, including during
the temporary move period.

Chapter 8. Planning for PowerHA 133


In Figure 8-1, there are eight active cores with IBM i and PowerHA SystemMirror entitlements
on the primary server. On the CBU box, we have one IBM i entitlement, one PowerHA
entitlement, and a temporary key (good for two years) with a usage limit of eight cores for
PowerHA SystemMirror.

Production CBU
Partition #1
Unused

Partition #2

Active standby
CBU Cores

Partition #3

Active standby
CBU Cores

IBM i, 5250, PowerHA

4+ 4 =8 1 Total = 9 PowerHA licenses

Figure 8-1 PowerHA licensing for CBU

8.1.2 PRPQ ordering information


You need to order 5799-HAS Program Request Pricing Quotation (PRPQ) for IBM PowerHA
SystemMirror for i to obtain these enhancements in PowerHA:
 PowerHA GUI.
 SVC/V7000 remote Copy Service support Metro Mirror, Global Mirror, and FlashCopy.
 New CL commands:
– CFGDEVASP to create an IASP
– CFGGEOMIR to configure geographic mirroring

This PRPQ (5799-HAS) is available in the English language (2924) only but can be installed
on other language systems. You must install the PRPQ with the LNG parameter as 2924
when you install PRPQ with RSTLICPGM. PTF SI44148 for LPP 5770-HAS is a prerequisite for
the PRPQ.

134 PowerHA SystemMirror for IBM i Cookbook


The order for the product should be placed via the ordering system particular to a location (for
example, AAS or WTAAS). Specify product ID 5799-HAS. See the following list for additional
information by geography:
 United States
PRPQs in the United States are ordered by the CSO via US ordering processes.
If you have any problems ordering the PRPQ, contact the Special Product
Marketing representative.
 EMEA
For EMEA ordering information, see SWORDERINFO (EPLC) on HONE. For price
information see the HONE SW Price application. For further assistance with EMEA
ordering information contact [email protected].
 Asia Pacific
In Asia Pacific, submit your request via Access GFS at:
http://cqrfa.vienna.at.ibm.com:8080/agfs/access/indexservlet
In Japan assistance is available from Process Fulfillment at [email protected].
 Canada
Assistance is available from Canada RPQ/Austria/IBM.
 Latin America
To order PRPQs in Latin America, contact the responsible fulfillment account
representative who normally orders your equipment via AAS. If you have any problems
ordering the PRPQ, contact your sales manual team in Vienna.

8.1.3 Power Systems requirements


IBM PowerHA SystemMirror for i requires IBM i 7.1. If the hardware is not supported by
IBM i 7.1, IBM PowerHA SystemMirror for i is not supported. The following platforms
supported by IBM i 7.1 are able to use IBM PowerHA SystemMirror for i:
 Power Systems servers and blades with POWER7® processors
 Power Systems servers and blades with POWER6/6+ processors
 System i servers with POWER6 processors
 System i servers with POWER5/5+ processors

When using external storage-based remote Copy Services together with IBM PowerHA
SystemMirror for i, make sure to use separate Fibre Channel adapters for SYSBAS and for
each IASP group. When using NPIV, you do not need separate physical adapters, but you
still must use separate virtual adapters for SYSBAS and each IASP group.

Chapter 8. Planning for PowerHA 135


For using the independent ASP, consider the disk arms of System ASP for good application
performance. This is particularly important when it comes to temporary storage in a system
configured with independent disk pools. All temporary storage is written to the SYSBAS disk
pool. You must also remember that the operating system and basic functions occur in the
SYSBAS disk pool. As a starting point use the guidelines provided in Table 8-2.

Table 8-2 Disk arms for SYSBAS guidelines


Disk arms in iASPs Arms for SYSBAS: Divide iASP arms by:

Less than 20 3

20 - 40 4

Greater than 40 5

Note: Every situation is going to be different. The rule of thumb above is presented as a
starting point. Your own requirements might vary.

Also, install the latest PTFs related with IBM PowerHA SystemMirror for i to all nodes in the
cluster. The latest known PTFs fix known problems and provide the enhanced benefits in an
IBM PowerHA environment. Periodically check and install the latest known PTF in
“Recommended Fixes for High Availability for Release 7.1" at the following IBM i Support
Recommended fixes website:
http://www-912.ibm.com/s_dir/slkbase.nsf/recommendedfixes

IBM PowerVM
IBM PowerVM is a virtualization technology for AIX®, IBM i, and Linux environments on IBM
POWER processor-based systems. The PowerVM Virtual I/O Server included in the
PowerVM provides a virtualization environment, in that the storage and network I/O
resources are shared. This feature is required for IBM i to attach to DS8000 via VSCSI or
NPIV and to SVC/V7000.

Virtualization technology is offered in three editions on Power Systems:


 PowerVM Express Edition
 PowerVM Standard Edition
 PowerVM Enterprise Edition

They provide logical partitioning (LPAR) technology by using either the Hardware
Management Console (HMC) or the Integrated Virtualization Manager (IVM), Dynamic LPAR
operations, Micro-Partitioning®, and Virtual I/O Server capabilities, and N_Port ID
Virtualization (NPIV). You can choose the editions of IBM PowerVM depending on what
features you want (Table 8-3).

Table 8-3 IBM PowerVM Editions


Features Express Standard Enterprise

Maximum VMs 3/Server 1000/Server 1000/Server

Management VMControl, IVM VMControl, IVM, HMC VMControl, IVM, HMC

Virtual I/O Server  (Dual) (Dual)

Suspend/resume  

NPIV   

136 PowerHA SystemMirror for IBM i Cookbook


Features Express Standard Enterprise

Shared Processor Pools  

Shared Storage Pools  

Thin Provisioning  

Active Memory™ 
Sharing

Live Partition Mobility 

8.1.4 Virtual I/O Server considerations


Virtual I/O Server (VIOS) is virtualization software that runs in a separate partition of your
Power System. Its purpose is to provide virtual storage and networking resources to one or
more client partitions. The IBM i client partition can also be hosted by VIOS.

The Virtual I/O Server owns the physical I/O resources, such as Ethernet and SCSI
(SAS)/FC adapters. It virtualizes those resources for its client LPARs to share them remotely
using the built-in hypervisor services. These client LPARs can be quickly created, typically
owning only physical memory and shares of processors without any physical disks or physical
Ethernet adapters.

Traditionally, IBM i has been supporting 520 bytes per sector storage. There are restrictions
to directly attach a common storage system to IBM i for this reason. However, VIOS brings to
he IBM i client partition the ability to use 512 bytes per storage sector by using the sector
conversion of IBM PowerVM Hypervisor, which converts IBM i traditional 8 x 520 byte sectors
into 9 x 512 byte sectors of a 4 KB memory page. Therefore, open storage volumes (or logical
units, LUNs) are physically attached to VIOS through a FC or a Serial-attached SCSI (SAS)
connection and then made available to IBM i (Figure 8-2).

Virtual I/O Server IBM i Client Partition Virtual I/O Server


FC Adapter

FC Adapter
Device Driver IBM i Multipathing Device Driver
Multi-pathing Multi-pathing

SCSI
LUNs
Hdisk #1-n Hdisk
#1-n #1-n
FC Adapter
FC Adapter

VSCSI VSCSI VSCSI VSCSI


Server Client Client Server
Adapter ID10 Adapter ID10 Adapter ID20 Adapter ID20

POWER Hypervisor

IBM System Storage


or SVC

Figure 8-2 VSCSI connection via VIOS with multipathing

Chapter 8. Planning for PowerHA 137


When you configure the IBM i client partition with VIOS, consider a dependency on VIOS. If
the VIOS partition fails, IBM i on the client will lose contact with the virtualized open storage
LUNs. The LUNs also become unavailable if VIOS is brought down for scheduled
maintenance or a release upgrade. To remove this dependency, you can configure
Redundant VIOS so that two or more VIOS partitions can be used to simultaneously provide
virtual storage to one or more IBM i client partitions using IBM i multipathing (Figure 8-2 on
page 137).

Virtual SCSI
From IBM i 6.1.1, IBM i VSCSI client driver supports MPIO through two or more VIOS
partitions to a single set of LUNs. This multipath configuration allows a VIOS partition to fail or
be brought down for service without IBM i loosing access to the disk volumes as the other
VIOS partitions remain active (Figure 8-2 on page 137).

When setting up external storage with VIOS and VSCSI, there are several VIOS settings that
should be changed. The first is the fc_err_recov and dyntrk attributes for each fscsiX device
in VIOS partitions. To show the current fscsiX devices, use lspath.

Run the following command to set the fast fail and dynamic tracking attributes on
these devices:
chdev–attr fc_err_recov=fast_fail,dyntrk=yes –perm –dev fscsiX

R to restart VIOS after these changes. The fast fail is designed for a multipath environment
not to retry failed paths for a long time. In single path configuration, do not set fast_fail.

Note: In a dual VIOS environment using IBM i multipath, the SCSI reserve policy for each
LUN (or hdisk) on both VIOS LPARs must be set to no_reserve to allow disk sharing.

This change must be made prior to mapping the LUNs to IBM i, and it does not require a
restart of VIOS.

To show the current reserve policy settings, run the following command:
lsdev –dev hdiskX –attr reserve_policy

You can set the no_reserve attribute by running the following command:
chdev –dev hdiskX –attr reserve_policy=no_reserve

You need to install the SVC Subsystem Device Driver Path Control Module (SDDPCM) on
VIOS to enable the management of multiple paths to the SAN Volume Controller or V7000
virtual Disks (VDisks). For more information about the Multipath Subsystem Device Driver,
see the latest user's guides, available here:
https://www-304.ibm.com/support/docview.wss?dc=DA400&rs=540&uid=ssg1S7000303&conte
xt=ST52G7&cs=utf-8&lang=en&loc=en_US

N_Port ID Virtualization (NPIV) with IBM System Storage DS8000


N_Port ID Virtualization (NPIV) is an industry-standard FC protocol that allows VIOS to
directly share a single FC adapter among multiple client LPARs, acting as a FC passthrough.
Unlike VSCSI, NPIV does not map a LUN to a virtual target device in VIOS, which the client
LPAR can then access as a generic SCSI disk. Instead, a port on the physical FC adapter is
mapped to a virtual FC server adapter in VIOS, which in turn is connected to a virtual FC
client adapter in IBM i (Figure 8-3 on page 139). When the virtual FC client adapter is
created, two unique world-wide port names (WWPNs) are generated for it. Through the link to
the server virtual FC adapter and then the physical adapter in VIOS, those WWPNs become

138 PowerHA SystemMirror for IBM i Cookbook


available on the SAN, and storage can be mapped to them as with any other FC host ports.
Note that these WWPNs are unique, not just within the Power server, but globally on the
SAN. When a virtual FC client adapter is deleted, the WWPNs are not reused. By default, the
PowerVM Hypervisor is capable of creating 32,000 WWPNs for virtual FC client adapters. If
additional WWPNs are required, clients can acquire an enablement code from IBM.

These are the requirements for IBM i DS8000 NPIV attachment:


 8 Gb IOP-less Fibre Channel IOA (CCIN 577D) or 10 Gb FCoE PCI Express Dual Port
Adapter (CCIN 2B3B)
 NPIV-capable SAN switch
 IBM i 6.1.1 or later
 HMC V7R3.5.0 or later
 POWER6 FW 350_038 or later
 VIOS 2.1 Fix Pack 22.1 or later

Virtual I/O Server IBM i Client Partition

Physical Adapter WWPN


DS8000 2107
eeeeffff55556666
LUNs
#1-n
FC Adapter
8 G #5735

WWPNs
aaaabbbb11112222
ccccdddd33334444

VFC VFC
Server Client
Adapter ID10 Adapter ID10
IBM i LUNs Mapped to
WWPN aaaabbbb11112222
POWER Hypervisor

Figure 8-3 NPIV for IBM i

For more information about connecting IBM i through VIOS to DS8000 using VSCSI and
NPIV, see DS8000 Copy Services for IBM i with VIOS, REDP-4584.

8.1.5 Storage considerations


You can check which of the external storages can attach directly to IBM i or serve for IBM i
partitions via VIOS in 4.1, “PowerHA technologies” on page 46.

IBM Storage System DS6000/DS8000


For the IBM PowerHA SystemMirror for i to be able to communicate with the external Storage
System you need to install the DS command-level interface (DS CLI) on all nodes in the
cluster. The DS CLI software can be found and downloaded here:
http://www-947.ibm.com/support/entry/portal/Downloads/Hardware/System_Storage/Storage_software
/Other_software_products/Copy_Services_CLI_%28Command_Line_Interface%29

For more information about using DS CLI with IBM i, see Chapter 8, "Using DS CLI with
System i," in IBM i and IBM System Storage: A Guide to Implementing External Disks on
IBM i, SF24-7120.

Chapter 8. Planning for PowerHA 139


The support of DS8000 multiple Global Mirror sessions have been added in Version 6.1 of the
DS8000 firmware. Prior to 6.1, only a single Global Mirror session was allowed to be active at
a time on the DS8000. However, the Global Mirror session on the DS8000 is different from
the PowerHA ASP session of type *GLOBALMIR. PowerHA does not support multiple Global
Mirror sessions, as there are additional parameters that are required on the DSCLI
commands. However, the legacy commands for Global Mirror from PowerHA work as long as
the user only has a single Global Mirror session on the DS8000.

Fibre Channel-attached LUNs are identified as the storage unit device type of 2107 on the
IBM i host system. You can specify 1 - 32 LUNs for each attachment to the IBM i Fibre
Channel adapter feature 2766, 2787, or 5760. Also, you can specify 1 - 64 LUNs for each
attachment to the IBM i Fibre Channel adapter feature 5749, 5774/5276, or 5735/5273.

Table 8-4 Capacity and Models of Disk Volumes for IBM i


Size Type Protected model Unprotected model

8.5 GB 2107 A01 A81

17.5 GB 2107 A02 A82

35.1 GB 2107 A05 A85

70.5 GB 2107 A04 A84

141.1 GB 2107 A06 A86

282.2 GB 2107 A07 A87

SVC/V7000
Attaching SVC/V7000 to IBM i is possible using IBM PowerVM Virtual I/O Server (VIOS).
IBM i runs in its own LPAR as a client of the VIOS server and accesses the SVC/V7000
storage via Virtual SCSI (vSCSI).

For this attachment, the minimum requirements for IBM PowerHA SystemMirror for i are
SVC/V7000 Version 6.1.x and IBM VIOS 2.2.1. For a SVC-supported hardware list, device
driver, firmware, and recommended software level, see the following web page:
https://www-304.ibm.com/support/docview.wss?uid=ssg1S1003697#_CodeLevel

To communicate between PowerHA and SVC/V7000 for issuing command-line interface


(CLI), you must prepare for using the Secure Shell (SSH) TCP port 22. You can generate the
SSH key pair from IBM i. For more information, see “Preparing for SSH connection between
IBM i and SVC/V7000” on page 373.

Table 8-5 lists the configuration maximums for VIOS supporting an IBM i client attached to SVC.

Table 8-5 SVC configuration maximums for IBM i servers


Object Maximum Descriptions

Volume (HDisk) 512 The maximum number of volumes that can be supported
by the SAN Volume Controller for a host running an IBM i
operating system (per host object).

Paths per volume 8 The maximum number of paths to each volume. The
suggested number of paths is 4.a
a. Subsystem device driver path-control module (SDDPCM) for AIX supports 16 paths per
volume, but the SAN Volume Controller supports a maximum of only eight paths for a
reasonable path-failover time.

140 PowerHA SystemMirror for IBM i Cookbook


There are also known issues and limitations with the SVC and IBM i host, so consider the
following items when attaching SVC to a host that runs IBM i:
 A maximum of 16 disk virtual LUNs and 16 optical virtual LUNs is supported for each IBM i
virtual I/O client SCSI adapter.
 SAN Volume Controller thin-provisioned volumes are supported for IBM i for use as
FlashCopy targets only.

8.2 PowerHA Copy Services Support considerations


In this section we describe the considerations for using Copy Services support in
various scenarios.

8.2.1 Global Mirror symmetrical and asymmetrical configurations


Global Mirror and IASPs offer a new and exciting opportunity for a highly available
environment. They enables customers to replicate their environment over an extremely long
distance without the use of traditional IBM i replication software. This environment comes in
two types:
 Asymmetrical
 symmetrical

Chapter 8. Planning for PowerHA 141


Asymmetrical configuration
With this configuration (Figure 8-4), Global Mirror can only be used from the production site to
the disaster/recovery site. This type of configuration would be typical for a disaster recovery
configuration where the production systems would run in the secondary location only if there
was an unplanned outage of the primary location. Only one consistency group is set up, and
it resides at the remote site. This means that you cannot do regular role swaps by reversing
the Global Mirror direction (disaster recovery to production).

IASP-A

Prod DS

Prod
System/LPAR

*SYSBAS
Global
Mirror

SAN
IASP-A IBM i Cluster & Device Domain
Global Copy
FlashCopy

D/R
System/LPAR

DR DS *SYSBAS

IASP-A
Flash Copy w/CG

Figure 8-4 Global Mirror with asymmetrical configuration

After production workloads are moved to the recovery site, Global Copy must be used to
return to the primary site. This is a manual procedure that is not supported by IBM PowerHA
SystemMirror for i. For more detailed steps for Global Mirror switchback with an asymmetrical
configuration, see in “Using CL commands for an asymmetrical Global Mirror switchback after
a planned outage” on page 339. Because no disaster recovery capability would be provided
in the reverse direction, it is unlikely that in this type of configuration we would choose to run
for extended periods of time in the secondary location unless forced to by unavailability of the
primary site.

Because Global Mirror uses two copies of data in the secondary location, there would be
twice as many physical drives in this location as in the production location if the same size
drives were used. In some situations, it might be cost effective to use larger drives in the
secondary location. Spreading the production data over all these drives should provide
equivalent performance in a disaster situation while reducing the overall cost of the solution.

142 PowerHA SystemMirror for IBM i Cookbook


Symmetrical configuration
With the configuration shown in Figure 8-5, an additional FlashCopy consistency group is
created on the source site production DS model. It provides all the capabilities of
asymmetrical replication, but adds the ability to do regular role swaps between the production
and the disaster recovery sites using PowerHA.

IASP-A

Prod DS

Prod
FlashCopy

System/LPAR

*SYSBAS

IASP-A
Flash Copy w/CG
Global

SAN
Mirror

IBM i Cluster & Device Domain


IASP-A
Global Copy

D/R
FlashCopy

System/LPAR

*SYSBAS
DR DS
IASP-A
Flash Copy w/CG

Figure 8-5 Global Mirror with symmetrical configuration

As we have FlashCopy capacity in both sites, it is possible to provide a disaster recovery


solution using Global Mirror in both directions between the two sites. This type of
configuration would typically be used where the production workloads might run for extended
periods of time in either location.

8.2.2 FlashCopy NoCopy/full copy/incremental/reverse


In this section we discuss the various FlashCopy options.

Full copy versus NoCopy


There are two variants to the copy operation in FlashCopy. Whether you do a full copy or
NoCopy depends on how you want to use the FlashCopy targets. In a real-world environment
where you might want to use the target volume over a longer time and with high I/O workload,
full copy is a better option to isolate your backup system I/O workload from your production
workload when all data has been copied to the target. For the purpose of using FlashCopy for

Chapter 8. Planning for PowerHA 143


creating a temporary system image for saving to tape during a low production workload, the
nocopy option is recommended.

With the standard FlashCopy, full copy is the default, while the nocopy option is the default
for FlashCopy SE.

Incremental and reverse FlashCopy


Incremental FlashCopy provides the capability to refresh a FlashCopy relationship, thus
refreshing the target volume. When refreshing a target volume, any writes previously written
to the target volume are always overwritten. The incremental FlashCopy is not available with
FlashCopy SE.

To perform an incremental FlashCopy, you must first establish the FlashCopy relationship
with the change data recording and persistent FlashCopy options enabled. You can do the
incremental copy at any time, and you do not have to wait for the previous background copy
to complete.

Usually you can use the incremental FlashCopy to minimize the amount of data that must be
regularly copied and save the time for the physical copy of FlashCopy.

With reverse FlashCopy, the FlashCopy relationship can be reversed by copying over
modified tracks from the target volume to the source volume. For DS6000 and DS8000, the
background copy process must complete before you can reverse the order of the FlashCopy
relationship to its original source and target relationship. The change data recording is a
prerequisite for reverse restore.

The reverse restore function can only be used when a full copy relationship is completed.
Therefore, it is not possible with FlashCopy SE. PowerHA supports reverse FlashCopy only
to volumes that are not a part of a Metro Mirror or Global Mirror relationship.

It is possible to establish up to 12 FlashCopy relationships using the same source for


DS8000. That is, a source volume can have up to 12 target volumes. However, a target
volume can still only have one source. Also, a FlashCopy target volume cannot be a
FlashCopy source volume at the same time.

For multi-relationship FlashCopy, an incremental FlashCopy relationship can be established


with one and only one target. For each source volume, only one FlashCopy relationship can
be reversed. With reversing one relationship of a DS8000 multi-relationship FlashCopy, all
FlashCopy targets for which the background copy process has not completed are lost.

A persistent relationship is necessary to do an incremental FlashCopy or a reverse FlashCopy.

8.3 Sizing and performance considerations


In this section we discuss sizing and performance considerations on various environments.

8.3.1 Geographic mirroring


With geographic mirroring, IBM i does the replication. It is important to consider performance
when planning to implement a geographic mirroring solution. While asynchronous geographic
mirroring does allow a bit more flexibility regarding the distance between systems, there are
still implications to undersizing the source, target, or the communications line between the two.

144 PowerHA SystemMirror for IBM i Cookbook


Minimizing the latency (that is, the time that the production system waits for the
acknowledgement that the information has been received on the target system) is key to
good application performance.

In the following sections we discuss how to size for good geographic mirroring performance.

General performance recommendations


When implementing geographic mirroring, different factors can influence the performance of
systems involved in this HA solution. To maximize the performance of your applications that
are used in this HA solution, several planning considerations must be taken into account. The
factors discussed in this section provide general planning considerations for maximizing
performance in a geographic mirroring environment.

There are two separate aspects to consider when sizing for a geographic mirroring
environment. During the normal run time of the production environment, there will be some
overhead added by geographic mirroring as the IBM i operating system is sending disk writes
to the target system. The second aspect is the overhead and time required for
synchronization, when the target IASP is reconnected to the source IASP and changes are
pushed from the source to the target to make the two equivalent again.

Source and target comparison


Geographic mirroring consumes resources on both the source and the target resource.
Especially for synchronous geographic mirroring, the best performance will be seen when the
source and target systems are fairly equivalent in CPU, memory, and the disk subsystem.

CPU considerations
There is extra overhead on CPU and memory when doing geographic mirroring, and this has
to be considered for both the source and the target system. Geographic mirroring increases
the CPU load to the system processors on both the system owning the production copy of the
IASP and the system owning the mirror copy of the IASP. There must be sufficient excess
CPU capacity to handle this overhead, but there is no formula to calculate this exactly as it
depends on many factors in the environment and the configuration. This CPU usage is
needed for both systems to communicate and replicate data from the source IASP to the
target IASP.

You might require additional processors to increase CPU capacity. As a general rule, the
partitions that you are using to run geographic mirroring needs more than a partial processor.
In a minimal CPU configuration, you can potentially see 5 - 20% CPU overhead while running
geographic mirroring.

Regarding the backup system, be especially careful in sizing that system's processor. It
should not be a small percentage of your production system, because this might slow down
synchronization times considerably. If your backup system has fewer processors in
comparison to your production system and there are many write operations, CPU overhead
might be noticeable and affect performance.

Memory considerations
Geographic mirroring also requires extra memory in the machine pool. For optimal
performance of geographic mirroring, particularly during synchronization, increase your
machine pool size by at least the amount given by the following formula and then use
WRKSHRPOOL to set the machine pool size:
Extra machine pool size = 300 MB + (0.3 * number of disk arms in the IASP)

This extra memory is needed particularly during the synchronization process on the system
that owns the mirror copy of the IASP. However, you must add extra storage on every cluster

Chapter 8. Planning for PowerHA 145


node involved in geographic mirroring (as defined in the cluster resource group). Any node in
the cluster can become the primary owner of the mirror copy of the IASP if a switchover or
failover occurs.

Important: The machine pool storage size must be large enough before starting the
resynchronization. Otherwise, increasing memory is not taken into account as soon as the
synchronization task is in progress, and the synchronization process can take longer.

If the system value QPFRADJ is equal to 2 or 3, then the system might make changes to the
storage pools automatically as needed. To prevent the performance adjuster function from
reducing the machine pool size take these steps:
1. Set the machine pool minimum size to the calculated amount (the current size plus
the extra size for geographic mirroring from the formula) by using the Work with
Shared Storage Pools (WRKSHRPOOL) command or the Change Shared Storage Pool
(CHGSHRPOOL) command.
2. Set the Automatically adjust memory pools and activity levels (QPFRADJ) system value to
zero, which prohibits the performance adjuster from changing the size of the machine pool.

Note: We recommend that you use WRKSHRPOOL for setting the machine pool size to the
calculated minimum. Disabling Performance Auto Adjuster can have other performance
implications in your environment.

Disk subsystem considerations


Disk unit and IOA performance can affect overall geographic mirroring performance. This is
especially true when the disk subsystem is slower on the mirrored system. When geographic
mirroring is in synchronous mode, all write operations on the production copy are gated by
the mirrored copy writes to disk. Therefore, a slow target disk subsystem can affect the
source-side performance. You can minimize this effect on performance by running
geographic mirroring in asynchronous mode. Running in asynchronous mode alleviates the
wait for the disk subsystem on the target side and sends confirmation back to the source side
when the changed memory page is in memory on the target side.

System disk pool considerations


Similar to any system disk configuration, the number of disk units available to the application
can have a significant affect on its performance. Putting additional workload on a limited
number of disk units might result in longer disk waits and ultimately longer response times to
the application. This is particularly important when it comes to temporary storage in a system
configured with independent disk pools. All temporary storage (such as objects in the QTEMP
library) is written to the SYSBAS disk pool. If your application does not use a lot of temporary
storage, then you can get by with fewer disk arms in the SYSBAS disk pool.

146 PowerHA SystemMirror for IBM i Cookbook


Although disk writes associated with production data will be on the IASP, there will continue
to be disk activity associated with the SYSBAS pool for the operating system and basic
functions. As a starting point, you can use the guidelines shown in Table 8-6.

Table 8-6 Disk arms for SYSBASE guidelines


Disk arms in IASPs Arms for SYSBAS: Divide IASP arms by:

Less than 20 3

20 - 40 4

Greater than 40 5

For example, if IASP contains 10 drives, then SYSBAS should have at least three. As another
example, if IASP contains 50 drives, then SYSBAS should have at least 10.

Note: Disk pool sizing is very application dependent, and the above guidelines are only
given as a starting point. You might find that fewer or more arms are required for your
application environment. Understanding performance monitoring for your application
environment is critical for sizing and capacity planning.

You will want to monitor the percent busy of the SYSBAS disk arms in your environment to
ensure that you have the appropriate number of arms. If it gets above 40% utilization, then
you must add more arms. Also, when possible, the disk assigned to the IASP should be
placed on a separate I/O adapter from the SYSBAS disk to reduce any potential contention. It
has also been found that IOA cache is very important and provides greater data integrity and
improved performance.

Communications lines
When you are implementing a PowerHA solution using geographic mirroring, plan for
adequate communication bandwidth so that the communications bandwidth does not become
a performance bottleneck in addition to system resources.

Geographic mirroring can be used for virtually any distance. However, only you can determine
the latency that is acceptable for your application. The type of networking equipment, the
quality of service, the distance between nodes, the number, and the characteristics of data
ports used can all affect the communications latency. As a result, these become additional
factors that can impact geographic mirroring performance.

To ensure better performance and availability, we recommend that you take the
following actions:
 To provide consistent response time, geographic mirroring should have its own redundant
communications lines. Without dedicated communication lines, there might be contention
with other services or applications that utilize the same communication line. Geographic
mirroring supports up to four communication lines (data port lines), and a cluster heartbeat
can be configured for up to two lines. However, we recommend utilizing Virtual IP
addresses to provide redundancy to the cluster heartbeat.
 It is important to know that a round-robin approach is used to send the data across the
lines. This implies that for best performance, when multiple dataport lines are configured,
they should have close to equivalent performance characteristics. If one slow line is
added, then this will gate the sending of the data to that line speed.
 Geographic mirroring replication should also be run on a separate line from the cluster
heartbeating line (the line associated with each node in the cluster). If the same line is
used, during periods of heavy geographic mirroring traffic, heartbeating could fail, causing

Chapter 8. Planning for PowerHA 147


a false partition. For the cluster heartbeat we recommend using two different IP addresses
for redundancy.
 From a high availability point of view, we recommend using different interfaces and routers
connected to different network subnets for the four data ports that can be defined for
geographic mirroring (Figure 8-6). It is better to install the Ethernet adapters in different
expansion towers, using different System i hardware buses. Also, if you use multiport IOA
adapters, use different ports to connect the routers.
 Virtual IP adapter (VIPA) can be used to define the geographic mirroring IP addresses.

Node 1 Node 2
PRODUCTION Copy MIRROR Copy

LAN 1

LAN 2
SYSBAS SYSBAS

Geographic
Mirroring
using multiple
Ethernet adapters
IASP and LANs IASP
source target

Site A Site B

Figure 8-6 Recommended network configuration for geographic mirroring

Runtime environment
In this section we discuss the runtime environment.

Delivery and mode


When configuring geographic mirroring, there are two main parameters that affect geographic
mirroring runtime performance. The DELIVERY parameter affects the performance of disk
writes to the IASP. With synchronous delivery, the disk write will not complete until the
affected page in storage has also been received on the target system. Asynchronous delivery
allows the disk write on the source to complete after the write has been cached. The actual
sending of the disk write to the target system happens outside the scope of the write on the
source. For synchronous delivery, there is also a synchronous or asynchronous MODE.
Synchronous mode ensures that the write has arrived at the disk cache on the target
(essentially on disk at that point) before returning. Asynchronous mode only ensures that the
write is on memory on the target.

Synchronous delivery and synchronous mode guarantee equivalent copies of the IASP on
the source and the target while geographic mirroring is active. It also provides the added
protection of a crash-consistent copy of the data in case of a target system failure, because all
writes will have been received into the disk subsystem.

Synchronous delivery and asynchronous mode can be beneficial for customers running with
a significantly slower disk subsystem on their target system. This allows the disk write on the

148 PowerHA SystemMirror for IBM i Cookbook


source to complete without waiting for the completion on the target. This delivery and mode
still guarantee equivalent data on the source and target IASPs in the case of a failure of the
source system.

With synchronous delivery, it is important to have the communications bandwidth available to


support the number of disk writes at all peak periods throughout the day or night. The
overhead of sending the data to the target will be added to the time for each disk write to
complete, which can significantly affect production performance. Even with a very fast line, if
the distance between the source and the target is too great, production performance will
suffer. For this reason, asynchronous delivery for geographic mirroring was introduced in
release 7.1.

Asynchronous delivery is best for those environments in which the source and target are
separated by too long of a distance for acceptable synchronous response times, or for
scenarios where the bandwidth cannot support the peak write rate.

Sizing for optimum performance


For the best runtime performance, it is important to know the write volume within the IASP.
We only consider writes because those are the only I/O that is transferred to the target
system. If the IASP has not yet been defined, the write volume in SYSBAS can be used as a
rough estimate, understanding that this might result in excess communications capacity. Both
the peak and average megabytes per second written should be collected, preferably over
short intervals, such as 5 minutes.

For synchronous delivery, the bandwidth of the communications lines must be able to keep
up with the peak write volume. If it cannot keep up, the writes will begin to stack up and
production performance will suffer.

For asynchronous delivery, the bandwidth of the lines must still keep up at least to the
average write volume. Because writes on the source are not waiting, it is acceptable for some
queuing to occur, but if the line cannot handle the average write volume, then geographic
mirroring will continue to get further and further behind.

It also is important to examine the variance of the write rate over time. If there is a large
variance between peak and average, then it might be advisable to size more for the peak.
Undersizing in this case affects the recovery point objective in the case of a source system
failure during the peak write rate.

Communications transports speeds


Just how fast is a T1 line? A data T1 transfers information at about 1.544 megabits every
second, or about 60 times more than the average conventional dialup modem. That translates to
.193 MBps theoretical throughput for a T1 line. The absolute best that you can hope to get out of
a T1 line is 70% effective throughput, and most network specialists say to plan for 30%.
Therefore, the best that a T1 line can transfer is .135 MBps. If you have a 2 Gigabyte file to
initially sync up, then that synch would take over 2 hours with nothing else running. As you scale
to 2 TB, that same sync would take over 80 days. As you can see, most systems need more
than a T1 line to achieve effective geographic mirroring transaction throughput.

T3 lines are a common aggregation of 28 T1 circuits that yield 44.736 Mbps total network
bandwidth or 5.5 MBps with a best effective throughput of 70%, which equals 3.9 MBps and a
planning number of 2 MBps.

The OC (the optical carrier fiber optic-based broadband network) speeds help you grow.

Chapter 8. Planning for PowerHA 149


Table 8-7 provides other communication line speeds.

Table 8-7 Communication line speeds


Type Raw speed Raw speed 30% planning GB/hour
(MBps) (MBps) (MBps) during synch

T1 1.544 0.193 0.06 0.22

DS3/T3 44.736 5.5 2 7.2

OC-1 51.840 6.5 2.1 7.6

OC-3 155.52 19.44 6 21.6

OC-9 455.56 56.94 18 64.8

OC-12 622.08 77.76 24 86.4

OC-18 933.12 116.64 35 126

OC-24 1244 155.5 47 169

OC-36 1866 233.25 70 252

OC-48 2488 311 93 335

OC-192 9953 1244.12 373 1342

1 Gb Ethernet 1000 125 38 (30% local) 225


local

Monitoring the runtime environment


When using asynchronous delivery, it might be useful to determine whether geographic
mirroring is “keeping up” with disk writes. On DSPASPSSN on the source system, the Total data
in transit field gives the amount of data in megabytes that has been sent to the target system,
but not acknowledged as received. This field is only shown when the transmission delivery is
*ASYNCH and the state is ACTIVE.

Synchronization
In this section we discuss synchronization.

Partial and full synchronizations


When you suspend mirroring for any planned activities or maintenance, any changes made
on the production copy of the independent disk pool are not being transmitted to the mirror
copy. So, when you resume geographic mirroring, synchronization is required between the
production and mirror copies.

If geographic mirroring is suspended without tracking, then full synchronization occurs. This
can be a lengthy process. If geographic mirroring is suspended with the tracking option,
PowerHA will track changes up to the tracking space limit specified on the ASP session.
When mirroring is resumed, the production and mirror copies are synchronized concurrently
with performing geographic mirroring.

Tracking is available on both the source side and the target side. Target side tracking greatly
reduces the need for a full synchronization. Usually a full synchronization is only required
when either the source or target IASP does not vary off normally, such as from a crash or an
abnormal vary-off.

150 PowerHA SystemMirror for IBM i Cookbook


While a synchronization is taking place, the environment is not highly available. This makes it
essential to calculate the time required to do a full synchronization to understand whether the
business can support that length of time exposed to an outage.

Tracking space
Tracking space is a reserved area within the IASP where the system tracks changed pages
while in suspended status, and the changes need to be synchronized when resuming
mirroring. Tracking space is needed only when the target copy of the IASP is suspended,
detached, or resuming. The changes themselves are not contained within the tracking space,
only a space-efficient indication of which pages require changes. The amount of tracking
space allocated can be defined by the user. The maximum is 1% of the total space within the
IASP. Using CHGASPSSN, a user can set the percentage of that 1%. For example, setting the
field to 10% means that the tracking space would be 10% of 1% or .1% of the total IASP size.
These parameters can be viewed using DSPASPSSN. The tracking space allocated is the
percentage of the maximum (it would show 10% in the above example) and the tracking
space used is the percentage of the available tracking space being used.

Note: If tracking space is exhausted (it reaches 100%), then no more changes can be
tracked, and when you resume geographic mirroring, a full synchronization is required.

Monitoring synchronization
To track how much data is left to be synchronized, DSPASPSSN can be used on the source
system. On the second screen, there are fields for Total data out of synch and Percent
complete. These fields will display the megabytes of data that need to be resynchronized and
how far the synchronization has progressed. Both of these fields are updated as the
synchronization runs. Each time the a synchronization starts or is resumed, these fields will
be reset. In the case of a resume, the percent complete will reset to 0, but you should also
see a reduced total data out of synch.

Calculating full synchronization time


It is fairly simple to determine the bandwidth needed for day-to-day operations and
initial synchronization. All write I/O will need to be sent from the primary system to the
secondary system.

To determine the megabytes of writes per second for each interval, run the performance tools
during a representative and peak period.

From the resulting QADMDSK file use these parameters:


 DSBLKW number of blocks written: A block is one sector on the disk unit. PD (11,0).
 INTSEC elapsed interval seconds: The number of seconds since the last sample interval.
PD (7,0).

Then you take these steps:


1. Calculate disk blocks written per second:
Disk blocks written per interval divided by the number of seconds in the
interval ((QAPMDISK.QAPMDISK.DSBLKW / QAPMDISK.QAPMDISK.INTSEC)
2. Convert disk blocks to bytes. Multiply by 520 to get the number of bytes.
3. Divide by a million to get megabytes per second.
4. Divide by 2 to get geographic mirror traffic because the disk writes are doubled if the
system is using mirrored disk.

Chapter 8. Planning for PowerHA 151


The formula to calculate the amount of traffic expressed as megabytes written per second is
as follows:
((QAPMDISK.QAPMDISK.DSBLKW / QAPMDISK.QAPMDISK.INTSEC) * 520) / 1000000 / 2

For example, if you determine that the amount of traffic is 5 MBps and you want to use
geographic mirroring for disaster recovery, then you need a pipe that can accommodate
5 MBps of data being transferred. If you are configuring two lines as data ports, then you
need 2.5 MBps per line.

From Table 8-7 on page 150, we can see the following facts:
 A DS3/T3 allows 5.6 MBps theoretical throughput with a 2 MBps with a best practice at
30% utilization.
 An OC-3 line allows 19.44 MBps theoretical throughput with 6 MBps with a best practice at
30% utilization.

You can initially start with two DS3 lines, but you might need to go to two OC-3 lines to
account and plan for growth.

To determine the time needed for initial synchronization, divide the total space utilized in the
IASP by the effective communications capability of the chosen communications lines. Speed
of the lines makes a big difference.

For example, if the IASP size is 900 GB and you are using 1 Gb Ethernet switches, then the
initial synchronization time will be less than an hour. However, if you are using two T3/DS3
lines, each having an effective throughput of 7.2 GB/hour, it would take around 63 hours to do
the initial synchronization. This was calculated by dividing the size of the IASP by the
effective GB/hour, that is, 900 GB divided by 14 GBps. A full resynchronization might also be
needed in the event of a disaster, so that must be factored into disaster recovery plans.

In most cases, the size of the data is used in the calculation, not the size of the IASP. An
exception to this is a *NWSSTG in an IASP. An *NWSSTG object is treated as one file, so the
size of the *NWSSTG is used instead of the amount of data within the *NWSSTG file.

To compute the initial synchronization time for *NWSSTG in an IASP, divide the size of the
network storage space of the IBM i hosted partition by the effective speed of the
communications mechanism.

For example, if the network storage space hosting IBM i was set up as 600 GB, it would take
42 hours to do the initial synchronization for a disaster recovery scenario using two DS3 lines.

To improve the synchronization time, a compression device can be used.

Synchronization priority
The synchronization priority setting (low, medium, or high) determines the amount of
resources allocated to synchronization. Lower settings will gate synchronization, which will
also allow more resources to be allocated to non-synchronized work.

Managing contention between run time and synchronization


Ideally, synchronization will run best when the system is quiet. However, most businesses
cannot support this amount of quiesced time. Thus, synchronization will most likely be
contending for system resources with normal production workload, in addition to the normal
geographic mirroring runtime workload.

152 PowerHA SystemMirror for IBM i Cookbook


For the least effect on production work, a synchronization priority of low can be selected.
However, this lengthens the amount of time required to complete the synchronization, also
lengthening the amount of time without a viable target.

For additional information about best practices for high-availability environments, see
Chapter 15, “Best practices” on page 397.

8.3.2 Virtual I/O Server


When you configure VIOS for IBM i client serving, dedicated VIOS processors will work better
than shared processors. Virtual SCSI devices generally have higher processor utilization
when compared with directly attached storage. The amount of processor required for Virtual
SCSI Server is based on the maximum I/O rates required of it.

The sizing methodology used is based on the observation that the processor time required to
perform I/O operating on the virtual SCSI server is fairly constant for a given I/O size.
Table 8-8 provides the approximate cycles per second for both physical disk and logical
volume operations on a 1.65 Ghz processor. For other frequencies, scaling by the ratio of the
frequencies is sufficiently accurate to produce a reasonable sizing.

Table 8-8 Approximate cycle per second on a 1.65Ghz logical partition


Disk type 4 KB 8 KB 32 KB 64 KB 128 KB

Physical disk 45,000 47,000 58,000 81,000 120,000

Logical volume 49,000 51,000 59,000 74,000 105,000

For example, consider a Virtual I/O Server that uses two client logical partitions on physical
disk-backed storage. The first client logical partition requires a maximum of 7,000 8-KB
operations per second. The second client logical partition requires a maximum of 10,000
8-KB operations per second. The number of 1.65 Ghz processors for this requirement is
approximately ((7,000 × 47,000 + 10,000 × 47,000) / 1,650,000,000) = 0.48 processors,
which rounds up to a single processor when using a dedicated processor logical partition.

As a sizing general rule, a single dedicated POWER6 processor for VIOS is good enough for
about 40,000 virtual SCSI I/O.

For the memory sizing of VIOS, we do not need to consider the additional requirement for
VSCSI as similar to the processor because there is no data caching required by VSCSI. With
large I/O configurations and very high data rates, a 1 GB memory allocation for the VIOS is
likely to be sufficient. For low I/O rate situations with a small number of attached disks,
512 MB will most likely suffice.

The Virtual I/O Server requires a minimum of 30 GB of disk space with the Virtual I/O Server
Version 2.2.0.11, Fix Pack 24, Service Pack 1, or later.

On the VIOS host the mapped logical drives are seen as hdisks and then assigned to an
IBM i partition. We recommend that you assign entire hdisks to IBM i. VIOS supports logical
volumes where you can subset an hdisk into multiple volumes to assign to a partition. This is
not recommended for IBM i partitions.

IBM i has always been architected to perform best with more disk arms. This does not change
with SAN disks. You need to create a good number of logical drives, not one large drive.

Chapter 8. Planning for PowerHA 153


We strongly recommend that only FC or SAS physical drives are used to create LUNs for
IBM i as a client of VIOS because of the performance and reliability requirements of IBM i
production workloads.

As the LUNs are virtualized by VIOS, they do not have to match IBM i integrated disk sizes.
The technical minimum for any disk unit in IBM i is 160 MB and the maximum is 2 TB, as
measured in VIOS. Actual LUN size is based on the capacity and performance requirements
of each IBM i virtual client partition and load source disk restrictions (17.5 GB minimum,
1.9 TB maximum). IBM i performs best with logical drives that are the same size.

When creating an open storage LUN configuration for IBM i as a client of VIOS, it is crucial to
plan for both capacity and performance. As LUNs are virtualized for IBM i by VIOS instead of
being directly connected, it might seem that the virtualization layer will necessarily add a
significant performance overhead. However, internal IBM performance tests clearly show that
the VIOS layer adds a negligible amount of overhead to each I/O operation. Instead, the tests
demonstrate that when IBM i uses open storage LUNs virtualized by VIOS, performance is
almost entirely determined by the physical and logical configuration of the storage subsystem.

It is possible to virtualize up to 16 LUNs to IBM i through a single VSCSI connection. Each


LUN typically uses multiple physical disk arms in the open storage subsystem. If more than
16 LUNs are required in an IBM i client partition, an additional pair of VSCSI server (VIOS)
and client (IBM i) adapters must be created.

8.3.3 Copy Service bandwidth


The SAN infrastructure between the local and remote storage system plays a critical role in
how much data and how fast remote Copy Services is able to replicate. If the SAN bandwidth
is too small to handle the traffic, then application write I/O response times will be longer.

Consider that the bandwidth can handle peak write workload requirements.

For IBM Storage System, collect the performance data to get the highest write rate and
calculate the needed bandwidth as follows:
1. Assume 10 bits per byte for network overhead.
2. If the compression of devices for remote links is known, you can apply it.
3. Assume a maximum of 80% utilization of the network.
4. Apply a 10% uplift factor to the result to account for peaks in the 5-minute intervals of
collecting data, and a 20 - 25% uplift factor for 15-minute intervals.

As an example, we show how to calculate the required bandwidth for a given write workload:
1. The highest reported write rate at an IBM i is 40 MBps.
2. Assume 10 bits per byte for network overhead:
40 MBps * 1.25 = 50 MBps
3. Assume a maximum of 80% utilization of the network:
50 MBps * 1.25 = 62.5 MBps
4. Apply a 10% uplift for 5-minute intervals:
62.5 MBps * 1.1 = app 69 MBps
5. The needed bandwidth is 69 MBps.

154 PowerHA SystemMirror for IBM i Cookbook


SVC Global Mirror is much more sensitive to a lack of bandwidth than DS8000. Also, it is
important to design and size a pipe that is able to handle normal and peak write I/O. You can
calculate the link bandwidth for SVC Global Mirror with the peak write load for all servers,
additional background copy rate, and the SVC intercluster heartbeat traffic size. The
background copy rate is usually 10 - 20% of the maximum peak load. The intercluster
heartbeat traffic size for the primary cluster and the secondary cluster can be found as a
table in the SVC information center. As a example for the bandwidth calculation, see the
following steps:
1. The peak write load for all servers is 10 MBps.
2. Add 10 - 20% for background copy rate:
10 MBps * 1.15 = 11.5 MBps
3. Add the SVC intercluster heartbeat traffic. If there are two nodes in cluster 1 and four
nodes in cluster 2, the size is 4 Mbps:
11.5 MBps + 0.5 MBps = 12 MBps
The needed bandwidth is 12 MBps.

A Recovery Point Object (RPO) estimation tool is available for IBM and IBM Business
Partners. This tool provides a method for estimating the RPO in a DS8000 Global Mirror
environment in relation to the bandwidth available and other environmental factors (see
Techdocs Document ID: PRS3246 for IBM and IBM Business Partners). For more
information, contact to IBM or IBM Business Partners.

8.3.4 FlashCopy space-efficient relation


With a FlashCopy space-efficient or thin-provisioned relation, disk space will only be
consumed for the target when a write to source needs to be hardened on disk or when a write
is directed to the target. For this reason, using FlashCopy SE requires less disk capacity than
using standard FlashCopy, which can help lower the amount of physical storage needed.

FlashCopy SE is designed for temporary copies, so FlashCopy SE is optimized for use cases
where a small percentage of the source volume is updated during the life of the relationship. If
much more than 20% of the source is expected to change, there might be a trade-off in terms
of performance versus space efficiency. Also, the copy duration should generally not last
longer than 24 hours unless the source and target volumes have little write activity.

DS8000 FlashCopy SE
The DS8000 FlashCopy SE repository is an object within an extent pool and provides the
physical disk capacity that is reserved for space-efficient target volumes. When provisioning a
repository, storage pool striping will automatically be used with a multi-rank extent pool to
balance the load across the available disks. FlashCopy SE is optimized to work with
repository extent pools consisting of four RAID arrays. In general, we recommend that the
repository extent pool contain between one and eight RAID arrays. Extent pools larger than
eight RAID arrays are not recommended for the repository. It is also important that adequate
disk resources are configured to avoid creating a performance bottleneck. It is advisable to
use the same disk rotational speed or faster (10 K RPM or 15 K RPM) for the target repository
as for the source volumes. We also recommend that the repository extent pool have as many
disk drives as the source volumes.

After the repository is defined in the extent pool it cannot be expanded, so planning is
important to ensure that it is configured to be large enough. If the repository becomes full, the
FlashCopy SE relationships will fail. After the relationship fails, the target becomes
unavailable for reads or writes, but the source volume continues to be available for reads and

Chapter 8. Planning for PowerHA 155


writes. You can estimate the physical space needed for a repository by using historical
performance data for the source volumes along with knowledge of the duration of the
FlashCopy SE relationship. In general, each write to a source volume consumes one track of
space on the repository (57 KB for CKD, 64 KB for FB). Thus, the following calculation can be
used to come up with a reasonable size estimate:
IO Rate x (% Writes/100) x ((100 – Rewrite%)/100) x Track Size x Duration in
seconds x ((100+Contingency%)) = Repository Capacity Estimate in KB

Because it is critical not to undersize the repository, a contingency factor of up to 50%


is suggested.

You can monitor and notify repository capacity and threshold using Simple Network
Management Protocol (SNMP) traps. You can set notification for any percentage of free
repository space with a default notification at 15% free and 0% free. Also, you can convert
and send these messages to QSYSOPR messages using the ACS toolkit. For more detailed
information, see in 10.2, “DS storage management” on page 178.

SVC/V7000 thin provisioning


For SVC/V7000, when you are using a fully allocated source with a thin-provisioned target,
you need to disable the background copy and cleaning mode on the FlashCopy map by
setting both the background copy rate and cleaning rate to zero. If these features are
enabled, then the thin-provisioned volume will be either offline or as large as the source. You
can select the grain size (32 KB, 64 KB, 128 KB, or 256 KB) for thin-provisioning. The grain
size that you select affects the maximum virtual capacity for the thin-provisioned volume. If
you select 32 KB for the grain size, the volume size cannot exceed 260,000 GB. The grain
size cannot be changed after the thin-provisioned volume has been created. In general,
smaller grain sizes save space and larger grain sizes produce better performance. For best
performance, the grain size of the thin-provisioned volume must match the grain size of the
FlashCopy mapping. However, if the grain sizes are different, the mapping still proceeds.

You can set the cache mode to readwrite for maximum performance when you create a
thin-provisioned volume. Also, to prevent a thin-provisioned volume from using up capacity
and getting offline, the autoexpand feature can be turned on.

156 PowerHA SystemMirror for IBM i Cookbook


9

Chapter 9. PowerHA user interfaces


There are several interfaces available for you to set up and manage your HA environment,
including the brand new PowerHA GUI, which effectively combines both the Cluster
Resource Services GUI and the High Availability Solution Manager GUI.

In this chapter we introduce you to the new PowerHA GUI and discuss the other interfaces:
 9.1, “Command line” on page 158
 9.2, “PowerHA GUI” on page 160
 9.3, “Cluster Resource Services GUI” on page 164
 9.4, “High Availability Solution Manager GUI” on page 165

© Copyright IBM Corp. 2012. All rights reserved. 157


9.1 Command line
Although there are multiple GUIs available to manage your HA environment, you might still
want to use the traditional 5250 interface for performing certain tasks either from a CL
program or a command line.

Note: The current release of IBM PowerHA for i GUI only supports the command-line
interface for working with SVC/V7000 copy services.

The IBM PowerHA for i licensed program provides IBM i command-line interfaces to
configure and manage your high-availability solution.

These are the categories of various IBM PowerHA for i commands:


 Cluster administrative domain commands
 Monitored resource entry commands
 Cluster commands
 Commands and APIs for working with ASP copy descriptions and sessions
 Geographic mirroring and iASP configuration commands

9.1.1 The Work with Cluster (WRKCLU) command


The Work with Cluster (WRKCLU) command is a good starting point because it gives you
one-stop access to most of the functions available within the HA environment. When you run
this command, the Work with Cluster Menu displays (Figure 9-1).

Work with Cluster


System: DEMOGEO1
Cluster . . . . . . . . . . . . . . . . : PWRHA_CLU

Select one of the following:

1. Display cluster information


2. Display cluster configuration information

6. Work with cluster nodes


7. Work with device domains
8. Work with administrative domains
9. Work with cluster resource groups
10. Work with ASP copy descriptions

20. Dump cluster trace

Selection or command
===>

F1=Help F3=Exit F4=Prompt F9=Retrieve F12=Cancel


Figure 9-1 Work with Cluster

158 PowerHA SystemMirror for IBM i Cookbook


The Work with Cluster page has menu options to display cluster-level information, such as
version and summary, to view and manage configuration and tuning parameters, to work with
the nodes, to work with the device and administrative domains, to work with CRGs and ASP
copy descriptions, and also an option to trace and gather debug information.

9.1.2 The Configure Device ASP command


The new Configure Device ASP (CFGDEVASP) command is part of the base operating system
and is available with PTF SI44141. The Configure Device ASP command can be used to
create or delete an independent auxiliary storage pool (ASP).

When used with the *CREATE action, this command does as follows:
 Creates the independent ASP using the specified non-configured disk units
 Creates an ASP device description by the same name if one does not already exist

When used with the *DELETE action, this command does as follows:
 Deletes the independent ASP
 Deletes the ASP device description if it was created by this command

See 2.2, “Creating an IASP” on page 18, for more information about creating an
independent ASP.

Configure Device ASP (CFGDEVASP)

Type choices, press Enter.

ASP device . . . . . . . . . . . > PWRHA_ASP1 Name


Action . . . . . . . . . . . . . > *CREATE *CREATE, *DELETE
ASP type . . . . . . . . . . . . *PRIMARY *PRIMARY, *SECONDARY, *UDFS
Primary ASP device . . . . . . . Name
Protection . . . . . . . . . . . *NO *NO, *YES
Encryption . . . . . . . . . . . *NO *NO, *YES
Disk units . . . . . . . . . . . *SELECT Name, *SELECT
+ for more values

Additional Parameters

Confirm . . . . . . . . . . . . *YES *YES, *NO

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 9-2 Configure Device ASP (CFGDEVASP) command

Note: The Configure Device ASP (CFGDEVASP) command does not rely on SST/DST access.
However, the user profile must still have *IOSYSCFG and *SERVICE special authorities.

Chapter 9. PowerHA user interfaces 159


9.1.3 Configure Geographic Mirroring command
The Configure Geographic Mirror (CFGGEOMIR) command (Figure 9-3) can be used to create a
geographic mirror copy of an existing independent ASP. This command can also create ASP
copy descriptions if they do not already exist and can start an ASP session. It performs all the
configuration steps necessary to take an existing standalone independent ASP and create a
geographic mirror copy.

Configure Geographic Mirror (CFGGEOMIR)

Type choices, press Enter.

ASP device . . . . . . . . . . . > PWRHA_ASP1 Name


Action . . . . . . . . . . . . . > *CREATE *CREATE, *DELETE
Source site . . . . . . . . . . * Name, *
Target site . . . . . . . . . . * Name, *
Session . . . . . . . . . . . . PWRHA_SSN1 Name, *NONE
Source ASP copy description . PWRHA_CPY1 Name
Target ASP copy description . PWRHA_CPY2 Name
Transmission delivery . . . . . *ASYNC *SYNC, *ASYNC
Disk units . . . . . . . . . . . *SELECT Name, *SELECT
+ for more values

Additional Parameters

Confirm . . . . . . . . . . . . *YES *YES, *NO


Cluster . . . . . . . . . . . . * Name, *
Cluster resource group . . . . . * Name, *
More...
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 9-3 Configure Geographic Mirror (CFGGEOMIR) command

9.2 PowerHA GUI


The new PowerHA GUI combines key features from the existing Cluster Resource
Services and High Availability Solutions Manager GUIs and is accessed via web-based IBM
Systems Director Navigator for i. PowerHA GUI can be ordered via 5799 HAS PRPQ and is
available for customers running 7.1. See 8.1.2, “PRPQ ordering information” on page 134, for
more information.

This new GUI makes configuration and management of a high availability environment easy
to use and intuitively obvious. The overall status of the high availability environment can be
easily determined and supports the ability to drill down to detect and correct specific problems
in the environment. The GUI enables basic functionality and allows configuration and
management of a geographic mirroring or DS8000-based replication environment.

Note: We recommend that PowerHA GUI is used for managing your high-availability
environment going forward, as the older Cluster Resource Services and High Availability
Solutions Manager GUIs will be discontinued after all of their functionality is brought into
the PowerHA GUI.

160 PowerHA SystemMirror for IBM i Cookbook


9.2.1 Accessing the PowerHA GUI
Using the new PowerHA GUI to create and mange your HA solution is straightforward and
intuitive, as it combines both the solution-based HASM GUI features and the task-based CRS
GUI features.

The PowerHA GUI is accessed via the IBM Systems Director Navigator for i from this website:
http://<system_name>:2001

You are then redirected automatically to port 2005, and your browser might issue a security
certificate warning. Choose Continue to this website and enter your user ID and password
to log in to IBM Systems Director Navigator for i. Expand the IBM i Management tasks tree to
see the new PowerHA option (Figure 9-4).

Figure 9-4 IBM Systems Director Navigator for i: PowerHA option

When PowerHA is accessed for the first time in the current session, it will look for any existing
configuration and inspect the HA environment (Figure 9-5).

Figure 9-5 PowerHA: Collecting information about the HA environment

See Chapter 11, “Creating a PowerHA base environment” on page 199, for the steps for
creating a PowerHA base environment.

9.2.2 Managing the cluster


All of the cluster management functions that are available via options 1 and 2 on WRKCLU are
also available on the PowerHA GUI. The pull-down menu next to the cluster name allows you

Chapter 9. PowerHA user interfaces 161


to display and modify cluster properties or to check whether the requirements for the HA
environment are satisfied.

Cluster properties
The cluster properties option of PowerHA GUI (Figure 9-6) gives you similar access as the
functions on the CHGCLU and CHGCLUVER command interfaces.

Figure 9-6 Cluster pull-down menu

162 PowerHA SystemMirror for IBM i Cookbook


Figure 9-7 shows the cluster properties of our demo environment.

Figure 9-7 Cluster Properties

Check Requirements option


The check requirements option of the cluster level pull-down menu runs through various
checks and highlights any condition that might prevent a failover execution.

Note: The Current PowerHA version must be 2.1 for most of the options on PowerHA GUI
to work, including the Check Requirements function.

Chapter 9. PowerHA user interfaces 163


In addition to showing the warnings and suggestions, PowerHA GUI also gives you an option
to fix the condition without having to leave the page (Figure 9-8).

Figure 9-8 Check Requirements: Fix option

Note: Choosing the fix option from the PowerHA warning message panel modifies the
QGPL/QBATCH job queue entry for the QSYS/QBATCH subsystem from a shipped
default value of 1 to *NOMAX.

Delete Cluster option


When you choose the Delete Cluster menu option, a warning message box (Figure 9-9)
displays. If you choose Yes, PowerHA deletes a cluster from all nodes currently in the cluster’s
membership list.

Figure 9-9 Delete cluster warning

9.3 Cluster Resource Services GUI


The Cluster Resource Services graphical user interface (GUI) is provided by IBM Systems
Director Navigator for i. This GUI provides you with a task-based approach for setting up and
maintaining a high-availability environment that uses PowerHA for i.

164 PowerHA SystemMirror for IBM i Cookbook


The Cluster Resource Services GUI interface (Figure 9-10) allows you to configure and
manage clustered resources and environments. Unlike the HASM GUI discussed in 9.4,
“High Availability Solution Manager GUI” on page 165, this cluster interface is based on
task-oriented goals. This interface allows you to take the following actions:
 Create and manage a cluster.
 Create and manage cluster nodes.
 Create and manage cluster resource groups (CRGs).
 Create and manage cluster administrative domains.
 Create and manage Monitored Resource Entries (MREs).
 Monitor the cluster status for high-availability-related events such as failover, cluster
partitions, and so on.
 Perform manual switchovers for planned outages, such as backups and scheduled
maintenance of software or hardware.

For more information about Cluster Resource Services GUI, see Chapter 7 of the IBM
Redbooks publication Implementing PowerHA for IBM i, SG24-7405:
http://www.redbooks.ibm.com/abstracts/sg247405.html

Figure 9-10 Cluster Resource Services GUI

9.4 High Availability Solution Manager GUI


The PowerHA for i High Availability Solutions Manager graphical interface provides a
solution-based approach to selecting, configuring, and managing your HA environment.

The High Availability Solutions Manager graphical interface (Figure 9-11 on page 166)
provides several predefined solutions. For each of these solutions, the dependent
technologies are configured based on your selection. The High Availability Solutions Manager
graphical interface provides easy-to-use tools to manage your high-availability solution.

Unlike the CRS GUI discussed in 9.3, “Cluster Resource Services GUI” on page 164, HASM
GUI employs a solution-based approach.The GUI operations are deployed via IBM Systems
Director Navigator for i, a web-based console that allows you to take the following actions:
 Select a high-availability solution.
 Verify requirements for your high-availability solutions.
 Set up a high availability solution.
 Manage a high availability solution.

For more information about configuring and managing your high-availability solution using the
HASM GUI, see Chapter 6 of Implementing PowerHA for IBM i, SG24-7405:
http://www.redbooks.ibm.com/abstracts/sg247405.html

Chapter 9. PowerHA user interfaces 165


Note: High Availability Solution Manager is an all-or-nothing type of tool. That is, to be able
to manage a solution, it must have been set up with the HASM GUI. You cannot manage
your solution if it was set up outside of this tool.

Figure 9-11 High Availability Solutions Manager GUI

166 PowerHA SystemMirror for IBM i Cookbook


10

Chapter 10. Advanced Copy Services for


PowerHA
Advanced Copy Services (ACS) for PowerHA is an extension to IBM PowerHA SystemMirror
for i that provides enhanced functionality and automation through greater integration and
management of an IBM System Storage DS6000/DS8000 Copy Services environment.

In this chapter we describe the following major functional enhancements provided by ACS:
 Easy-to-use interface for managing a DS storage and Copy Services environment
including externalized DSCLI commands, additional sanity checks, and scripting, which
can be used also for non-IBM i workloads (for example, WRKCSE and STRDSMGT)
 Support for customized programming and faster resolution for Copy Services-related
issues by using default scripts provided and maintained for the Copy Services environment
 Full automation of the FlashCopy process from the backup node
 Additional safety measures and checks to prevent users from accessing inconsistent IASP
data after a FlashCopy relationship ended
 Verification tests for PPRC switch-readiness, which can be user-initiated (CHKPPRC)
 Ability to view and manage DS storage host connections and associated volumes
 Simplified creation of new host connection when replacing a defective Fibre Channel IOA
 Detailed audit trail for debugging issues (VIEWLOG)
 Delivery of DS storage SNMP messages to the IBM i QSYSOPR message queue
 Automated creation of a D volume FlashCopy at a Global Mirror target site
 Support for Metro/Global Mirror (MGM) 3-site disaster recovery solutions
 Support for integration with TPC-R in Metro Mirror or Metro/Global Mirror environments

ACS is implemented as a service offering from IBM STG Lab Services and is supported via
the regular IBM i software support structure. For further information about ACS see the IBM
STG Lab Services website for Power Systems here:
http://www-03.ibm.com/systems/services/labservices/platforms/labservices_power.html

© Copyright IBM Corp. 2012. All rights reserved. 167


10.1 Advanced Copy Services interface
The implementation of Advanced Copy Services for PowerHA on top of IBM PowerHA
SystemMirror for i requires IBM i clustering to be set up using regular PowerHA commands,
with the exception of the cluster resource group for which ACS requires a data CRG to be
created for a switchable IASP using CRTCSECRG.

ACS requires the user QLPAR on each IBM i cluster node and a user qlpar on the
DS6000/DS8000 systems. Instead of the user having to specify a user ID and password for
any of the ACS functions, it uses a DSCLI password file, which is set up only once from the
ACS storage management interface, as like described in 10.2, “DS storage management” on
page 178.

The main entry point to the ACS user interface for Copy Services is the Work with Copy
Services Environments accessed by WRKCSE (Figure 10-1).

Copy Services Environments

Type options, press Enter.


1=Add 2=Change 4=Delete 5=Display 12=Work with
14=List Stream files 16=Define host connections 18=Make PPRC Paths

Opt Name Type Text

DEMOGM GMIR
FLASH01 FLASH
FLASH02 FLASH
TOOLKIT FLASH
TOOLKIT MMIR

Bottom
Command
===>
F1=Help F3=Exit F4=Prompt F9=Viewlog F12=Cancel
Figure 10-1 ACS Copy Services Environments panel (WRKCSE)

As the available options from Work with Copy Services Environments shows, the user can
create a new Copy Services Environment or change, delete, display, or manage an existing
Copy Services Environment for Metro Mirror, Global Mirror, FlashCopy, or LUN-level switching.

168 PowerHA SystemMirror for IBM i Cookbook


When displaying a Copy Services environment with ACS using option 5=Display from the
WRKCSSE panel (Figure 10-2), all relevant information is presented on one panel for both source
and target relationships (in our case, for the nodes DEMOPROD and DEMOHA). ACS does not
use ASP sessions like PowerHA, but under the cover still creates ASP copy descriptions.

Display a PPRC Environment


Press Enter to continue.

Environment . . . . . . : TOOLKIT
Type . . . . . . . . . . : MMIR
ASP Device name . . . . : TOOLKIT
Source Copy Description : TKMMIRPS
Target Copy Description : TKMMIRPT
TPC Replication . . . . : *NO

Source node . . . . . . : DEMOPROD


Target node . . . . . . : DEMOHA
Primary ASP . . . . . . : 222
Source device . . . . . : IBM.2107-75AY031
Target device . . . . . : IBM.2107-75AY032

Source hmc1 . . . . . . : 9.5.168.55


Target hmc1 . . . . . . : 9.5.168.55
Volume sets . . . . . . : 6
PPRC Paths . . . . . . . : 2

Bottom
Volume relationships:
DEMOPROD DEMOHA
Volumes Volumes
7000-7002 7000-7002
7100-7102 7100-7102

Bottom
F1=Help F3=Exit F8=PPRC Paths F12=Cancel
Figure 10-2 ACS displaying a PPRC environment panel

The ACS has built-in checks to prevent the user from erroneously updating copy descriptions
by favoring the remote replication copy descriptions over FlashCopy descriptions and stores
all Copy Services environment data in the device domain data to make sure that it is kept
consistent within the IBM i cluster. It also provides protection against a user accidentally
manually changing the ASP copy descriptions instead of changing them via the ACS Copy
Services environment interface by refusing any ASP copy description changes with the
following message:
Copy Descriptions and ACS data do not match.

Chapter 10. Advanced Copy Services for PowerHA 169


Using option 12 (Work with Volumes) from the Work with Copy Services environment panel
(Figure 10-3) helps the user to manage the DS Copy Services configuration by working with
volumes, setting directions for replication after a failover, suspending/resuming PPRC, or
viewing out-of-sync tracks.

Work with MMIR Environment

Environment . : TOOLKIT Status . . . . : Running


Direction . . : Normal

Select one of the following:

2. Pause
3. Resume

6. Start Replication after failover

12. Work with Volumes


13. Display Out of Sync sectors
14. List Stream files

Selection

F1=Help F3=Exit F5=Refresh Status F9=Viewlog F12=Cancel


Figure 10-3 ACS Work with MMIR Environment

Using option 12 (Work with Volumes) for a Metro Mirror environment, the user can view the
PPRC state and suspend or resume Metro Mirror volume relationships (Figure 10-4).

Work with MMIR PPRC Volumes

Environment . : TOOLKIT Direction . . : Normal


Copy Service Source device : IBM.2107-75AY031
Type . . . . : MMIR Target device : IBM.2107-75AY032

Type Volume options; 2=Pause, 3=Resume, press Enter.

Opt Src : Tgt Preferred Source Status Preferred Target Status


7000:7000 Full Duplex - Metro Mirror 70 Target Full Duplex - Metro
7001:7001 Full Duplex - Metro Mirror 70 Target Full Duplex - Metro
7002:7002 Full Duplex - Metro Mirror 70 Target Full Duplex - Metro
7100:7100 Full Duplex - Metro Mirror 71 Target Full Duplex - Metro
7101:7101 Full Duplex - Metro Mirror 71 Target Full Duplex - Metro
7102:7102 Full Duplex - Metro Mirror 71 Target Full Duplex - Metro

Bottom
F1=Help F3=Exit F5=Refresh Status F9=Viewlog F12=Cancel
Figure 10-4 ACS Work with MMIR PPRC Volumes panel

170 PowerHA SystemMirror for IBM i Cookbook


Using option 14 (List Stream files) shows all the stream files for DSCLI profiles and scripts
that get automatically created by ACS for a Copy Services environment with the preferred
source DS storage system designated as PS and the preferred target designated as PT.
These stream files are not supposed to be changed by the user, as ACS changes them any
time that the user invokes an option of ACS, but they can be used by the user to easily view
information about the DS8000 Copy Services configuration or used for custom programming.

CS Environment Stream Files


Type options; 2=Change, 4=Delete, 5=Display, 9=Run, press Enter.

Opt Stream file name IFS directory


pprc_PS.profile ...profiles/TOOLKIT_MMIR
pprc_PT.profile ...profiles/TOOLKIT_MMIR
pprc_9_PS.profile ...profiles/TOOLKIT_MMIR
pprc_9_PT.profile ...profiles/TOOLKIT_MMIR
runds_PS.profile ...profiles/TOOLKIT_MMIR
runds_PT.profile ...profiles/TOOLKIT_MMIR
chhostconn_Add.script ...scripts/TOOLKIT_MMIR
chhostconn_Drop.script ...scripts/TOOLKIT_MMIR
failoverpprc_to_PS.result ...scripts/TOOLKIT_MMIR
failoverpprc_to_PS.script ...scripts/TOOLKIT_MMIR
failoverpprc_to_PT.script ...scripts/TOOLKIT_MMIR
lsavailpprcport_PS.result ...scripts/TOOLKIT_MMIR
lsavailpprcport_PS.script ...scripts/TOOLKIT_MMIR
lsfbvol_PS.result ...scripts/TOOLKIT_MMIR
lsfbvol_PS.script ...scripts/TOOLKIT_MMIR
More...
Command
===>
F1=Help F3=Exit F4=Prompt F9=Viewlog F12=Cancel
Figure 10-5 ACS Copy Services Environment Stream Files panel

However, while working within option 14 (List Stream files), ACS does not change the stream
files, so the user can still make adjustments before invoking a stream file as a DSCLI script
using option 9 (Run) with its output shown on the panel. This directory also has the *.result
files from any ACS command function, like CHKPPRC or SWPPRC, which is where a user
would find DSCLI errors if the command fails. The user can also run one of the existing
stream file profiles with option 9 (Run) to quickly get an interactive DSCLI session with the
corresponding DS storage system.

Chapter 10. Advanced Copy Services for PowerHA 171


Using option 18 (Make PPRC Paths) from the Work with Copy Services Environments
(WRKSCE) panel determines the available PPRC paths between the DS storage systems
and lets the user create the PPRC paths in both directions for all logical subsystems (LSSs)
in the configured environment (Figure 10-6).

Available PPRC Paths

Environment . : TOOLKIT Source device : IBM.2107-75AY031


Type . . . . . : MMIR Target device : IBM.2107-75AY032

Select all connection pairs to be used, press Enter.


These selections replace all paths currently in use for this environment.
1=Select

Opt PPRC Connection Path


I0040 : I0240
I0140 : I0340

Bottom
F1=Help F3=Exit F5=Refresh Status F9=Viewlog F12=Cancel
Figure 10-6 ACS Available PPRC Paths panel

To easily set up Metro Mirror remote replication on the DS storage system with ACS, the user
can simply use a combination of the WRKCSE option 18 (Make PPRC Paths) and run the
corresponding mkpprc_from_PS.script that ACS does automatically generates based on the
DS storage system configuration information that the user has already provided when
creating the Copy Services environment.

For setting up Global Mirror remote replication on the DS storage system with ACS, the user
can simply invoke the mkpprc_GM_from_PS.script, the mksession scripts for both sides, the
chsession_GM_add_PS.script to add the primary volumes to the Global Mirror session, the
mkflash_GM_CG_PT.script to create the FlashCopy consistency group C volumes, and
finally the mkgmir_PS.script without ever needing to bother with manually entering complex
DSCLI commands.

172 PowerHA SystemMirror for IBM i Cookbook


When setting up a Global Mirror environment, the user has the option (Figure 10-7) to specify
a FlashCopy D volume on the Global Mirror target site for backup or testing purposes and
also to set up an asymmetrical Global Mirror environment (that is, without FlashCopy
consistency group C volumes on the preferred primary site). As with native PowerHA, also for
ACS the failback with an asymmetrical Global Mirror setup is a manual step, but ACS will tell
the user which scripts to run when invoking the failback option.

Display a PPRC Environment


Press Enter to continue.

Symmetric . . . . . . . : *YES
D-Copy Flash normal . . : *YES
D-Copy Flash reversed . : *NO
Extra CG Flash . . . . . : *NO
Override Master LSS . . : *YES
Source Mstr LSS . . . . : 02
More...
Volume relationships:
DEMOPROD DEMOHA DEMOHA DEMOPROD
PPRC Vols PPRC Vols CG Flash Vols CG Flash Vols
0200-0201 0200-0201 0250-0251 0250-0251
0300-0301 0300-0301 0350-0351 0350-0351

Bottom
F1=Help F3=Exit F8=PPRC Paths F12=Cancel
Figure 10-7 ACS Display a PPRC Environment panel for Global Mirror

Chapter 10. Advanced Copy Services for PowerHA 173


The user invokes FlashCopy for the D volumes on the Global Mirror target site either using
the Work with Copy Services Environments option 12 (Work with) for the Global Mirror
environment (Figure 10-8) or using STRFLASH with *GMIRTARGET (Figure 10-9 on page 175)
to create a consistent D volume involving a Global Mirror failover and FlashCopy fast reverse
restore from C to B and FlashCopy from B to D. The ACS Copy Services environment name
for the D copy needs to match its Global Mirror environment name.

Work with GMIR Environment

Environment . : DEMOGM GMIR Status . : Running


Direction . . : Normal PPRC Status . : Running

Select one of the following:

2. Pause
3. Resume
4. Failover
5. Symmetrical switchover

7. Make target D-Copy

12. Work with Volumes


13. Display Out of Sync sectors
14. List Stream files
Bottom
Selection

F1=Help F3=Exit F5=Refresh Status F9=Viewlog F12=Cancel


Figure 10-8 ACS Work with GMIR Environment panel

174 PowerHA SystemMirror for IBM i Cookbook


Start a FlashCopy Backup (STRFLASH)

Type choices, press Enter.

Environment name . . . . . . . . DEMOGM Name


From volumes . . . . . . . . . . *GMIRTARGET *SOURCE, *GMIRTARGET

Bottom
F3=Exit F4=Prompt F5=Refresh F10=Additional parameters F12=Cancel
F13=How to use this display F24=More keys
Figure 10-9 ACS Start a FlashCopy Backup panel (STRFLASH)

Chapter 10. Advanced Copy Services for PowerHA 175


Creating a data CRG for ACS is done using CRTCSECRG, which also allows you to set up full
FlashCopy automation, which automatically runs the CHGASPACT to quiesce/resume the
IASP before and after taking the FlashCopy, as shown with the Quiesced flash *YES setting
in the CHGCSEDTA output shown in Figure 10-10.

Change CSE CRG Data

Supply all required values, press Enter.

FlashCopy information:
FlashCopy node . . . . . . . . DEMOFC Name
Status . . . . . . . . . . . . *FLASHED *NONE, *FLASHED, number
Warm flash . . . . . . . . . . *YES *YES, *NO
Incremental flash . . . . . . *NO *YES, *NO
Quiesced flash . . . . . . . . *YES *YES, *NO

Second FlashCopy information:


FlashCopy node . . . . . . . . Name

Third FlashCopy information:


FlashCopy node . . . . . . . . Name

More...
F1=Help F3=Exit F12=Cancel
Figure 10-10 ACS Change CSE CRG Data panel (CHGCSEDTA)

At any time the user can verify the PPRC switch-readiness of an ACS Copy Services
environment by running CHKPPRC, which communicates with the DS storage system, similar to
displaying an ASP session in PowerHA, to verify that PPRC is running full-duplex and is set
up correctly with regard to the ACS PPRC setup.

176 PowerHA SystemMirror for IBM i Cookbook


A switch of PPRC is by design generally not automated by ACS. Rather, it can be explicitly
switched using SWPPRC (Figure 10-11).

Switch PPRC (SWPPRC)

Type choices, press Enter.

Environment name . . . . . . . . > TOOLKIT Name


Switch type . . . . . . . . . . > *UNSCHEDULED *SCHEDULED, *UNSCHEDULED...
Type . . . . . . . . . . . . . . * *, *GMIR, *LUN, *MMIR
Auto Vary On . . . . . . . . . . *YES *YES, *NO
Auto replicate . . . . . . . . . *DFT *DFT, *YES, *NO
Switch paused MMIR . . . . . . . *NO *YES, *NO

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 10-11 ACS Switch PPRC command panel (SWPPRC)

When using the Auto replicate *DFT option, for a scheduled Metro Mirror switch, the
command looks at the CSE configuration parameter Automatic PPRC Replicate (whose
default value is *YES) to determine whether the PPRC direction will be established in the
reverse direction. For an unscheduled Metro Mirror switch, Auto replicate *DFT means *NO.
That is, no replication in the reverse direction is started. For Global Mirror it always changes
the replication direction for a scheduled switch, and also for an unscheduled switch if the
source DS storage system is available. ACS allows the user to also switch a paused PPRC
relationship, but only if all PPRC volumes are in the same paused state.

ACS is installed in the QZRDHASM library with its logs written to


/QIBM/Qzrdhasm/qzrdhasm.log, which can easily be viewed using VIEWLOG.

DMPINF allows the user to dump all ACS Copy Services environment information, such as
DSPCSEDTA and CSE scripts joblogs, to a stream file like
/QIBM/Qzrdhasm/qzrdhasm_DEMOPROD_110919_1202.txt.

Chapter 10. Advanced Copy Services for PowerHA 177


10.2 DS storage management
ACS provides an integrated DS storage management interface accessed via STRDSMGT, which
after configuring a DS storage system connection the user can use to easily manage DS host
connections, check volumes, Copy Services, or multi-path configurations; configure SNMP
trap notifications; or manage DS storage system authentications (Figure 10-12).

Storage Management Menu


System: CTCIHA8Y
Storage Device . . . . . . . . . . . . . . . . . . . . : IBM.2107-75AY031

SNMP Trap status . . . . . . . . . . . . . . . . . . . : Not active

Select one of the following:

1. Configure Storage Management


2. Work with Storage Connections
3. Save Storage Connections
4. Delete Saved Connections
5. Display Saved Connections
6. Compare Saved Connections

8. Check Storage Volumes


9. Check Copy Service Volumes

11. Check System i Multipath

14. List Storage Management files

17. Start Storage SNMP Traps


18. Stop Storage SNMP Traps

20. Manage Authority to Storage

Selection

F3=Exit F9=Viewlog F12=Cancel


Figure 10-12 ACS Storage Management Menu

178 PowerHA SystemMirror for IBM i Cookbook


After setting up the DS connection information with option 1 (Configure Storage
Management) (Figure 10-13), option 20 (Manage Authority to Storage) allows the user to add
any user to a DS storage system (option 1), change the qlpar user (option 2) on the DS
storage system, or automatically update the DSCLI password file on the IBM i client (option 4)
(Figure 10-14 on page 180).

Configure Storage Management

Press Enter to continue

Storage:
Device name . . . . . . . . IBM.2107-75AY031 Name
Primary HMC . . . . . . . . 9.5.168.55 IPv4
Second HMC . . . . . . . . . IPv4

SNMP Trap:
Message Queue Name . . . . . *SYSOPR name, *SYSOPR
Message Queue Library . . . name
Issue storage commands . . . *NO *YES, *NO

Verbose message logging . . . *NO *YES, *NO

Copy Services installed . . : *YES

F1=Help F3=Exit F12=Cancel


Figure 10-13 ACS Configure Storage Management panel

Chapter 10. Advanced Copy Services for PowerHA 179


Storage Authorization Management Menu
System: CTCIHA8Y
Select one of the following:

1. Add Storage User


2. Change Storage User
3. Set password expiration interval
4. Update password file

Selection

F3=Exit F12=Cancel
Figure 10-14 ACS Storage Authorization Management Menu

180 PowerHA SystemMirror for IBM i Cookbook


Option 2 (Work with Storage Connections) from the Storage Management Menu builds a map
from the IBM i Fibre Channel IOAs to the DS showing all the DS host connection, port login,
and volume group information.

Work with Disk Connections

Storage Device . . . . . . . . . . . . . . . . : IBM.2107-75AY031

Type options, press Enter.


5=Display 10=mkhostconnect 11=Replace IOA

Resource IOA IOA IOA Host Login Volume


Opt Name Type Bus Slot Host Connect Name ID Port Group
DC07 280E 768 C2 8YTPCR 0029 I0010 V37
DC03 280E 775 C5 8yTPCMGM 000F I0110 V25
DC01 576B 293 C5 0 HKTEST 0004 I0010 V46
DC01 576B 293 C5 1 *****
DC04 6B25 255 C4 *****

Bottom
Command
===>
F1=Help F3=Exit F5=Refresh F12=Cancel F17=Save F18=Display saved
F21=Print report
Figure 10-15 ACS Work with Disk Connections panel

Chapter 10. Advanced Copy Services for PowerHA 181


Option 5 (Display) on the Work with Disk Connections panel allows the user to display details
for an IBM i Fibre Channel connection to the DS showing all the volume and WWPN
information (Figure 10-16).

Connection Details

Press Enter to continue

Resource Name . : DC03 HostConnect Name: 8yTPCMGM

IO Adapter: Host ID . . . . : 000F


Type . . . . : 280E Login Port . . : I0110
Bus . . . . . : 775 Volume Group . : V25
Slot . . . . : C5
Storage Port
System i Tower : U5790.001.12940A1 WWPN . . . . : 50050763070902B8
System i WWPN . : 10000000C967E453
Volumes . . . . : 8

Volume IDs . . : EA00 EA01 EA02 EA03 EB00 EB01 EB02 EB03

Bottom
F1=Help F3=Exit F12=Cancel
Figure 10-16 ACS Connection Details panel

Using F17 (Save) on the Work with Disk Connections panel allows the user to save a known
good configuration, so whenever a link failure occurs (represented by a **** not logged-in
information), the user can easily debug this connection from getting the original port login
information back with F18 (Display saved).

Option 11 (Replace IOA) on the Work with Disk Connections panel can be used to have the
host connections on the DS automatically updated (deleted and recreated) with the new
WWPN information after a replacement of a Fibre Channel IOA.

182 PowerHA SystemMirror for IBM i Cookbook


Option 9 (Check Copy Services Volumes) from the Storage Management Menu allows the
user to have ACS verify the ACS Copy Services environment configuration with the IASP
configuration. A conflict would be reported as such with the following panel message
(Figure 10-17):
System i and Copy Services volumes in conflict

Check Copy Services Volumes

Make selections below to subset the information processed.


-
Press Enter to continue

System i:
ASP list . . 179 ASP numbers

Copy Services Environment:


Name . . . . TOOLKIT Name
Type . . . . MMIR FLASH, GMIR or
MMIR

F1=Help F3=Exit F4=Prompt F5=Display report F12=Cancel


F14=Command string
System i and Copy Services volumes in conflict.
Figure 10-17 ACS Check Copy Services Volumes panel

With the option F5 (Display report) a detailed log can be displayed (Example 10-1) where the
ACS expects IASP volumes in LSS 70/71 while the actual IASP 179 is configured on LSS 02/03.

Example 10-1 /QIBM/Qzrdhasm/Management/CopyServiceVolume.report


************Beginning of data**************
#
# Storage Management - Copy Services Volume Report
#
# Generated: Mon Sep 19 12:34:12 2011
#
# Selection input:
# Storage System . . : IBM.2107-75AY031
# System i ASPs . . . . . . . . : 179
# Copy Services Environment Name: TOOLKIT
# Type: MMIR
#
System i Volume Copy Services . . . . . .
ASP ID Environment Type Role
-------- ---- ---------- ----- ----
Error ******** 7000 TOOLKIT MMIR PSrc
Error ******** 7001 TOOLKIT MMIR PSrc

Chapter 10. Advanced Copy Services for PowerHA 183


Error ******** 7002 TOOLKIT MMIR PSrc
Error ******** 7100 TOOLKIT MMIR PSrc
Error ******** 7101 TOOLKIT MMIR PSrc
Error ******** 7102 TOOLKIT MMIR PSrc
Warning 179 0200 ********** ***** ****
Warning 179 0201 ********** ***** ****
Warning 179 0300 ********** ***** ****
Warning 179 0301 ********** ***** ****
------------------------------------------------------
Summary:
System i volumes: 4
Matching Copy Services volumes: 0
Copy Services volumes not on System i: 6
System i volumes not in Copy Services: 4
Copy Services volumes: 6
************End of Data********************

Option 11 (Check System i Multipath) from the Storage Management Menu allows the user to
check whether the IASP disk units report in as multipath units since the last IPL or IASP
switch (that is, multi-path reset) (Example 10-2).

Example 10-2 /QIBM/Qzrdhasm/Management/MultiPath.report


************Beginning of data**************
#
# Storage Management - Multipath Report
#
# Generated: Mon Sep 19 12:40:42 2011
#
# Selection input:
# Storage System . . : IBM.2107-75AY031
# System i ASPs: *ALL
#
# Non-Multipath volumes are listed below:
#

ASP 179: 0200 0201 0300 0301


ASP 180: 6000 6002 6100 6101 6001 6102
--------------------------------------------
Summary:
Non-Multipath System i volumes: 10
************End of Data********************

184 PowerHA SystemMirror for IBM i Cookbook


The ACS toolkit allows you to receive SNMP messages from the DS storage system for
System Reference Code errors or Copy Services events and converts them to QSYSOPR
messages (Figure 10-18), which customers usually monitor any time.

Additional Message Information

Message ID . . . . . . : IAS1202 Severity . . . . . . . : 80


Message type . . . . . : Diagnostic
Date sent . . . . . . : 10/03/11 Time sent . . . . . . : 10:02:42

Message . . . . : *Primary remote mirror and copy devices on an LSS were


suspended.
Cause . . . . . : A PPRC volume pair on the storage decice was suspended due
to an error. . Use the VIEWLOG command for specific details.
Recovery . . . : Use the appropriate Copy Services management tools on the
storage HMC to analyze the problem.

Bottom
Press Enter to continue.

F3=Exit F6=Print F9=Display message details F12=Cancel


F21=Select assistance level
Figure 10-18 ACS QSYSOPR SNMP message

Corresponding detailed SNMP trap information is logged by ACS in the


/QIBM/Qzrdhasm/qzrdsnmp.log file (Example 10-3), which can easily be viewed using
VIEWLOG *SNMP.

Example 10-3 ACS SNMP trap log file


2011-10-03 10:02:42 DSM-SNMPTRAP: public, type = 6/202, bindings = 1
2011/10/03 10:10:39 CDT
Primary PPRC Devices on LSS Suspended Due to Error
UNIT: Mnf Type-Mod SerialNm LS LD SR
PRI: IBM 2107-9B2 75-AY031 EA 00 03
SEC: IBM 2107-000 75-AY032 EC 00
Start: 2011/10/03 10:10:39 CDT
PRI Dev Flags (1 bit/Dev, 1=Suspended):
8000000000000000000000000000000000000000000000000000000000000000

Chapter 10. Advanced Copy Services for PowerHA 185


10.3 FlashCopy on Global Mirror target site
Advanced Copy Services for PowerHA extends the Global Mirror base functionality provided
by PowerHA SystemMirror for i while also allowing a FlashCopy to be taken from a Global
Mirror target site (for example, for a tape backup done at the remote site) (Figure 10-19).

IBM Power Systems IBM Power Systems


Production Backup

IBM Power Storage IBM Power Storage


Primary Secondary
Global Mirror
Session
SYSBAS SYSBAS

Production Backup
LPAR A Volumes Global Copy B Volumes LPAR
FlashCopy FlashCopy
Consistency Consistency
Groups Groups
FlashCopy only FlashCopy
C Volumes C Volumes
after GM Failover LPAR
and FRR from B
D Volumes

SYSBAS

Figure 10-19 ACS FlashCopy on Global Mirror target site

Configuration of a FashCopy D volume with ACS is done when creating a Global Mirror
environment (Figure 10-7 on page 173) by specifying D-Copy Flash normal *YES to set up a
FlashCopy D volume configuration on the Global Mirror target site, and by specifying D-Copy
Flash reversed *YES to allow for a FlashCopy on the target site also when the Global Mirror
direction is reversed (that is, the DS storage system at the target or secondary site acts as the
Global Mirror source).

Creating a FlashCopy to the D volumes, which involves an automatic prior Global Mirror
failover and FlashCopy fast-reverse-restore from the C volumes to the B volumes before
consistent data is on the B volumes, which can finally be flashed to the D volumes, can be
initiated by the user either with an option from the Global Mirror environment (Figure 10-8 on
page 174) or by using STRFLASH (Figure 10-9 on page 175).

186 PowerHA SystemMirror for IBM i Cookbook


10.4 Metro/Global Mirror and TPC-R support
When an IBM System Storage DS8000 3-site disaster recovery solution with Metro/Global
Mirror (MGM) support is desired (Figure 10-20), it is supported with ACS only with the IBM
Tivoli® Storage Productivity Center for Replication (TPC-R) software product.

IBM Power Systems IBM Power Systems IBM Power Systems


Production Backup 1 Backup 2

Production Backup 1 Backup 2


LPAR LPAR LPAR

IBM Power Storage TPC-R IBM Power Storage DS8000 IBM Power Storage
Primary Secondary Tertiary
Metro/Global Mirror Global Mirror
Session Session

SYSBAS SYSBAS SYSBAS

MM Target/
MM Source Metro Mirror GC Source Global Copy GC Target

GM FLC CG

Site 1 Site 2 Site 3

Figure 10-20 DS8000 3-site disaster recovery solution with Metro/Global Mirror

Chapter 10. Advanced Copy Services for PowerHA 187


Due to its session concept, TPC-R makes managing consistency across multiple Copy
Services volume sets and failover or corrective actions for a session in a failed state easy
(Figure 10-21).

Figure 10-21 TPC-R Metro/Global Mirror session

Other reasons for using TPC-R with ACS are if a customer likes to use Metro Mirror
consistency groups or uses TPC-R already with another environment and wants the IBM i
server platform included. Even with configured TPC-R support, for FlashCopy ACS always
uses its DSCLI interface. Similarly, ACS also does not make use of TPC-R’s Metro/Global
Mirror with Practice session, as it always fails over to the practice volume instead of the
GlobalCopy target volumes.

188 PowerHA SystemMirror for IBM i Cookbook


ACS interaction with TPC-R is supported for either Metro Mirror or Metro/Global Mirror such
that the user sets up communication with the TPC-R server for the ACS Copy Services
environment (Figure 10-22).

Change a GMIR Environment


Type choices, press Enter.

Environment . . . . . . : IASP01

Global Mirroring Power HA, ASP information:


Device name . . . . . . IASP01 Name
GMIR Src is MMIR Src . : *YES
Source Copy Description PROD Name
Target Copy Description DR Name

TPC information:
TPC Replication . . . . *YES *YES, *NO
Metro-Global Mirroring *YES *YES, *NO
Primary server . . . . . 9.5.167.82 IPv4
Secondary server . . . . 9.5.167.57 IPv4
Session name . . . . . . MGM Name
User . . . . . . . . . . tpcruser Profile name
Password . . . . . . . .

More...
F1=Help F3=Exit F12=Cancel
Figure 10-22 ACS Change a GMIR Environment panel

ACS will then stop using the DSCLI and instead use its Java API interface into the TPC-R
server for managing the Copy Services functions. The CSE stream file scripts for DSCLI still
get created but without accompanying result files. Instead, there are Tpcr*Rtn.xml result files
showing the return status of the ACS Java API communication with the TPC-R server.

Note: Before planning to implement a Metro/Global Mirror environment with ACS for IBM i,
contact IBM STG Lab Services, as it requires careful planning to ensure that customer
requirements are met:
http://www-03.ibm.com/systems/services/labservices/contact.html

10.5 Custom programming


ACS supports custom programming for a Copy Services environment by allowing a user to
retrieve information from the DS storage system or TPC-R server that can be further
processed by a user-written program and used to invoke desired actions either on the storage
system or the TPC-R server.

Using RUNDSCMD, a user can run a DSCLI script for retrieving the output in a comma-separated
text file that can be validated using a CL program, which in turn can make use again of the
existing ACS Copy Services environment stream files.

Chapter 10. Advanced Copy Services for PowerHA 189


For example, the verification of an existing FlashCopy relationship can be done by querying
for the DSCLI message code CMUC00234I, which is returned for a non-existing FlashCopy
relationship, as shown by the validation routine example in Figure 10-23.

Run DS Scripted Command (RUNDSCMD)

Type choices, press Enter.

Result validation list:


Column position . . . . . . . > 1 1-20
Expected value . . . . . . . . > 'CMUC00234I'
Logic to next in list . . . . *AND, *OR
+ for more values
Result file rows . . . . . . . . *ALL *ONE, *ALL
Summation column . . . . . . . . *NONE *NONE, 1-20
CL variable for returned total TYPE(*DEC) LEN(9 0)
Return column . . . . . . . . . *NONE *NONE, 1-20
Return key value . . . . . . . . *NONE
CL variable for returned value TYPE(*CHAR) LEN(80)
Comment . . . . . . . . . . . .

Bottom
F3=Exit F4=Prompt F5=Refresh F10=Additional parameters F12=Cancel
F13=How to use this display F24=More keys
Figure 10-23 ACS Run DS Scripted Command panel

Another example of using validation with RUNDSCMD is using its Summation column and CL
variable for returned total parameters to query the sum of certain values like the
out-of-sync tracks from all listed PPRC relationships (for example, to write a customized
program to predict the remaining synchronization time).

By using the Return column, Return key value, and CL variable for returned value in the
validation program, the user can search for the specified key value in the DSCLI output and
get the value at the specified return column returned in a CL variable for further processing in
customized programming.

190 PowerHA SystemMirror for IBM i Cookbook


RUNTPCACT of ACS allows a user to run any TPC-R session commands, such as ‘Start
H1->H2->H3’ or ‘Suspend’ (Figure 10-24).

Run TPC Action (RUNTPCACT)

Type choices, press Enter.

Environment name . . . . . . . . > IASP01 Name


Type . . . . . . . . . . . . . . *MMIR *GMIR, *MMIR
Action . . . . . . . . . . . . . > *RUN *CHK, *RTV, *RUN, *ACS, *RCS
TPC command . . . . . . . . . . 'Suspend'

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 10-24 ACS Run TPC Action command panel

RTVTPCCMD can be used to list the available TPC session actions as output in a CL variable,
which can be processed/verified in a CL program before running an available action to take
control of TPC-R from an IBM i client.

Chapter 10. Advanced Copy Services for PowerHA 191


10.6 IBM i full-system FlashCopy replication
STG Lab Services also offers a Full-System FlashCopy Toolkit (FSFC) for IBM i 5.4 and later
that is installed on a production partition to be cloned via FlashCopy and a managing partition
that controls the entire FlashCopy process via IP communication to the production and
FlashCopy backup partition including SSH communication through the HMC (Figure 10-25).

3
Partition returns to
1 IBM Power Systems production duties
Write "all" memory
content to disk
IBM System Storage

Production LPAR

2
Managing LPAR Sysbas
SYSBAS FlashCopy the disks

Flas
hCo
Backup LPAR

py
4 Sysbas’
SYSBAS
IPL the target
partition

5 6
Backup the The whole process is
target partition managed by the
to tape controller

Figure 10-25 Full-System FlashCopy Toolkit

FSFC is licensed separately from Advanced Copy Services for PowerHA but uses the same
Copy Services Environment menus for setting up the profiles and scripts to communicate with
the DS storage system. It needs to be installed on the managing and production partition in
the QZRDIASH5 library.

Note: While taking a warm FlashCopy with the full-system or IASP online, even when
quiesced, any object that is resident in memory and not journaled, such as files, data
queues, data areas, and so on, is subject to potential damage due to object
inconsistencies from parts of the object not on disk. Journaling allows the target partition to
recover from possible damaged journaled objects by applying journal entries to journaled
files during the IPL.

It is only with a full-system FlashCopy that there is a chance that on taking a warm
FlashCopy some IBM i internal objects can be found corrupted at IPL time. You might need
a SLIP install, or another FlashCopy might be needed for a recovery.

In addition to controlling the IBM i full-system FlashCopy process itself, FSFC also provides a
high level of integration with IBM i Backup Recovery and Media Services (BRMS) to make
sure that the backup by BRMS run from the FlashCopy backup partition appears in the BRMS
history containing all the backup and media information, as was done from the production

192 PowerHA SystemMirror for IBM i Cookbook


partition with the BRMS history (QUSRBRM) automatically transferred to the production
partition after the backup has finished (Figure 10-26).

Production system name: PROD Backup system name: PROD


BRMS system name: PROD BRMS system name: PROD

1 Stop all BRMS activities System attributes and network


6 configuration changed by IPL
startup program
Prepare BRMS
2
for FlashCopy
Set BRMS system state to
7
FlashCopy mode
Quiesce system or
PWRDWNSYS to
3 Backup system treated as
insure all main storage 8
"PROD" in BRMS network
content is written to disk

9 Complete saves using BRMS


4 FlashCopy
Set BRMS state to FlashCopy
10
backup complete mode
Resume or restart
5 system
Save QUSRBRM and transfer
11
the library to PROD system

Call post FlashCopy BRMS


12
function / restore QUSRBRM

13 Resume BRMS activities on PROD

Figure 10-26 Full-system FlashCopy toolkit integration with BRMS

The FSFC full-system FlashCopy functions are configured on the managing partition using
CRTSYSCPY (Figure 10-27 on page 194). In our example we show a FlashCopy configuration
where CTCIHA7A is the production partition to be flashed via the managing partition to the
target partition CTCIHA7C with quiescing the production partition (OPTYPE *QUIESCE)
before taking the FlashCopy, resuming the production partition, and activating the backup
partition by starting its specified IP interface and BRMS backup job.

Chapter 10. Advanced Copy Services for PowerHA 193


Create Full Sys Flash Copy (CRTSYSCPY)

Type choices, press Enter.

Configuration Name . . . . . . . FLC_HA7A Character value


Environment name . . . . . . . . HA7A F4 to prompt
Source partition host name . . . CTCIHA7A.RCHLAND.IBM.COM

Source HMC partition name . . . CTCIHA7A


Source partition profile . . . . DEFAULT
Source managing system . . . . . CTCIHA7
Source HMC1 address . . . . . . 9.5.168.169
Source HMC2 address . . . . . . *NONE
Shutdown target . . . . . . . . *YES *YES, *NO
Restart target partition . . . . *YES *YES, *NO, *INQ, *OFF
Target LPAR IPL source . . . . . *PANEL *PANEL, A, B, D
Target LPAR keylock position . . *PANEL *PANEL, *AUTO, *MANUAL
Target LPAR pre-activation pgm *NONE *NONE, name
Library . . . . . . . . . . . *LIBL *LIBL, library name

More...
Target partition host name . . . CTCIHA7C.RCHLAND.IBM.COM

Target HMC partition name . . . CTCIHA7C


Target partition profile . . . . DEFAULT
Target managing system . . . . . *SOURCE
Target HMC1 address . . . . . . *SOURCE
Target HMC2 address . . . . . . *NONE
Backup Application . . . . . . . *BRMS *NATIVE, *BRMS
Lock BRMS . . . . . . . . . . . *BOTH *BOTH, *NO, *SRCONLY...
Restricted BRMS media classes . *NONE F4 to prompt
+ for more values

More...
Target LPAR Device Config:
Backup device name . . . . . . TS3400 *NONE, device name
Backup device serial number . 78-1011003 *NONE, serial number
Robot host . . . . . . . . . . *NOCHANGE
Local internet address . . . . *NOCHANGE
+ for more values
Program to move QUSRBRM . . . . *SAVF_SOCK *SAVF_SOCK,
*VRTTAP,*NONE,name
Library . . . . . . . . . . . *LIBL *LIBL, library name
Save compression for QUSRBRM . . *DEV *DEV, *YES, *NO, *LOW...
Operation type . . . . . . . . . *QUIESCE *QUIESCE, *IPL, *NOIPL
Storage Type . . . . . . . . . . *DS8K *DS5K, *DS8K, *SVC
Issue IPL confirmation message *NO *YES, *NO
Source LPAR shutdown command . .

More...
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 10-27 Full-system FlashCopy toolkit CRTSYSCPY command panel

194 PowerHA SystemMirror for IBM i Cookbook


Figure 10-28 shows the remaining pages from the CRTSYSCPY command panel.

Minutes to wait for power down 60 minutes


Force *SYSBAS resume . . . . . . 60 seconds
Force flash copy . . . . . . . . *NO *YES, *NO
FlashCopy exit program . . . . . *NONE *NONE, name
Library . . . . . . . . . . . *LIBL *LIBL, library name
Minutes to wait for power up . . 600 minutes
Target LPAR startup program . . *NONE *ORIG, *NONE, name
Library . . . . . . . . . . . *LIBL *LIBL, library name
Startup program job queue . . . QCTL *NONE, name
Library . . . . . . . . . . . QSYS *LIBL, library name
Hold scheduled jobs on target . *YES *YES, *NO
Target LPAR default route:
Binding interface . . . . . . *NOCHANGE
Next hop . . . . . . . . . . .

More...
Target Comm Interfaces:
IO card serial number . . . . 00-EC7560A *NONE, Serial number
Line Description . . . . . . . FSFCL *NEW, line name
IO card port number . . . . . 0 0-32
IO card IP address . . . . . . 9.5.168.117
Network Mask . . . . . . . . . '255.255.255.0'
+ for more values
Target partition backup cmd . . STRBKUBRM USERDATA SBMJOB(*NO)

+ for more values

Wait for final notification . . *NO *YES, *NO


More...
Stop target after process . . . *RMV *YES, *NO, *RMV
Figure 10-28 Full-system FlashCopy toolkit CRTSYSCPY command panel (continued)

After the Copy Services environment has been created by using WRKCSE and the configuration
using CRTSYSCPY, a full-system FlashCopy can be started by using MKSYSCPY (Figure 10-29).

Make Full System Copy (MKSYSCPY)

Type choices, press Enter.

Configuration Name . . . . . . . FLC_HA7A F4 to prompt


User Profile . . . . . . . . . . POWERHA *DDM, User Profile
Password . . . . . . . . . . . . Character value

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 10-29 Full-system FlashCopy toolkit STRFLASH command panel

Chapter 10. Advanced Copy Services for PowerHA 195


For further information about the full-system FlashCopy toolkit service offering contact STG
Lab Services here:
http://www-03.ibm.com/systems/services/labservices/contact.html

196 PowerHA SystemMirror for IBM i Cookbook


Part 3

Part 3 Implementation
examples and best
practices
In this part, the final part of the book, we show you practical scenarios and
implementation examples with step-by-step instructions for how to configure and manage
your high-availability environment using IBM PowerHA SystemMirror for i.

This part has the following chapters:


 Chapter 11, “Creating a PowerHA base environment” on page 199
 Chapter 12, “Configuring and managing Geographic Mirroring” on page 235
 Chapter 13, “Configuring and managing DS8000 Copy Services” on page 263
 Chapter 14, “Configuring and managing CSVC/V7000 Copy Services” on page 371

We also discuss best practices to consider and follow in setting up and managing your
high-availability environment in Chapter 15, “Best practices” on page 397.

© Copyright IBM Corp. 2012. All rights reserved. 197


198 PowerHA SystemMirror for IBM i Cookbook
11

Chapter 11. Creating a PowerHA base


environment
In this chapter, we show you how to set up the basic building blocks of a highly available
environment using PowerHA. This chapter provides a step-by-step approach for the
following tasks:
 Creating a cluster
 Setting up cluster monitors and advanced node failure detection
 Creating an IASP
 Setting up an administrative domain

In this section, we set up a basic cluster environment with two nodes in it. We then create an
IASP on the production site. We set up advanced failure detection by adding a cluster monitor
and registering the cluster nodes with the HMC CIM server. In addition, we create an
administrative domain and add monitored resource entries to it. The entire setup is done
using the new PowerHA GUI. In addition, we provide you with the CL commands that you can
use alternatively to do this setup.

© Copyright IBM Corp. 2012. All rights reserved. 199


11.1 Creating a cluster
To create your basic cluster setup using the new PowerHA GUI, take the following steps:
1. Access and log in to IBM Systems Director Navigator for i from the following URL:
http://<your_server_ip_address>:2001
Figure 11-1 shows you the list of possible tasks.

Figure 11-1 IBM Systems Director Navigator for i: Welcome panel

200 PowerHA SystemMirror for IBM i Cookbook


2. Choosing PowerHA leads you to the new PowerHA GUI. The GUI connects to your
system and checks whether any cluster configuration has already been done. This is not
the case in our environment (Figure 11-2). Click Create a new Cluster to start the wizard.

Figure 11-2 PowerHA GUI: Create new cluster

3. The wizard guides you through the steps necessary to create the cluster and asks you for
all information required for the setup. Figure 11-3 shows the steps.

Figure 11-3 PowerHA GUI: Welcome screen

Chapter 11. Creating a PowerHA base environment 201


4. Skipping over the welcome page takes you to the page shown in Figure 11-4. Here,
you enter the name of your cluster (PWRHA_CLU in our case). PowerHA version,
cluster version, and cluster modification default to the most current levels the system that
you are connected to supports. These values can be changed. However, PowerHA
Version 2.1 with cluster Version 7 is the minimum level required for the Power HA GUI to
work properly.

Figure 11-4 PowerHA GUI: Cluster name and version

202 PowerHA SystemMirror for IBM i Cookbook


5. On the next page you enter cluster node information pertaining to the system that you are
connected to. As can be seen in Figure 11-5, the node name defaults to the current
system name value from the network attributes of the system, but it can also be changed.
The cluster IP addresses you provide here are used for cluster heartbeat and internal
cluster communication. For a production environment it is a good practice to use two
dedicated, redundant Ethernet ports with cluster heartbeat addresses, especially when
not using IBM i 7.1 advanced node failure detection to help prevent cluster partition
conditions because of single-points-of-failure in the cluster communication setup.

Figure 11-5 PowerHA GUI: Local node information

Chapter 11. Creating a PowerHA base environment 203


6. In the next step, you additional nodes into the cluster. You have to provide the node name
and the cluster IP addresses used for heartbeat (Figure 11-6). You can also specify
whether that cluster node should be started when the creation of the cluster is finished.

Note: To have clustering on a node started automatically after an IPL, change the
system’s startup program with adding a STRCLUNOD entry.

Figure 11-6 PowerHA GUI: Add additional nodes

204 PowerHA SystemMirror for IBM i Cookbook


7. Specify whether you want to define a cluster-wide failover message queue. Should the
primary node fail, a message is sent to this message queue on the node that the cluster
would fail to. You can also specify how long this message should wait for an answer and
what the default action should be if there is no answer within that time frame (Figure 11-7).
This setting is then valid for all cluster resource groups (CRGs) within this cluster.
Alternatively, you can define failover message queues for each CRG individually.

Note: If you do not specify any failover message queue, then failover happens
immediately without any message being sent in advance.

Figure 11-7 PowerHA GUI: Specify cluster message queue

Chapter 11. Creating a PowerHA base environment 205


8. As shown in Figure 11-8, the summary page provides you with an overview of the data
that you entered in the wizard. Click Finish to create the cluster.

Figure 11-8 PowerHA GUI: Create cluster summary

9. You will see the page shown in Figure 11-9 after the cluster is created and all nodes are
started. The PowerHA GUI can then help you determine whether all requirements for
successfully running a cluster environment are met. Click Check Requirements to do so.

Figure 11-9 PowerHA GUI: Cluster successfully created

206 PowerHA SystemMirror for IBM i Cookbook


10.In our example, we receive a list of warnings and suggestions (Figure 11-10). For
example, the system value QRETSVRSEC has to be set to 1 if you want to use an
administrative domain to synchronize user passwords between the nodes of your cluster.

Figure 11-10 PowerHA GUI: Check requirements

11.Clicking the toggle beside the message opens a Fix icon (Figure 11-11). If you click this
button corrective action is taken on the respective cluster node.

Figure 11-11 PowerHA GUI: Fix requirements

Chapter 11. Creating a PowerHA base environment 207


12.After all the requirements are met, you will see a page like that show in Figure 11-12.
Close this information page to continue with your cluster setup.

Figure 11-12 PowerHA GUI: Cluster requirements met

11.2 Setting up cluster monitors


Cluster monitors were introduced with IBM i 7.1. Setting them up can help to avoid cluster
partition situations that occur because a cluster node was not able to send out a panic
message before going down. To allow the HMC CIM server to notify registered IBM i cluster
nodes of sudden partition or system failures, you need to set up SSH communication
between your cluster nodes and the HMC.

Steps to set up SSH


Take these steps to set up SSH:
1. License program 5733-SC1 (IBM Portable Utilities for i) with options base and 1 have to
be installed on your cluster nodes.
2. License program 5770-UME (IBM Universal Manageability Enablement for i) has to be
installed on your cluster nodes.
3. TCP/IP server *CIMOM has to be started.
4. The *CIMOM TCP server must be configured and started on each cluster node that has a
cluster monitor configured on it. The default configuration of the *CIMOM server that is
provided by the installation of the 5770-UME license program must be changed so that the
IBM i system can communicate with the CIM server. To do that, two configuration
attributes that control security aspects need to be changed by running cimconfig within a
PASE shell:
a. With the *CIMOM server running, start a PASE shell from the command line with
call qp2term.
b. Enter:
/QOpenSys/QIBM/ProdData/UME/Pegasus/bin/cimconfig -s
enableAuthentication=false -p

208 PowerHA SystemMirror for IBM i Cookbook


c. Enter:
/QOpenSys/QIBM/ProdData/UME/Pegasus/bin/cimconfig -s
sslClientVerificationMode=optional -p
d. End the PASE shell and restart the CIMOM server using ENDTCPSVR *CIMOM and
STRTCPSVR *CIMOM.
5. Add a digital certificate from your HMC to the certificate truststore using these steps:
a. TCP server *SSHD must be running. If this is not the case start it with
STRTCPSVR *SSHD.
b. The IBM i user profile used to copy the certificate has to have a home directory
associated with the profile, and that directory must exist.
c. You must use the physical monitor and keyboard attached to your HMC. You cannot
use Telnet or a web interface to the HMC.
d. Open a restricted shell on the HMC.
e. Use the secure copy command to copy a file to your IBM i cluster node:
scp /etc/Pegasus/server.pem QSECOFR@CTCIHA9V:/server_name.pem
In the above command, change QSECOFR to the IBM i profile that you want to use,
change CTCIHA9V to your IBM i IP name, and change server_name.pem to the file
name that you want to use for the certificate file, for example, my_hmc.pem.
6. Sign off the HMC.
7. On your IBM i system, start the PASE shell environment using call qp2term:
a. Move the HMC digital certificate:
mv /myhmc.pem /QOpenSys/QIBM/UserData/UME/Pegasus/ssl/truststore/myhmc.pem)
Replace the name, myhmc.pem, with your specific file name.
b. Add the digital certificate to the truststore:
/QOpenSys/QIBM/ProdData/UME/Pegasus/bin/cimtrust -a -U QSECOFR -f
/QOpenSys/QIBM/UserData/UME/Pegasus/ssl/truststore/myhmc.pem -T s
Replace the name, myhmc.pem, with your specific file name.
c. Exit the PASE shell by pressing F3.
8. Restart the CIM server to pick up the new certificate by doing an ENDTCPSVR *CIMOM and a
STRTCPSVR *CIMOM.

Chapter 11. Creating a PowerHA base environment 209


9. When all these requirements are met you can add a cluster monitor to each of your cluster
nodes. From the PowerHA GUI main menu, choose Cluster Nodes and then open the
properties information of your first cluster node (Figure 11-13).

Figure 11-13 Cluster Nodes: Show properties

210 PowerHA SystemMirror for IBM i Cookbook


As can be seen in Figure 11-14, there is currently no cluster monitor defined for node
CTCIHA9V.

Figure 11-14 Cluster monitor not defined

Chapter 11. Creating a PowerHA base environment 211


10.From the Select Action pull-down menu, choose Add Cluster Monitor (Figure 11-15).

Figure 11-15 Add Cluster Monitor

11.Provide the information shown in Figure 11-16. The CIM server host name is the IP name
of your HMC. The user ID and password are also those used to access the HMC.

Figure 11-16 Provide cluster monitor information

212 PowerHA SystemMirror for IBM i Cookbook


In Figure 11-17 you can see that the cluster monitor was successfully added for node
CTCIHA9V.

Figure 11-17 Cluster monitor added

Chapter 11. Creating a PowerHA base environment 213


12.Perform the same steps for all nodes in your cluster that should have a cluster monitor
defined. In our example, we set up cluster monitors for both nodes (Figure 11-18).

Figure 11-18 Cluster monitor added and active

The setup that we have done so far can also be done using the commands shown in
Example 11-1.

Example 11-1 Command to create a 2-node cluster with cluster monitors


CRTCLU CLUSTER(PWRHA_CLU) NODE((CTCIHA9V ('10.10.10.1'))) CLUMSGQ(QSYS/QSYSOPR) FLVWAITTIM(*NOMAX)
ADDCLUNODE CLUSTER(PWRHA_CLU) NODE(CTCIHA9W ('10.10.10.2'))
CHGCLUVER CLUSTER(PWRHA_CLU) HAVER(*UP1MOD)
ADDCLUMON CLUSTER(PWRHA_CLU) NODE(CTCIHA9V) CIMSVR(CTCHMC04.RCHLAND.IBM.COM hmc_user (hmc_password))
ADDCLUMON CLUSTER(PWRHA_CLU) NODE(CTCIHA9W) CIMSVR(CTCHMC04.RCHLAND.IBM.COM mhc_user (hmc_password))

214 PowerHA SystemMirror for IBM i Cookbook


11.3 Creating an IASP
From the page shown in Figure 11-19 you can proceed to create an independent ASP (IASP).

Figure 11-19 PowerHA GUI: Cluster overview

Chapter 11. Creating a PowerHA base environment 215


To proceed:
1. Figure 11-20 shows you a list of highly available IASPs. If you want to create a new IASP,
you first have to click Show all others.

Figure 11-20 PowerHA GUI: IASP overview

216 PowerHA SystemMirror for IBM i Cookbook


2. From the Select Action pull-down menu you can select Create Independent ASP
(Figure 11-21).

Figure 11-21 PowerHA GUI: Create independent ASP

Chapter 11. Creating a PowerHA base environment 217


3. A wizard guides you through the steps required to create an IASP. The Welcome page
(Figure 11-22) gives you an overview of the information that you need to provide.

Figure 11-22 PowerHA GUI: IASP Welcome panel

218 PowerHA SystemMirror for IBM i Cookbook


4. In the first step you have to decide which node in the cluster you want to create the IASP.
By default, this is the node on the system that you are connected to (Figure 11-23).

Figure 11-23 PowerHA GUI: Node selection

Chapter 11. Creating a PowerHA base environment 219


5. Provide a name for the IASP and decide whether it is a primary, secondary, or UDFS
IASP. You can also encrypt the IASP (Figure 11-24). The Protected field allows you to
specify whether you want to have the new IASP protected by either parity protection or
IBM i mirroring. Depending on your selection for this field, only corresponding disk units
with the required protection level will be shown. SVC/V7000 LUNs will always show up
as unprotected disk units regardless of whether they are RAID protected by the
storage system.

Figure 11-24 PowerHA GUI: IASP name and type

220 PowerHA SystemMirror for IBM i Cookbook


6. The wizard then shows you a list of unconfigured disks on your system. Notice that to
create an IASP with the new PowerHA GUI you do not have to provide an SST password
or have an SST user ID with the same name and password as your normal IBM i user ID
anymore. Choose the disks that you want to add into your IASP and click Add
(Figure 11-25).

Figure 11-25 PowerHA GUI: Select disks for IASP

Chapter 11. Creating a PowerHA base environment 221


7. The wizard updates the information shown with your choice (Figure 11-26). You can still
remove individual disks from your selection.

Figure 11-26 PowerHA GUI: Disks selected for IASP

222 PowerHA SystemMirror for IBM i Cookbook


8. The Summary page (Figure 11-27) provides an overview of the configuration settings that
you have chosen. Click Finish to create the IASP.

Figure 11-27 PowerHA GUI: IASP creation summary

9. The IASP is created. The wizard regularly updates the completion status (Figure 11-28).

Figure 11-28 PowerHA GUI: IASP is created

Chapter 11. Creating a PowerHA base environment 223


Alternatively, you can use a CL command to create your iASP. Figure 11-29 shows the
required parameters for CFGDEVASP.

Configure Device ASP (CFGDEVASP)

Type choices, press Enter.

ASP device . . . . . . . . . . . > IASP1 Name


Action . . . . . . . . . . . . . > *CREATE *CREATE, *DELETE
ASP type . . . . . . . . . . . . *PRIMARY *PRIMARY, *SECONDARY, *UDFS
Protection . . . . . . . . . . . *NO *NO, *YES
Encryption . . . . . . . . . . . *NO *NO, *YES
Disk units . . . . . . . . . . . *SELECT Name, *SELECT
+ for more values

Bottom
F1=Help F9=Calculate Selection F11=View 2 F12=Cancel

Figure 11-29 CFGDEVASP to create IASP

Specifying *SELECT for the disk unit parameter shows the panel shown in Figure 11-30. It
provides you with a list of unconfigured disks on your system. Choose which ones you want to
add to your IASP and press Enter to create the IASP.

Select Non-Configured Disk Units

ASP device . . . . . . . . . . . . . . . : IASP1


Selected capacity . . . . . . . . . . . : 0
Selected disk units . . . . . . . . . . : 0

Type options, press Enter.


1=Select

Resource
Opt Name Serial Number Type Model Capacity Rank Eligible
1 DD007 YQKJGD54BUK6 6B22 0050 19088 002 Yes
1 DD006 YDP4V2FVUK63 6B22 0050 19088 002 Yes
1 DD008 YUNHA7W9URJL 6B22 0050 19088 002 Yes
1 DD005 YWPZGH6N8LA9 6B22 0050 19088 002 Yes

Bottom
F1=Help F9=Calculate Selection F11=View 2 F12=Cancel
Configuration of ASP device IASP1 is 8% complete.
Figure 11-30 CFGDEVASP: Select disks to put into IASP

A message on the bottom of the page shows the progress of the IASP creation. Your age is
locked as long as the creation of the IASP is running.

224 PowerHA SystemMirror for IBM i Cookbook


11.4 Setting up an administrative domain
An administrative domain can be used to synchronize a number of object types that cannot
reside in an IASP between cluster nodes. The following steps can be used to set up an
administrative domain and add monitored resource entries into it:
1. Start from the main menu of the PowerHA GUI by choosing Cluster Administrative
Domain (Figure 11-31).

Figure 11-31 Cluster administrative domain setup

Chapter 11. Creating a PowerHA base environment 225


As you can see in Figure 11-32, we currently have not set up an administrative domain.

Figure 11-32 Cluster administrative domain not configured

2. To start the setup process choose Create Cluster Administrative Domain from the
pull-down menu (Figure 11-33).

Figure 11-33 Create Cluster Administrative Domain

226 PowerHA SystemMirror for IBM i Cookbook


3. Provide a name for the administrative domain and decide on the synchronization option.
This can be either last change or active domain. For a detailed description of these options
see “Monitored resources” on page 56. Add the nodes that you want to be part of the
administrative domain by choosing them from the list of available nodes and adding them
to the list of selected nodes (Figure 11-34).

Figure 11-34 Cluster Administrative Domain: Add nodes

4. After you have selected all nodes to be part of the administrative domain click OK
(Figure 11-35).

Figure 11-35 Cluster Administrative Domain: All nodes selected

Chapter 11. Creating a PowerHA base environment 227


The administrative domain is created but not yet started (Figure 11-36).

Figure 11-36 Cluster administrative domain created

5. Choose Start (Figure 11-37).

Figure 11-37 Start cluster administrative domain

228 PowerHA SystemMirror for IBM i Cookbook


The administrative domain becomes active (Figure 11-38). There are currently no
monitored resource entries defined.

Figure 11-38 Cluster administrative domain active

6. Choose Monitored Resources (Figure 11-39).

Figure 11-39 Cluster administrative domain: Check monitored resources

Chapter 11. Creating a PowerHA base environment 229


7. Choose Add Monitored Resources from the pull-down menu (Figure 11-40).

Figure 11-40 Cluster administrative domain: Add MRE

230 PowerHA SystemMirror for IBM i Cookbook


8. On the page shown in Figure 11-41 we add the subsystem QBATCH as a monitored
resource to the administrative domain. CTCIHA9V is the system that is used as a source
for the first synchronization. Be aware that, in general, there is no leading system in an
administrative domain. Changes on any node in the administrative domain are propagated
to all nodes in the domain. To add resources, you can either type their name or select
them from the corresponding list. You can also decide whether you want to synchronize all
attributes of a specific MRE or just subsets of attributes.

Figure 11-41 Cluster administrative domain: Add MRE for subsystem description

Chapter 11. Creating a PowerHA base environment 231


9. When using the list functions, you can also add several MREs of one object type into the
administrative domain with one configuration step (Figure 11-42).

Figure 11-42 Cluster administrative domain: Add MRE for user profile

10.The monitored resource entries are added to the administrative domain and a first
synchronization occurs (Figure 11-43). If the MREs do not exist on any of the other nodes
in the administrative domain, they are automatically created.

Figure 11-43 Cluster administrative domain: MRE successfully added

232 PowerHA SystemMirror for IBM i Cookbook


11.Figure 11-44 shows that the added MREs are in a consistent status within the
administrative domain.

Figure 11-44 Cluster administrative domain: MRE overview

The setup that we have done so far can also be done using the commands shown in
Example 11-2. Notice that you have to use individual commands for each monitored resource
that you want to add to the administrative domain.

Example 11-2 Command to create an administrative domain and add monitored resource entries
CRTCAD CLUSTER(PWRHA_CLU) ADMDMN(PWRHA_CAD) DMNNODL(CTCIHA9V CTCIHA9W)
SYNCOPT(*LASTCHG)

ADDCADMRE CLUSTER(PWRHA_CLU) ADMDMN(PWRHA_CAD) RESOURCE(QBATCH) RSCTYPE(*SBSD)


RSCLIB(QSYS)
ADDCADMRE CLUSTER(PWRHA_CLU) ADMDMN(PWRHA_CAD) RESOURCE(TESTSJ) RSCTYPE(*USRPRF)

Chapter 11. Creating a PowerHA base environment 233


234 PowerHA SystemMirror for IBM i Cookbook
12

Chapter 12. Configuring and managing


Geographic Mirroring
This chapter describes a scenario that uses the new PowerHA GUI for setting up and
managing your geographic mirroring high-availability environment.

A geographic mirroring solution requires a cluster with a minimum of two cluster nodes, an
IASP, and optionally cluster monitors and an administrative domain. These components
make up the base environment for a geographic mirroring solution. The steps are discussed
in Chapter 11, “Creating a PowerHA base environment” on page 199.

© Copyright IBM Corp. 2012. All rights reserved. 235


12.1 Setting up geographic mirroring
We set up geographic mirroring for an existing independent ASP that was created using the
steps in 11.3, “Creating an IASP” on page 215.

Geographic mirroring is set up using the Make Highly Available wizard, which is found on
PowerHA GUI → Independent ASPs → Show All Others → pop menu of the iASP → Make
Highly Available (Figure 12-1).

Figure 12-1 Make Highly Available

Follow the steps below to set up geographic mirroring:


1. When you choose the Make Highly Available option, the Welcome screen of the wizard
displays (Figure 12-2). This screen page you an overview of the steps involved in setting
up the environment. Click Next.

Figure 12-2 Make Highly Available Wizard’s Welcome panel

236 PowerHA SystemMirror for IBM i Cookbook


2. From the Choose Configuration page, click the Geographic Mirroring radio button
(Figure 12-3).

Figure 12-3 Choose Configuration

3. The IASP must be part of a device cluster resource group (CRG) for setting up geographic
mirroring. On the Configuration page, you can either create a new device CRG or use an
existing one. For our scenario, we choose to create a new device CRG named
PWRHA_CRG. Click Next to work with the recovery domain.

Chapter 12. Configuring and managing Geographic Mirroring 237


4. From the Recovery Domain page, you can add the cluster nodes and specify their roles
and the data port IP addresses (Figure 12-4). First we must select a node for the primary
role and assign a site name for identification. PowerHA allows you to choose up to four
data port IP addresses to ensure communication redundancy.

Figure 12-4 Add a primary node in the recovery domain

5. Add backup node 1 and specify a name for the site and up to four data port IP addresses
(Figure 12-5).

Figure 12-5 Add backup node to the recovery domain

238 PowerHA SystemMirror for IBM i Cookbook


6. The Recovery Domain page shows a summary of the node configuration options. From
this page, you can either click a node to modify the properties or click Next to continue
with the setup wizard (Figure 12-6).

Figure 12-6 Recovery Domain summary

7. If a device domain does not already exist, the wizard prompts you to and add the nodes to
a domain (Figure 12-7). Click OK to add them.

Figure 12-7 Add Nodes to Device Domain

Chapter 12. Configuring and managing Geographic Mirroring 239


The wizard shows progress pop-ups when adding the nodes to a device domain
(Figure 12-8).

Figure 12-8 Adding nodes to device domain progress pages

8. On the Devices panel, choose Modify from the IASP’s popup menu (Figure 12-9) to
specify a server takeover IP address and vary on options for the IASP.

Figure 12-9 Server takeover IP address

240 PowerHA SystemMirror for IBM i Cookbook


9. For our environment, we choose to automatically vary on the device on the backup node
and specify a server takeover IP address for the users to connect to, which will be
switched/failed over together with the device CRG. Click Save to go back to the Devices
page (Figure 12-10).

Figure 12-10 Modify IASP properties

10.You can specify an exit program to be executed after a switchover. This is optional, but if
specified, it must exist on all nodes in the recovery domain of the CRG. The exit program
can be used to set up an application environment and any other priming steps that are
necessary before users are allowed on the new production node after the switchover. For
our scenario, we leave the default Exit Program option as No and click Next
(Figure 12-11).

Figure 12-11 Specify Exit Program

Chapter 12. Configuring and managing Geographic Mirroring 241


11.Specify an optional failover message queue to prevent an automatic failover in case of any
unplanned node or server failures. For our environment, we use the QSYSOPR message
queue with an indefinite wait time parameter so that we can respond to the CPABB02
inquiry message to either proceed with or cancel the failover. We choose the default
message action as Cancel failover (Figure 12-12).

Figure 12-12 Specify Failover Message Queue

242 PowerHA SystemMirror for IBM i Cookbook


12.The summary page (Figure 12-13) lets you view the configuration options chosen up until
this point and gives you an opportunity to go back and modify any of the options. If you are
satisfied with the summary, click Finish to make the IASP highly available.

Figure 12-13 Summary page of configuration options

13.If there are no issues with the configuration, you will see a success dialog box
(Figure 12-14). PowerHA will add the nodes to the recovery domain, set up ASP copy
descriptions, and set up an ASP session for geographic mirroring.

Figure 12-14 Make Highly Available success dialog box

Chapter 12. Configuring and managing Geographic Mirroring 243


14.If the IASP was never set up for geographic mirroring, the wizard asks you whether you
want to configure mirroring (Figure 12-15). Choose Yes to start the Configure Mirroring
wizard in PowerHA GUI.

Figure 12-15 Prompt to continue setting up mirroring for the IASP

The Configure Mirroring wizard can also be started using the IASP context menu
(Figure 12-16).

Figure 12-16 Configure Mirroring for an IASP

244 PowerHA SystemMirror for IBM i Cookbook


The geographic mirroring for the IASP can also be managed by PowerHA if it was set up
elsewhere. When PowerHA detects that geographic mirroring was configured but that
there are no ASP copy descriptions or an ASP session that can be managed from
PowerHA GUI, it prompts you to create them (Figure 12-17).

Figure 12-17 Manage geographic mirroring with PowerHA GUI

15.The Configure Mirroring wizard (Figure 12-18) can be used to set up an ASP session,
ASP copy descriptions, and so on, to be used for geographic mirroring. On the Welcome
page, click Next to choose the mirroring configuration options.

Figure 12-18 Configure Mirroring wizard Welcome panel

Chapter 12. Configuring and managing Geographic Mirroring 245


16.On the Choose Configuration page, select Geographic Mirroring as the type and specify
a name for the ASP session (Figure 12-19).

Figure 12-19 Specify mirroring type and a name for the ASP mirroring session

246 PowerHA SystemMirror for IBM i Cookbook


17.The Recovery Domain page (Figure 12-20) gives you a summary and also lets you make
adjustments to the site names, data ports, and so on.

Figure 12-20 Recovery Domain summary

Chapter 12. Configuring and managing Geographic Mirroring 247


18.The Configure Mirroring wizard then takes you to the Mirroring Options panel, where you
can add copy descriptions and specify advanced mirroring options. Choose Modify from
the IASP context menu (Figure 12-21).

Figure 12-21 Context menu to modify IASP

19.Specify names for the ASP copy descriptions and click Save (Figure 12-22).

Figure 12-22 Add copy descriptions

248 PowerHA SystemMirror for IBM i Cookbook


20.On the Mirroring Options panel, click Show Advanced Options to change any of the
mirroring attribute defaults. This is where you can specify whether you want synchronous
or asynchronous delivery mode and synchronous or asynchronous data transmission.
The synchronous or asynchronous mirroring mode determines whether the writes on
the backup node are acknowledged back to the primary node after they are received in
main memory (synchronous) or after they are written to the disk (asynchronous),
respectively, on the mirror copy. Figure 12-23 shows the available advanced options for
the mirroring session.

Figure 12-23 Advanced options for geographic mirroring session

Chapter 12. Configuring and managing Geographic Mirroring 249


21.The Configure Mirroring wizard then takes you to Disk Unit configuration page for creating
the IASP mirror copy on the backup node. As shown in Figure 12-24, you can select the
disk units to create the IASP from a list of available unconfigured disk units on the backup
node. Click Add to add them to the selected Disk Units panel on the wizard.

Figure 12-24 Select disk units to create the IASP mirror copy

22.Verify the total disk capacity on the Selected Disk Units pane (Figure 12-25) and make
sure that you have adequate capacity to match the production copy. For more information
about sizing considerations refer to “Disk subsystem considerations” on page 146.

Figure 12-25 Selected Disk Units and their capacities

250 PowerHA SystemMirror for IBM i Cookbook


23.If the total capacity for the mirror copy does not match the capacity of the production copy,
PowerHA issues a warning (Figure 12-26). For our scenario, we choose Yes to continue
with the setup.

Figure 12-26 Capacity mismatch warning

24.The Configure Mirroring wizard shows you a summary page with all the details for the
mirroring session, recovery domain, ASP copy descriptions, and the disk unit selection for
the mirror copy (Figure 12-28 on page 252).

Figure 12-27 Configure Mirroring progress

Chapter 12. Configuring and managing Geographic Mirroring 251


Click Finish to complete the setup. You should see a configuration progress and
successful completion message (Figure 12-27 on page 251).

Figure 12-28 Configure Mirroring summary

Setting up geographic mirroring with CL commands


The setup shown above can also be done using CL commands, including the new CFGGEOMIR
command (Example 12-1).

Example 12-1 CL commands used for similar geographic mirroring setup


CRTDEVASP DEVD(IASP1) RSRCNAME(IASP1)
ADDDEVDMNE CLUSTER(PWRHA_CLU) DEVDMN(PWRHA_DMN) NODE(DEMOGEO1)
ADDDEVDMNE CLUSTER(PWRHA_CLU) DEVDMN(PWRHA_DMN) NODE(DEMOGEO2)
CRTCRG CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG) CRGTYPE(*DEV) EXITPGM(*NONE)
USRPRF(*NONE) RCYDMN(
(DEMOGEO1 *PRIMARY *LAST SITE1 ('192.170.86.11' '192.170.87.11'))
(DEMOGEO2 *BACKUP 1 SITE2 ('192.170.86.10' '192.170.87.10')))
CFGOBJ((IASP1 *DEVD *ONLINE '10.0.0.1'))
FLVMSGQ(QSYS/QSYSOPR) FLVWAITTIM(*NOMAX) FLVDFTACN(*CANCEL)
CFGGEOMIR ASPDEV(IASP2) ACTION(*CREATE)
SRCSITE(SITE1) TGTSITE(SITE2)
SSN(SSN1CPYD2/SSN1CPYD1/ASPGEOSSN1)
DELIVERY(*ASYNC) UNITS(DD021)
CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG) MODE(*ASYNC)
STRCRG CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG)

252 PowerHA SystemMirror for IBM i Cookbook


12.2 Managing geographic mirroring
The PowerHA GUI provides you with one-stop access to managing your entire HA
environment including managing these:
 The cluster and its properties
 Cluster nodes
 Independent ASPs
 Cluster administrative domains
 Cluster resource groups
 TCP/IP interfaces

In the following sections we show you how to manage various aspects of your high-availability
environment, including performing a planned switchover.

12.2.1 Administrative domain


In a high-availability environment it is necessary that the application and operational
environment remain consistent among the nodes that participate in high availability. The
cluster administrative domains interface on the PowerHA GUI allows you to maintain
environment resiliency and ensures that the operational environment remains consistent
across the nodes. The cluster administrative domain (Figure 12-29) provides the mechanism
that keeps resources synchronized across systems and partitions within a cluster.

Figure 12-29 Create Cluster Administrative Domain option

Chapter 12. Configuring and managing Geographic Mirroring 253


You can add all the nodes that would participate in a cluster administrative domain
(Figure 12-30).

Figure 12-30 Add nodes to cluster administrative domain

Monitored resources
A monitored resource is a system object or a set of attributes not associated with a specific
system object, such as the set of system environment variables. A resource represented by
an MRE is monitored for changes by a cluster administrative domain. When a monitored
resource is changed on one node in a cluster administrative domain, the change is
propagated to the other active nodes in the domain. You can access the Monitored
Resources panel from the Cluster Administrative Domain context menu (Figure 12-31).

Figure 12-31 Monitored Resources in a Cluster Administrative Domain

254 PowerHA SystemMirror for IBM i Cookbook


On the Monitored Resources interface, click the Select Action drop-down menu and choose
Add Monitored Resources (Figure 12-32).

Figure 12-32 Add Monitored Resources option

On the Add Monitored Resources interface (Figure 12-33), you can type the name of the
resource or select from a list. When adding an MRE for a system object, the resource name is
the name of the system object. One or more attributes can be specified, and only attributes
that are specified will be monitored for changes.

Figure 12-33 Add Monitored Resources

Chapter 12. Configuring and managing Geographic Mirroring 255


You can also access the attributes interface for a resource from the context menu of the
Monitored Resource page (Figure 12-34).

Figure 12-34 Access attributes option for an MRE

On the Attributes interface, you can see which attributes are consistent and which are
inconsistent. As shown in Figure 12-35, you might find it easier to click the Global Status
column to sort the list and identify the attributes that are in an inconsistent state.

Figure 12-35 Attribute details for a user profile MRE

256 PowerHA SystemMirror for IBM i Cookbook


CL commands for administrative domain and MREs
The setup shown above can also be done using the commands shown in Example 12-2.

Example 12-2 CL commands used for Cluster Administrative Domain


CRTCAD CLUSTER(PWRHA_CLU) ADMDMN(PWRHA_CAD) DMNNODL(DEMOGEO1)
ADDCADNODE CLUSTER(PWRHA_CLU) ADMDMN(PWRHA_CAD) NODE(DEMOGEO2)
ADDCADMRE CLUSTER(PWRHA_CLU) ADMDMN(PWRHA_CAD) RESOURCE(QBATCH)
RSCTYPE(*SBSD) RSCLIB(QSYS)
ADDCADMRE CLUSTER(PWRHA_CLU) ADMDMN(PWRHA_CAD) RESOURCE(POWERHA)
RSCTYPE(*USRPRF) ATTRIBUTE(*ALL)

12.2.2 Planned switchover


Before attempting a planned switchover, it is very important that all the components in the HA
environment are functional and in a green status. You can verify this from the main PowerHA
GUI landing page that shows you the status of these:
 Cluster nodes: All cluster nodes should be active.
 Independent ASPs: All independent ASPs are available.
 Cluster administrative domains: Active and all monitored resources are consistent.
 Cluster resource groups: All CRGs are active.
 TCP/IP interfaces: All interfaces are active or in an expected status.

Figure 12-36 shows a fully functional HA environment.

Figure 12-36 PowerHA main panel with status of all components

Chapter 12. Configuring and managing Geographic Mirroring 257


To perform a switchover, choose Switchover from the context menu of the cluster resource
group (Figure 12-37).

Figure 12-37 Switchover option for a cluster resource group

When you select the Switchover option, PowerHA gives you a preview of how the nodes and
their roles in a cluster resource group will look before and after a switchover (Figure 12-38).

Figure 12-38 Switchover summary with current and new node order

258 PowerHA SystemMirror for IBM i Cookbook


On the recovery domain preview screen, verify that the roles of various nodes are as
expected and click OK to proceed with the switchover.

PowerHA shows you various progress messages and gives you a success dialog if
everything went well (Figure 12-39).

Figure 12-39 Switchover progress and completion messages

After you click Close on the Switch completed successfully dialog, you will be back on the
cluster resource groups interface. Click Refresh to view the updated role status for all the
nodes in the recovery domain (Figure 12-40).

Figure 12-40 CRG status after node role reversal

Chapter 12. Configuring and managing Geographic Mirroring 259


To verify the status of geographic mirroring and the current direction of replication, go to the
Independent ASP details interface (Figure 12-41).

Figure 12-41 IASP details after switchover

12.2.3 Deconfiguring geographic mirroring


If you no longer want the capability to use geographic mirroring for a specific disk pool or disk
pool group, you can select Deconfigure Geographic Mirroring. If you deconfigure
geographic mirroring, the system stops geographic mirroring and deletes the mirror copy of
the disk pools on the nodes in the mirror copy site.

To deconfigure geographic mirroring, the disk pool must be offline. First vary off the
independent ASP and then select Deconfigure Geographic Mirroring from the IASP pop-up
menu on the PowerHA GUI (Figure 12-42).

Figure 12-42 Deconfigure Geographic Mirroring menu option

260 PowerHA SystemMirror for IBM i Cookbook


Click OK on the confirmation dialog to continue deconfiguring geographic mirroring.

Figure 12-43 Confirmation dialog to deconfigure geographic mirroring

Chapter 12. Configuring and managing Geographic Mirroring 261


262 PowerHA SystemMirror for IBM i Cookbook
13

Chapter 13. Configuring and managing


DS8000 Copy Services
In this chapter we describe the scenario that uses DS8000 Copy Services with IBM PowerHA
SystemMirror for i.

In various sections of this chapter, we discuss the following activities using an IBM i DS8000:
 Setting up a Copy Services environment
 Configuring Metro Mirror
 Configuring FlashCopy
 Configuring Global Mirror
 Configuring LUN-level switching
 Managing IBM i DS8000 Copy Services

© Copyright IBM Corp. 2012. All rights reserved. 263


13.1 Setting up IBM i DS8000 Copy Services
PowerHA SystemMirror for i controls IBM i clustering functions such as switchover and
failover. When used with DS8000 Copy Services, it also needs to be able to control the Copy
Services functions like PPRC failover/failback or FlashCopy. This requires the installation of
the DS command-line interface (DS CLI) on the IBM i partitions. There is no need for a user to
take care of the way the DSCLI works on an IBM i partition because it is used by PowerHA
“under-the-cover”.

The DS CLI installation package can be downloaded from the following URL:
ftp://ftp.software.ibm.com/storage/ds8000/updates/DS8K_Customer_Download_Files/CLI

The install packages are provided as ISO image files, which can be used by any virtual CD
drive tool on Windows.

For DS CLI installation procedures, refer to the IBM System Storage DS Command-Line
Interface User's Guide for the DS6000 series and DS8000 series, GC53-1127, included in the
install package and section 8.3 in IBM i and IBM System Storage: A Guide to Implementing
External Disks on IBM i, SG24-7120.

For the IBM i partitions to be able to manage DS8000 Copy Services, we need to provide a
DS storage user profile and its password. Whenever the password is changed on the
DS8000, the configuration on the IBM i partition needs to be changed at the same time. For
more information refer to section 9.5, “User management,” in IBM System Storage DS8000:
Architecture and Implementation, SG24-8886.

Note: User profiles and passwords are case sensitive on the DS8000.

The communication via DS CLI requires an IP connection between the IBM i partition and the
DS8000. This connection is initiated by IBM i to the DS8000, which listens on TCP port 1750.

Note: If there is a firewall between the IBM i partition and the DS8000, TCP port 1750 must
be authorized.

13.1.1 Configuring IBM i DS8000 Metro Mirror (GUI and CL commands)


This section covers a scenario using a DS8000 Metro Mirror environment as the hardware
replication technology.

Environment overview
For this scenario, we need to create several cluster items. Cluster, cluster monitor, IASP, and
administrative domain setup are covered in Chapter 11, “Creating a PowerHA base
environment” on page 199. They are common to all scenarios that we describe in this chapter.

These are cluster items that we specifically configure for our Metro Mirror environment:
 Cluster resource group
 Device domain
 ASPcopy descriptions and session

264 PowerHA SystemMirror for IBM i Cookbook


Table 13-1 and Figure 13-1 on page 266 show the setup that we use for our environment.

Table 13-1 Metro Mirror scenario settings


Preferred production Preferred backup

System name DEMOPROD DEMOHA

Cluster name PWRHA_CLU

Cluster resource group PWRHA_CRG1

Device domain PWRHA_DMN

Administrative domain PWRHA_CAD

IASP name, number IASP1, 33 IASP1, 33

Cluster resource group site name SITE1 SITE2

ASP copy description IASP1_MM1 IASP1_MM2

ASP session IASP1_MM

Takeover IP 10.0.0.1

Heartbeat Cluster IP 10.10.10.1 10.10.10.2

Management access IP 9.5.168.129 9.5.168.130

DS8000 device ID IBM.2107-75AY031 IBM.2107-75AY032

DS8000 IP addressa 9.5.168.55 9.5.168.55

Volume group ID V2 V2

Volume IDsb 6000-6002, 6100-6102 6000-6002, 6100-6102

FC IO adapter resource namec DC03 DC07

Host connection ID 3 1

Host connection WWPN 10000000C948031F 10000000C9523F21


a. IP addresses are identical because, in our environment, we use a single DS8000 configured
with two distinct storage images.
b. Volume IDs are not needed to be the same on the source and target Metro Mirror relationships.
c. FC stands for Fibre Channel.

Chapter 13. Configuring and managing DS8000 Copy Services 265


User access network
Heartbeat network

10.10.10.1

10.10.10.2
IBM Power
Systems server

DEMOPROD DEMOHA
IBM Power IBM i partition IBM i partition
Systems server

Host conn. Id 1 for 10000000C9523F21 on DC07


Host conn. id 3 for 10000000C948031F on DC03

9.5.168.130
Production Site 9.5.168.129 Backup Site

Management network
9.5.168.55

9.5.168.55

IBM.2107-75AY031 IBM.2107-75AY032

IBM System Storage IBM System Storage

Metro Mirror pair


60
00

60
01

60
02

60
00

60
01

60
02

Metro Mirror pair


61
00

61
01

61
02

61
00

61
01

61
02

Volumes group V2 Volumes group V2

A volumes B volumes

Figure 13-1 Metro Mirror scenario settings

Setting up a Metro Mirror environment


The Metro Mirror environment setup itself is not handled by IBM PowerHA SystemMirror for i.
You have to configure the logical storage configuration, including the Copy Services setup,
directly on the DS8000. Refer to 6.2, “Metro Mirror” on page 93, for more information.

266 PowerHA SystemMirror for IBM i Cookbook


Creating the PPRC paths
To find the possible links for a path, you can use the DS CLI command lsavailpprcport. In our
scenario, the source:target LSSs are 60:60 and 61:61. The target WWNN is 5005076307FFCAB8.
The I0040 port on the source DS8000 and the I0240 port on the target DS8000 are available. As
you can see in Example 13-1, we can create the PPRC paths with mkpprcpath.

Example 13-1 Creating PPRC path


dscli> lsavailpprcport -l -dev IBM.2107-75AY031 -remotedev IBM.2107-75AY032 -remotewwnn
5005076307FFCAB8 60:60
Date/Time: September 22, 2011 4:27:57 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
Local Port Attached Port Type Switch ID Switch Port
===================================================
I0040 I0240 FCP NA NA
dscli> lsavailpprcport -l -dev IBM.2107-75AY031 -remotedev IBM.2107-75AY032 -remotewwnn
5005076307FFCAB8 61:61
Date/Time: September 22, 2011 4:28:02 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
Local Port Attached Port Type Switch ID Switch Port
===================================================
I0040 I0240 FCP NA NA
dscli> mkpprcpath -dev IBM.2107-75AY031 -remotedev IBM.2107-75AY032 -remotewwnn 5005076307FFCAB8
-srclss 60 -tgtlss 60 I0040:I0240
Date/Time: September 22, 2011 4:37:44 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
CMUC00149I mkpprcpath: Remote Mirror and Copy path 60:60 successfully established.
dscli> mkpprcpath -dev IBM.2107-75AY031 -remotedev IBM.2107-75AY032 -remotewwnn 5005076307FFCAB8
-srclss 61 -tgtlss 61 I0040:I0240
Date/Time: September 22, 2011 4:37:54 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
CMUC00149I mkpprcpath: Remote Mirror and Copy path 61:61 successfully established.
dscli> lspprcpath -l 60-61
Date/Time: September 22, 2011 4:38:51 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
Src Tgt State SS Port Attached Port Tgt WWNN Failed Reason PPRC CG
================================================================================
60 60 Success FF60 I0040 I0240 5005076307FFCAB8 - Disabled
61 61 Success FF61 I0040 I0240 5005076307FFCAB8 - Disabled

Note: Make sure that the paths are established as bi-directional, that is, on both DS8000s
from one to the other for the volumes’ LSS.

Creating the Metro Mirror relationships


After the path exists from production DS8000 to backup DS8000, we can establish the Metro
Mirror relationships using mkpprc (Example 13-2). In our Metro Mirror environment, Metro
Mirror relationships are to be established between primary volume IDs 6000 - 6002 and
secondary volume IDs 6000 - 6002 for a first set, and primary volume IDs 6100 - 6102 and
secondary volume IDs 6100 - 6102 for a second set. The relationship is properly established
when the status is full duplex.

Example 13-2 Creating Metro Mirror relationships


dscli> mkpprc -dev IBM.2107-75AY031 -remotedev IBM.2107-75AY032 -type mmir -mode full 6000-6002:6000-6002
6100-6102:6100-6102
Date/Time: September 22, 2011 4:46:48 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
CMUC00153I mkpprc: Remote Mirror and Copy volume pair relationship 6000:6000 successfully created.
CMUC00153I mkpprc: Remote Mirror and Copy volume pair relationship 6001:6001 successfully created.
CMUC00153I mkpprc: Remote Mirror and Copy volume pair relationship 6002:6002 successfully created.
CMUC00153I mkpprc: Remote Mirror and Copy volume pair relationship 6100:6100 successfully created.
CMUC00153I mkpprc: Remote Mirror and Copy volume pair relationship 6101:6101 successfully created.

Chapter 13. Configuring and managing DS8000 Copy Services 267


CMUC00153I mkpprc: Remote Mirror and Copy volume pair relationship 6102:6102 successfully created.
dscli> lspprc 6000-61FF
Date/Time: September 22, 2011 4:47:12 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
===================================================================================================
6000:6000 Copy Pending - Metro Mirror 60 60 Disabled Invalid
6001:6001 Copy Pending - Metro Mirror 60 60 Disabled Invalid
6002:6002 Copy Pending - Metro Mirror 60 60 Disabled Invalid
6100:6100 Copy Pending - Metro Mirror 61 60 Disabled Invalid
6101:6101 Copy Pending - Metro Mirror 61 60 Disabled Invalid
6102:6102 Copy Pending - Metro Mirror 61 60 Disabled Invalid
dscli> lspprc 6000-61FF
Date/Time: September 22, 2011 4:50:37 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
==================================================================================================
6000:6000 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6001:6001 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6002:6002 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6100:6100 Full Duplex - Metro Mirror 61 60 Disabled Invalid
6101:6101 Full Duplex - Metro Mirror 61 60 Disabled Invalid
6102:6102 Full Duplex - Metro Mirror 61 60 Disabled Invalid

Note: For DS8000 releases prior to R4.3, to avoid a possible conflict with Volume Flash Init
DS8000 capabilities for volumes in a remote Copy Services relationship, the formatting of
volumes by DS8000 and IBM usually done when being added into an ASP must be
handled as follows:
1. Initialize and format disks using System Service Tools (STRSST):
a. Option 3. Work with disk units.
b. Option 3. Work with disk unit recovery.
c. Option 2. Disk unit recovery procedures.
d. Option 1. Initialize and format disk unit option. Notice that this operation is only a
marker, which disappears if there is an IPL.
2. Establish the Metro Mirror pair and wait for the source’s volumes to become synchronized.
3. Add volumes to the ASP or create the IASP to which volumes are added.

Creating all configuration items with the GUI


The device cluster resource group, device domain, and ASP copy descriptions can be created
with the PowerHA GUI using a specific wizard.

268 PowerHA SystemMirror for IBM i Cookbook


As for other PowerHA objects, the creation process starts from the main PowerHA panel:
1. Click Independent ASPs (Figure 13-2).

Figure 13-2 Starting Independent ASPs wizard

Chapter 13. Configuring and managing DS8000 Copy Services 269


2. In our case, there is no highly available independent ASP. Click Show all others to
retrieve all existing IASPs (Figure 13-3).

Figure 13-3 Starting Independent ASPs wizard

270 PowerHA SystemMirror for IBM i Cookbook


3. Select Make Highly Available for the IASP that you want to work with (Figure 13-4), in our
case IASP1.

Figure 13-4 Launch the Make Highly Available wizard

Chapter 13. Configuring and managing DS8000 Copy Services 271


4. Click Next (Figure 13-5) to start the Make Highly Available wizard.

Figure 13-5 Start the Make Highly Available wizard

272 PowerHA SystemMirror for IBM i Cookbook


5. The first step requires providing the type of mirroring to configure and the name of the
cluster resource group to be used or created. As shown in Figure 13-6, in our case, we
select Metro Mirror and provide the CRG name PWRHA_CRG1.

Figure 13-6 Specify type of mirroring and name of cluster resource group

Chapter 13. Configuring and managing DS8000 Copy Services 273


6. As shown in Figure 13-7, the next step is to add the primary node into the recovery
domain. In our case, DEMOPROD is this first node. Select the appropriate node in the list
and provide a site name, SITE1 in our case, and click Add.

Figure 13-7 Add the primary node to the cluster resource group

7. We continue by adding all necessary nodes to the cluster resource group. In our case
DEMOHA, which our backup node one, is located at the backup site with a site name of
SITE2 (Figure 13-8).

Figure 13-8 Add the second node to cluster resource group

274 PowerHA SystemMirror for IBM i Cookbook


8. After all nodes exist in the cluster resource group (Figure 13-9), click Next to continue.

Figure 13-9 Continue when all nodes are in the cluster resource group

9. All nodes in a cluster resource group whose type is device must also be members of a
device domain. There is no starting point for creating a device domain in the PowerHA
GUI. Device domains are created by giving them a name (Figure 13-10). If there is already
an appropriate existing device domain, we can select it from the drop-down list. In our
environment, no device domain exists. Giving it a name and then clicking OK creates it.

Figure 13-10 Create or select a device domain

Chapter 13. Configuring and managing DS8000 Copy Services 275


10.As shown in Figure 13-11, each node is added to the device domain. Independent ASP
and ASP device description creation occur at this time for all newly added nodes but the
first one.

Figure 13-11 Nodes being added to device domain

11.When all nodes are included (Figure 13-12), we can click Close to continue.

Note: If PowerHA does not succeed to add all nodes into the device domain, you
receive a Cancel/Retry panel. This provides you with the opportunity to analyze the
reason for the failure, solve the cause, and retry adding nodes into the device domain.

Note: At this time, the cluster resource group is not yet created, but the device domain
has been created and nodes have been included.

Figure 13-12 Nodes successfully added to device domain

276 PowerHA SystemMirror for IBM i Cookbook


12.After each node is a member of the same device domain (Figure 13-13), we can continue
by clicking Next.

Figure 13-13 All nodes are included in both a device domain and the cluster resource group

13.The next step is to launch the device configuration wizard. As shown in Figure 13-14,
select Modify for the device that you are working with (in our environment, IASP1).

Figure 13-14 Launch the device configuration

Chapter 13. Configuring and managing DS8000 Copy Services 277


14.As shown in Figure 13-15, you can select the automatic vary-on during a switchover and
specify the server takeover IP address (the one that the users connect to). When done,
click Save.

Figure 13-15 IASP settings

15.After the device is configured (Figure 13-16), click Next to continue.

Figure 13-16 Continue when the device is configured

278 PowerHA SystemMirror for IBM i Cookbook


16.The next panel is related to providing the name of an exit program that might handle
cluster resource group events like failover. In our case, there is no such program. We skip
this step by clicking Next (Figure 13-17).

Figure 13-17 Specify an optional exit program

Chapter 13. Configuring and managing DS8000 Copy Services 279


17. The last wizard panel is related to providing information for a failover message queue. As
shown in Figure 13-18, for our environment we do not provide a CRG failover message queue,
and just have to click Next to go to the summary wizard panel. If no CRG message queue is
specified here, the cluster message queue is used for all CRGs in the device domain.

Figure 13-18 Specify an optional failover message queue

280 PowerHA SystemMirror for IBM i Cookbook


18.The last panel is a summary allowing you to review the provided information and go back
to the related panel if something is wrong. Click Finish to create the CRG (Figure 13-19).

Figure 13-19 Finish the operation

The wizard displays several panels with the progress indicator (Figure 13-20).

Figure 13-20 Cluster resource group is being created

Chapter 13. Configuring and managing DS8000 Copy Services 281


19.At the end of this step, the cluster resource group has been created and the device
domain is configured. Clicking Close (Figure 13-21) ends this part of the wizard.

Figure 13-21 Cluster resource group is created

20.The next panel asks whether you want to continue the wizard to create the ASP copy
descriptions for both independent ASP copies included in the cluster resource group. Click
Yes to launch the Configure Mirroring wizard (Figure 13-22).

Figure 13-22 Question for configuring mirroring

282 PowerHA SystemMirror for IBM i Cookbook


Note: If you do not click Yes at this time, you can still create the ASP copy descriptions
later through the PowerHA GUI by using the Configure mirroring option on the
Independent ASP details panel (Figure 13-23 on page 283). You will then use the same
wizard described in step 21 on page 284.

Figure 13-23 Select Configure Mirroring from IASP Details panel

Chapter 13. Configuring and managing DS8000 Copy Services 283


21.On the Configure Mirroring Welcome panel, click Next to start the wizard.

Figure 13-24 Starting Configure Mirroring wizard

284 PowerHA SystemMirror for IBM i Cookbook


22.The first panel is related to choosing the mirroring configuration. As we come from a
previous wizard where we decided to use Metro Mirror, no other selection is possible. As
shown in Figure 13-25, we provide the name of the new ASP session, IASP1_MM in our
case, then we click Next to continue.

Figure 13-25 Choose configuration

Chapter 13. Configuring and managing DS8000 Copy Services 285


23.The recovery domain information is already populated with values from the cluster
resource group. Click Next (Figure 13-26).

Figure 13-26 Recovery domain reminder

24.The next step is related to creating ASP copy descriptions for both the source copy and
the target copy. As shown in Figure 13-27, select Modify Source Copy for the IASP (in
our case, this is still IASP1).

Figure 13-27 Beginning of copy descriptions creation

286 PowerHA SystemMirror for IBM i Cookbook


25.On the panel shown in Figure 13-28, you will, at the same time, either select or provide the
copy description name (in our case we provide IASP1_MM1), provide the DS8000 device
ID (IBM.2107-75AY031), the user profile (prowerha), the password, the IP address
(9.5.168.55), and the first range of volumes (6000 - 6002). Click Add to validate the panel.

Note: For the DS8000 user ID and password, make sure that your browser settings do
not override the information that you provide with one that could already exist in the
saved passwords for Mozilla Firefox or AutoComplete settings for Microsoft Internet
Explorer, for example.

Figure 13-28 Specify first logical unit range for source copy description

26.You will need to add the logical unit ranges as needed. For each new range (in our case
6100-6102), click Add. When a range is not possible for a single volume, you still have to
enter it as a range, for example, 6000-6000.

Chapter 13. Configuring and managing DS8000 Copy Services 287


27.When all logical unit ranges are OK, click Save to validate the source copy description
(Figure 13-29).

Figure 13-29 Validate logical unit ranges for source copy description

288 PowerHA SystemMirror for IBM i Cookbook


28. After the source copy description, we have to handle the target copy description creation. The
steps are the same as for the source copy description. Only the information differs:
a. As shown in Figure 13-30, we select Modify Target Copy against our IASP.
b. We provide the required information for the target copy description (Figure 13-31 on
page 290). In our case, the target copy description name is IASP1_MM2. It applies to
DS8000 device ID IBM.2107-75AY032. We connect to this DS8000 with the user
profile powerHA at the IP address 9.5.168.55. The first logical unit range is 6000 - 6002
and the second one is 6100 - 6102.
c. When all logical volume ranges are OK, click Save to continue (Figure 13-32 on
page 291).

Figure 13-30 Continue with target copy description creation

Chapter 13. Configuring and managing DS8000 Copy Services 289


Figure 13-31 Specify first logical unit range for target copy description

290 PowerHA SystemMirror for IBM i Cookbook


Figure 13-32 Validate logical unit ranges for target copy description

29.When both source and target copy descriptions are complete (Figure 13-33), click Next to
review the configuration.

Figure 13-33 Copy descriptions are complete

Chapter 13. Configuring and managing DS8000 Copy Services 291


30.When you reviewed your intended configuration, click Finish to proceed with the creation
of the ASP copy descriptions (in our case IASP1_MM1 and IASP1_MM2) and the starting
of the ASP session for Metro Mirror (in our case IASP1_MM) (Figure 13-34).

Figure 13-34 Submit copy descriptions creation

31.The wizard displays several panels with progress indicator (Figure 13-35).

Figure 13-35 Copy descriptions are being created and session started

292 PowerHA SystemMirror for IBM i Cookbook


32.The cluster resource group, device domain, and copy descriptions have been created and
the CRG is ready to be started when needed. Click Close to go back to the wizard starting
point (Figure 13-36). Independent ASP is now highly available and ready to be varied on
(Figure 13-37).

Figure 13-36 Mirroring configuration complete

Figure 13-37 IASP is highly available ready

Using CL commands
The corresponding CL commands for creating a device domain and a cluster resource group
are ADDDEVDMNE and CRTCRG.

The device domain must be properly populated with the nodes participating in the cluster
resource group before we can create it.

Chapter 13. Configuring and managing DS8000 Copy Services 293


As shown in Example 13-3, one command is necessary for each node to add it into the device
domain. For our environment DEMOPROD owns the IASP and therefore must be the first
node added into the device domain, and DEMOHA is the second node. Therefore, for any
existing Independent ASP on DEMOPROD, the corresponding device descriptions need to be
manually created on DEMOHA using CRTDEVASP. This is in contrast to the PowerHA GUI,
which creates the ASP device descriptions on the other nodes in the device
domain automatically.

Example 13-3 Adding nodes to a device domain and creating ASP device description
ADDDEVDMNE CLUSTER(PWRHA_CLU) DEVDMN(PWRHA_DMN) NODE(DEMOPROD)
ADDDEVDMNE CLUSTER(PWRHA_CLU) DEVDMN(PWRHA_DMN) NODE(DEMOHA)

CRTDEVASP DEVD(IASP1) RSRCNAME(IASP1) RDB(IASP1)

After the appropriate device domain and all IASP devices descriptions have been created, we
can create the cluster resource group (Example 13-4).

Example 13-4 Creating a cluster resource group


CRTCRG CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG1) CRGTYPE(*DEV)
EXITPGM(*NONE) USRPRF(*NONE)
RCYDMN((DEMOPROD *PRIMARY *LAST SITE1) (DEMOHA *BACKUP 1 SITE2))
CFGOBJ((IASP1 *DEVD *ONLINE '10.0.0.1'))
TEXT('DS8000 Cluster Resource Group')

The related command for adding a Metro Mirror ASP copy description is ADDASPCPYD
(Example 13-5). You have to create one for each IASP copy (that is, one for the Metro Mirror
source and one for the target site).

Example 13-5 Adding Metro Mirror ASP copy descriptions


ADDASPCPYD ASPCPY(IASP1_MM1) ASPDEV(IASP1) CRG(PWRHA_CRG1) LOCATION(*DEFAULT)
SITE(SITE1) STGHOST('powerha' ('password') ('9.5.168.55'))
LUN('IBM.2107-75AY031' ('6000-6002' '6100-6102') ())
ADDASPCPYD ASPCPY(IASP1_MM2) ASPDEV(IASP1) CRG(PWRHA_CRG1) LOCATION(*DEFAULT)
SITE(SITE2) STGHOST('powerha' ('password') ('9.5.168.55'))
LUN('IBM.2107-75AY032' ('6000-6002' '6100-6102') ())

Starting ASP session and cluster resource group using CL commands


When creating the ASP copy descriptions through the GUI, at the end of the procedure, the
session is automatically started. However, if you need to start it again, STRASPSSN can be
used. Also, STRCRG can be used to start the cluster resource group when needed
(Example 13-6).

Example 13-6 Starting the ASP session and cluster resource group
STRASPSSN SSN(IASP1_MM) TYPE(*METROMIR) ASPCPY((IASP1_MM1 IASP1_MM2))
STRCRG CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG1)

13.1.2 Configuring IBM i DS8000 FlashCopy (CL commands)


This section discusses a scenario using the DS8000 Copy Services FlashCopy function for
creating a point-in-time copy of an IASP and making it available for another IBM i partition for
backup purposes.

294 PowerHA SystemMirror for IBM i Cookbook


Environment overview
For this scenario, we need to create several cluster items. Cluster, cluster monitors, IASP, and
administrative domain setup are covered in Chapter 11, “Creating a PowerHA base
environment” on page 199. They are common to all scenarios that we describe in this chapter.

These are cluster items that we specifically configure for FlashCopy:


 Device domain
 FlashCopy ASP copy descriptions and sessions

Table 13-2 and Figure 13-38 on page 296 show the setup that we are going to use for our
FlashCopy environment.

Table 13-2 Settings for FlashCopy scenario


Source IASP Target IASP

System name DEMOHA DEMOFC

Cluster name PWRHA_CLU

Device domain PWRHA_DMN

IASP name, number IASP2,34 IASP2,34

Management access IP 9.5.168.129 9.5.168.130

DS8000 device ID IBM.2107-75AY032 IBM.2107-75AY032

DS8000 IP address 9.5.168.55 9.5.168.55

Volume group ID V26 V48

Volumes IDsa A010-A013 A020-A021


b
FC IO adapter resource name DC05 DC04

Host connection ID 5 2

Host connection WWPN 10000000C94122A2 10000000C9523E9D


a. Volumes need to be the same size on both sides.
b. FC stands for Fibre Channel.

Chapter 13. Configuring and managing DS8000 Copy Services 295


IBM System Storage
Management IP: 9.5.168.55
IBM Power Systems Device ID: IBM.2107-75AY032

VolGrp: V26

WWPN 10000000C94122A2
DEMOHA A010 A012
A011 A013

FlashCopy
WWPN 10000000C9523E9D
DEMOFC A020
A022
A021
A023
VolGrp: V48

Figure 13-38 The schematics of our environment

Volume groups in the DS8000


We assume that we already have the IASP2 configured on the system DEMOHA, and we
have the volumes for the target copy prepared. We list them using the DS CLI showvolgrp
command (Figure 13-39).

dscli> showvolgrp v26


Date/Time: September 28, 2011 10:13:25 PM CEST IBM DSCLI Version: 7.6.10.530
DS: IBM.2107-75AY032
Name V48
ID V26
Type OS400 Mask
Vols A010 A011 A012 A013
dscli> showvolgrp v48
Date/Time: September 28, 2011 10:13:30 PM CEST IBM DSCLI Version: 7.6.10.530
DS: IBM.2107-75AY032
Name V48fc
ID V48
Type OS400 Mask
Vols A020 A021 A022 A023
Figure 13-39 Volumes for source and target FlashCopy

IASP device description and ASP copy descriptions


On the DEMOHA system there is IASP2 already configured. To be able to use the copy of it
on the backup DEMOHA system, we need to create an IASP2 device description on it using
CRTDEVASP, as follows:
CRTDEVASP DEVD(IASP2) RSRCNAME(IASP2)

296 PowerHA SystemMirror for IBM i Cookbook


Then we need to create the ASP copy descriptions, which describe both copies of the IASP2
(that is, the FlashCopy source and target volumes) by using ADDASPCPYD on either one of the
two systems. Figure 13-40 and Figure 13-41 on page 298 are examples of our commands.

Add ASP Copy Description (ADDASPCPYD)

Type choices, press Enter.

ASP copy . . . . . . . . . . . . > ASPCPYD1 Name


ASP device . . . . . . . . . . . > IASP2 Name
Cluster resource group . . . . . *NONE Name, *NONE
Cluster resource group site . . *NONE Name, *NONE
Storage host:
User name . . . . . . . . . . > USER_ID
Password . . . . . . . . . . . > PASSWORD
Internet address . . . . . . . > 9.5.168.55

Location . . . . . . . . . . . . > DEMOHA Name, *DEFAULT, *NONE

Logical unit name:


TotalStorage device . . . . . IBM.2107-75AY032
Logical unit range . . . . . . A010-A013 Character value
+ for more values
Consistency group range . . . Character value
+ for more values
Recovery domain:
Cluster node . . . . . . . . . *NONE Character value, *NONE
Host identifier . . . . . . . Character value
+ for more values
Volume group . . . . . . . . . Character value
+ for more values
+ for more values

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 13-40 ADDASPCPYD for DEMOHA system

Chapter 13. Configuring and managing DS8000 Copy Services 297


Add ASP Copy Description (ADDASPCPYD)

Type choices, press Enter.

ASP copy . . . . . . . . . . . . > ASPCPYD2 Name


ASP device . . . . . . . . . . . > IASP2 Name
Cluster resource group . . . . . *NONE Name, *NONE
Cluster resource group site . . *NONE Name, *NONE
Storage host:
User name . . . . . . . . . . > USER_ID
Password . . . . . . . . . . . > PASSWORD
Internet address . . . . . . . > '9.5.168.55'

Location . . . . . . . . . . . . > DEMOFC Name, *DEFAULT, *NONE

Logical unit name:


TotalStorage device . . . . . > 'IBM.2107-75AY032'
Logical unit range . . . . . . > 'A020-A023' Character value
+ for more values
Consistency group range . . . Character value
+ for more values
Recovery domain:
Cluster node . . . . . . . . . *NONE Character value, *NONE
Host identifier . . . . . . . Character value
+ for more values
Volume group . . . . . . . . . Character value
+ for more values
+ for more values

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys

Figure 13-41 ADDASPCPYD for DEMOFC system

Here we describe parameters that we used in the ASP copy descriptions:


 Username and password
This specifies the DS user ID and password. In the case of the ASP copy description used
for FlashCopy, all source and all target volumes need to be on the same TotalStorage
device. Therefore, the user and password for both ASP copy descriptions will be the same.
 Internet address
This is the IP address for managing the TotalStorage system.
 Location
This parameter specifies the IBM i cluster node where the given copy of the IASP can be
used. In our case the source for the FlashCopy is being used by DEMOHA and the target
copy of the IASP will be used by DEMOFC.

298 PowerHA SystemMirror for IBM i Cookbook


Doing the FlashCopy of an IASP
After completing the configuration steps described in previous sections, we are ready to start
the FlashCopy operation for IASP2 on the DEMOHA system:
1. Vary off the IASP or quiesce the database activity for the IASP on the production system.
A quiesce brings the database to a consistent state by flushing modified data from main
memory to disk and suspending database operations and transactions. The vary-on of the
FlashCopied IASP will still be abnormal, but a lengthy database recovery is avoided,
which makes the vary-on process shorter. This operation must be run on the production
system. We run:
CHGASPACT ASPDEV(IASP1) OPTION(*SUSPEND) SSPTIMO(60)
2. Start the ASP FlashCopy session.
In this step we establish a point-in-time copy of the IASP with the FlashCopy operation.
On the backup system we run the following command:
STRASPSSN SSN(ASPSSN2) TYPE(*FLASHCOPY) ASPCPY((ASPCPYD1 ASPCPYD2))
Figure 13-42 shows an example of the command starting the FlashCopy process in
our environment.

Start ASP Session (STRASPSSN)

Type choices, press Enter.

Session . . . . . . . . . . . . > IASPSSN Name


Session type . . . . . . . . . . > *FLASHCOPY *GEOMIR, *METROMIR...
ASP copy:
Preferred source . . . . . . . > IASCPYD1 Name
Preferred target . . . . . . . > IASCPYD2 Name
+ for more values
FlashCopy type . . . . . . . . . *NOCOPY *COPY, *NOCOPY
Persistent relationship . . . . *NO *YES, *NO

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 13-42 Start ASP session

Here we explain parameters used by STRASPSSN that might need more clarification:
– Session type
In our case we use *FLASHCOPY.
– Preferred source and target
This specifies the ASP copy descriptions the we created previously.The correct order
is important, as the source copy description points to the volumes that will be treated
as a source for the FlashCopy and the target description points to the target volumes.

Chapter 13. Configuring and managing DS8000 Copy Services 299


– Flashcopy type
This parameter specifies whether all content of the source volumes will be copied onto
the target volumes. The differences between those settings are described in “Full
volume copy” on page 110 and “Nocopy option” on page 110.
– Persistent relationship
This specifies whether the FlashCopy relation is persistent. When the relationship is
persistent, the DS storage system maintains the FlashCopy relationship and keeps
track of the changes to the IASP volumes even when all tracks have already been
copied. You need to set it to *YES when you are going to use incremental FlashCopy
or FlashCopy reverse.
3. Resume the IASP activity on the production system.
To resume the normal operations for a previously quiesced IASP on the production
system, you need to run the following command:
CHGASPACT ASPDEV(IASP2) OPTION(*RESUME)
Otherwise, if the IASP has been varied off on the production system before taking the
FlashCopy, it can now be varied on again using VRYCFG.
4. Vary on the IASP copy on the backup system.
Immediately after the FlashCopy session is started you can make the IASP copy available
to your backup system by issuing the following VRYCFG command on the backup system:
VRYCFG CFGOBJ(IASP2)
CFGTYPE(*ASP)
STATUS(*ON)
When the IASP2 becames AVAIALABLE you can use it.
5. End the FlashCopy session.
When you finish using your IASP2 copy you can end the FlashCopy session on the target
system as follows:
a. Vary off the IASP2:
VRYCFG CFGOBJ(IASP2) CFGTYPE(*ASP) STATUS(*OFF)
b. End the ASP session:
ENDASPSSN SSN(ASPSSN2)

300 PowerHA SystemMirror for IBM i Cookbook


Monitoring the status of the FlashCopy session
You can check the status of the FlashCopy session with DSPASPSSN (Figure 13-43). In the
output you can see the status of the IASP2 on each system. The UNKNOWN state for a
remote system is normal.

Display ASP Session DEMOFC


09/30/11 13:42:13
Session . . . . . . . . . . . . . . . . . . : ASPSSN2
Type . . . . . . . . . . . . . . . . . . : *FLASHCOPY
Persistent . . . . . . . . . . . . . . . : *NO
FlashCopy type . . . . . . . . . . . . . : *NOCOPY
Number sectors copied . . . . . . . . . . : 1536
Number sectors remaining to be copied . . : 67107328

Copy Descriptions

ASP
device Name Role State Node
IASP2 ASPCPYD1 SOURCE UNKNOWN DEMOHA
IASP2 ASPCPYD2 TARGET VARYOFF DEMOFC

Bottom
Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=Automatic refresh

Display ASP Session DEMOHA


09/30/11 13:42:07
Session . . . . . . . . . . . . . . . . . . : ASPSSN2
Type . . . . . . . . . . . . . . . . . . : *FLASHCOPY
Persistent . . . . . . . . . . . . . . . : *NO
FlashCopy type . . . . . . . . . . . . . : *NOCOPY
Number sectors copied . . . . . . . . . . : 1536
Number sectors remaining to be copied . . : 67107328

Copy Descriptions

ASP
device Name Role State Node
IASP2 ASPCPYD1 SOURCE AVAILABLE DEMOHA
IASP2 ASPCPYD2 TARGET UNKNOWN DEMOFC

Bottom
Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=Automatic refresh

Figure 13-43 ASP session status for FlashCopy on target and source system

Chapter 13. Configuring and managing DS8000 Copy Services 301


On the DS8000 you can check the status of the FlashCopy volume relationships with lsflash
(Figure 13-44).

dscli> lsflash -fmt default a010-a013


Date/Time: October 3, 2011 4:12:40 PM CEST IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
ID SrcLSS SequenceNum Timeout ActiveCopy Recording Persistent Revertible SourceWriteEnabled TargetWriteEnabled BackgroundCopy
====================================================================================================================================
A010:A020 A0 0 60 Disabled Enabled Enabled Disabled Enabled Enabled Enabled
A011:A021 A0 0 60 Disabled Enabled Enabled Disabled Enabled Enabled Enabled
A012:A022 A0 0 60 Disabled Enabled Enabled Disabled Enabled Enabled Enabled
A013:A023 A0 0 60 Disabled Enabled Enabled Disabled Enabled Enabled Enabled

Figure 13-44 lsflash for FlashCopy session

Using FlashCopy of the Metro Mirror target


For this scenario we use the configuration shown in Figure 13-45 and Table 13-3 on
page 303. In addition, the FlashCopy target volumes will be space efficient.

SITE 1 SITE 2

IBM Power Systems IBM Power Systems

DEMOHA

DEMOPROD DEMOFC

IASP MetroMirror IASP CPY1


SRC
FlashCopy

IASP CPY2

IBM System Storage IBM System Storage


Management IP: 9.5.168.55 Management IP: 9.5.168.55
Device ID: IBM.2107-75AY032 Device ID: IBM.2107-75AY032

Figure 13-45 Environment for Metro Mirror target FlashCopy

302 PowerHA SystemMirror for IBM i Cookbook


Table 13-3 Parameters for Metro Mirror target FlashCopy
Source IASP Target IASP

System name DEMOHA DEMOFC

Cluster name PWRHA_CLU

Device domain PWRHA_DMN

IASP name, number IASP1, 33 IASP1, 33

Management access IP 9.5.168.129 9.5.168.130

DS8000 device ID IBM.2107-75AY032 IBM.2107-75AY032

DS8000 IP address 9.5.168.55 9.5.168.55

Volume group ID V2 V48

Volumes IDsa 6000-6002 6100-6102 A600-A602 A700-A702

FC IO adapter resource nameb DC07 DC04

Host connection ID 1 2

Host connection WWPN 10000000C9523F21 10000000C9523E9D


a. Volumes need to be the same size on both sides.
b. FC stands for Fibre Channel.

We have configured Metro Mirror replication between DEMOPROD and DEMOHA systems.
Figure 13-46 shows the status of this replication. On the DS storage system we can list the
volumes used by IASP1 on the DEMOHA system and their properties (we show only one
volume’s details) (Figure 13-47 on page 304).

dscli> lspprc -fmt default 6000-6002 6100-6102


Date/Time: October 3, 2011 4:37:56 PM CEST IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
=========================================================================================================
6000:6000 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6001:6001 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6002:6002 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6100:6100 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid
6101:6101 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid
6102:6102 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid

Figure 13-46 Metro Mirror status

Chapter 13. Configuring and managing DS8000 Copy Services 303


dscli> showvolgrp v2
Date/Time: September 30, 2011 9:34:05 PM CEST IBM DSCLI Version: 7.6.10.530 DS:
IBM.2107-75AY032
Name demoha_powerha
ID V2
Type OS400 Mask
Vols 6000 6001 6002 6100 6101 6102
dscli> showfbvol 6000
Date/Time: September 30, 2011 9:34:11 PM CEST IBM DSCLI Version: 7.6.10.530 DS:
IBM.2107-75AY032
Name demoprod_powerha
ID 6000
accstate Online
datastate Normal
configstate Normal
deviceMTM 2107-A01
datatype FB 520P
addrgrp 6
extpool P0
exts 8
captype iSeries
cap (2^30B) 8.0
cap (10^9B) 8.6
cap (blocks) 16777216
volgrp V2
ranks 1
dbexts -
sam Standard
repcapalloc -
eam rotatevols
reqcap (blocks) 16777216
realextents 8
virtualextents 0
migrating 0
perfgrp -
migratingfrom -
resgrp -
Figure 13-47 Volume group and volume details for IASP1

304 PowerHA SystemMirror for IBM i Cookbook


To use space-efficient (SE) volumes, we need to create the space-efficient storage pool. The
SE storage is created in extension storage pool P5. The SE storage will have virtual size of
100 GB and the allocated physical size of 20 GB (Figure 13-48).

dscli> mksestg -repcap 20 -vircap 100 p5


Date/Time: October 5, 2011 6:21:37 PM CEST IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
CMUC00342I mksestg: The space-efficient storage for the extent pool P5 has been created successfully.
dscli> lssestg
Date/Time: October 5, 2011 6:26:10 PM CEST IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
extpool stgtype datastate configstate repcapstatus repcap(GiB/Mod1) vircap
==========================================================================
P0 fb Normal Normal below 300.0 2200.0
P1 fb Normal Normal below 300.0 2200.0
P3 fb Normal Normal below 300.0 3000.0
P5 fb Normal Normal below 20.0 100.0

Figure 13-48 Creation of the SE storage pool

Now we have to create the target volumes for the FlashCopy. The volumes need to be the
same type as the source ones, in this case A01. Using the DS CLI mkfbvol commands
(Figure 13-49), we create the volumes in the SE storage created earlier. We specify that the
volumes are track space efficient (TSE). After creating the FlashCopy target volumes we
assign them to a new volume group using the DS CLI mkvolgrp and chvolgrp command.

dscli> mkfbvol -extpool p5 -os400 a01 -sam tse 11a0-11a5


Date/Time: October 5, 2011 6:32:22 PM CEST IBM DSCLI Version: 7.6.10.530 DS:
IBM.2107-75AY032
CMUC00025I mkfbvol: FB volume 11A0 successfully created.
CMUC00025I mkfbvol: FB volume 11A1 successfully created.
CMUC00025I mkfbvol: FB volume 11A2 successfully created.
CMUC00025I mkfbvol: FB volume 11A3 successfully created.
CMUC00025I mkfbvol: FB volume 11A4 successfully created.
CMUC00025I mkfbvol: FB volume 11A5 successfully created.
dscli> mkvolgrp -type os400mask mmflashTGT
Date/Time: October 5, 2011 6:37:59 PM CEST IBM DSCLI Version: 7.6.10.530 DS:
IBM.2107-75AY032
CMUC00030I mkvolgrp: Volume group V49 successfully created.
dscli> chvolgrp -action add -volume 11A0-11A5 v49
Date/Time: October 5, 2011 6:38:43 PM CEST IBM DSCLI Version: 7.6.10.530 DS:
IBM.2107-75AY032
CMUC00031I chvolgrp: Volume group V49 successfully modified.

Figure 13-49 Creation of the volumes and adding them to a volume group

Chapter 13. Configuring and managing DS8000 Copy Services 305


Now we need to connect the volumes to the DEMOFC system for which we use an already
existing host connection for DEMOFC on the DS storage system and assign it our newly
created volume group V50 containing the FlashCopy target volumes (Figure 13-50).

dscli> showhostconnect 2
Date/Time: October 5, 2011 6:40:18 PM CEST IBM DSCLI Version: 7.6.10.530 DS:
IBM.2107-75AY032
Name demofc_powerha
ID 0002
WWPN 10000000C9523E9D
HostType iSeries
LBS 520
addrDiscovery reportLUN
Profile IBM iSeries - OS/400
portgrp 0
volgrpID -
atchtopo -
ESSIOport all
speed Unknown
desc -
dscli> chhostconnect -volgrp v49 2
Date/Time: October 5, 2011 6:40:35 PM CEST IBM DSCLI Version: 7.6.10.530 DS:
IBM.2107-75AY032
CMUC00013I chhostconnect: Host connection 0002 successfully modified.
dscli> showhostconnect 2
Date/Time: October 5, 2011 6:40:39 PM CEST IBM DSCLI Version: 7.6.10.530 DS:
IBM.2107-75AY032
Name demofc_powerha
ID 0002
WWPN 10000000C9523E9D
HostType iSeries
LBS 520
addrDiscovery reportLUN
Profile IBM iSeries - OS/400
portgrp 0
volgrpID V49
atchtopo -
ESSIOport all
speed Unknown
desc -
Figure 13-50 Changing host connect for volume group V50 and DEMOFC system

306 PowerHA SystemMirror for IBM i Cookbook


Now we can see the disks in the DEMOFC system (Figure 13-51).

Logical Hardware Resources Associated with IOP

Type options, press Enter.


2=Change detail 4=Remove 5=Display detail 6=I/O debug
7=Verify 8=Associated packaging resource(s)

Serial Part
Opt Description Type-Model Number Number
Combined Function IOP 2844-001 53-7141345 0000039J1719
Storage IOA 2787-001 1F-C5500BA 0000080P6417
Disk Unit 2107-A01 50-11A0BB8
Disk Unit 2107-A01 50-11A1BB8
Disk Unit 2107-A01 50-11A2BB8
Disk Unit 2107-A01 50-11A3BB8
Disk Unit 2107-A01 50-11A4BB8
Disk Unit 2107-A01 50-11A5BB8

F3=Exit F5=Refresh F6=Print F8=Include non-reporting resources


F9=Failed resources F10=Non-reporting resources
F11=Display logical address F12=Cancel

Figure 13-51 Volumes in the DEMOFC system

Chapter 13. Configuring and managing DS8000 Copy Services 307


We have created the required target volumes for the FlashCopy and connected them to our
backup system DEMOFC. Now we need to create the objects needed to use FlashCopy with
PowerHA SystemMirror for i:
1. Create the ASP copy descriptions.
Figure 13-52 and Figure 13-53 on page 309 shows the commands used for creation of the
ASP Copy descriptions needed for the FlashCopy.

Add ASP Copy Description (ADDASPCPYD)

Type choices, press Enter.

ASP copy . . . . . . . . . . . . > ASP1FC1 Name


ASP device . . . . . . . . . . . > IASP1 Name
Cluster resource group . . . . . *NONE Name, *NONE
Cluster resource group site . . *NONE Name, *NONE
Storage host:
User name . . . . . . . . . . > userid
Password . . . . . . . . . . . > password
Internet address . . . . . . . > '9.5.168.55'

Location . . . . . . . . . . . . > DEMOHA Name, *DEFAULT, *NONE


Logical unit name:
TotalStorage device . . . . . > 'IBM.2107-75AY032'
Logical unit range . . . . . . > '6000-6002' Character value
+ for more values > '6100-6102'
Consistency group range . . . Character value
+ for more values
Recovery domain:
Cluster node . . . . . . . . . *NONE Character value, *NONE
Host identifier . . . . . . . Character value
+ for more values
Volume group . . . . . . . . . Character value
+ for more values
+ for more values

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys

Figure 13-52 ADDASPCPYD for the source of the FlashCopy

308 PowerHA SystemMirror for IBM i Cookbook


Add ASP Copy Description (ADDASPCPYD)

Type choices, press Enter.

ASP copy . . . . . . . . . . . . > ASP1FC2 Name


ASP device . . . . . . . . . . . > IASP1 Name
Cluster resource group . . . . . *NONE Name, *NONE
Cluster resource group site . . *NONE Name, *NONE
Storage host:
User name . . . . . . . . . . > userid
Password . . . . . . . . . . . > password
Internet address . . . . . . . > '9.5.168.55'

Location . . . . . . . . . . . . > DEMOFC Name, *DEFAULT, *NONE


Logical unit name:
TotalStorage device . . . . . > 'IBM.2107-75AY032'
Logical unit range . . . . . . > '11A0-11A5' Character value
+ for more values >
Consistency group range . . . Character value
+ for more values
Recovery domain:
Cluster node . . . . . . . . . *NONE Character value, *NONE
Host identifier . . . . . . . Character value
+ for more values
Volume group . . . . . . . . . Character value
+ for more values
+ for more values

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys

Figure 13-53 ADDASPCPYD for the target FlashCopy

Chapter 13. Configuring and managing DS8000 Copy Services 309


2. Start the ASP FlashCopy session.
We start the FlashCopy with the following command:
STRASPSSN SSN(ASPFC1) TYPE(*FLASHCOPY) ASPCPY((ASP1FC1 ASP1FC2))
Figure 13-54 shows the status of the FlashCopy. The FlashCopy target on the
DEMOFC system can be varied on and used. Figure 13-55 on page 311 shows the
status on the DS8000.

Display ASP Session DEMOFC


10/05/11 12:00:50
Session . . . . . . . . . . . . . . . . . . : ASPFC1
Type . . . . . . . . . . . . . . . . . . : *FLASHCOPY
Persistent . . . . . . . . . . . . . . . : *NO
FlashCopy type . . . . . . . . . . . . . : *NOCOPY
Number sectors copied . . . . . . . . . . : 177536
Number sectors remaining to be copied . . : 100485760

Copy Descriptions

ASP
device Name Role State Node
IASP1 ASP1FC1 SOURCE UNKNOWN DEMOHA
IASP1 ASP1FC2 TARGET AVAILABLE DEMOFC

Bottom
Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=Automatic refresh

Figure 13-54 ASP session status

310 PowerHA SystemMirror for IBM i Cookbook


dscli> lsflash 11a0-11a5
Date/Time: October 5, 2011 7:11:27 PM CEST IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
ID SrcLSS SequenceNum Timeout ActiveCopy Recording Persistent Revertible SourceWriteEnabled TargetWriteEnabled BackgroundCopy
====================================================================================================================================
6000:11A0 60 0 60 Disabled Disabled Disabled Disabled Enabled Enabled Disabled
6001:11A1 60 0 60 Disabled Disabled Disabled Disabled Enabled Enabled Disabled
6002:11A2 60 0 60 Disabled Disabled Disabled Disabled Enabled Enabled Disabled
6100:11A3 61 0 60 Disabled Disabled Disabled Disabled Enabled Enabled Disabled
6101:11A4 61 0 60 Disabled Disabled Disabled Disabled Enabled Enabled Disabled
6102:11A5 61 0 60 Disabled Disabled Disabled Disabled Enabled Enabled Disabled
dscli> lssestg
Date/Time: October 5, 2011 7:11:31 PM CEST IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
extpool stgtype datastate configstate repcapstatus repcap(GiB/Mod1) vircap
==========================================================================
P0 fb Normal Normal below 300.0 2200.0
P1 fb Normal Normal below 300.0 2200.0
P3 fb Normal Normal below 300.0 3000.0
P5 fb Normal Normal below 20.0 100.0
dscli> showsestg p5
Date/Time: October 5, 2011 7:11:33 PM CEST IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
extpool P5
stgtype fb
datastate Normal
configstate Normal
repcapstatus below
%repcapthreshold 0
repcap(GiB) 20.0
repcap(Mod1) -
repcap(blocks) 41943040
repcap(cyl) -
repcapalloc(GiB/Mod1) 0.1
%repcapalloc 0
vircap(GiB) 100.0
vircap(Mod1) -
vircap(blocks) 209715200
vircap(cyl) -
vircapalloc(GiB/Mod1) 48.0
%vircapalloc 48
overhead(GiB/Mod1) 2.0
reqrepcap(GiB/Mod1) 20.0
reqvircap(GiB/Mod1) 100.0

Figure 13-55 lsflash for the FlashCopied volumes

3. End the FlashCopy session.


When you finish your activities on the target FlashCopy IASP, you need to end the ASP
session. To do so you need to vary off the IASP1 and then use ENDASPSSN SSN(ASPFC1).

13.1.3 Configuring IBM i DS8000 Global Mirror (CL commands)


This section describes a scenario using a DS8000 Global Mirror environment as the
hardware replication technology. Most of the PowerHA GUI panels wizard are common
between Metro Mirror and Global Mirror. Differences exist only when selecting a Metro Mirror
or Global Mirror mirroring solution, and for the ASP copy descriptions for which a consistency
group is mandatory for Global Mirror. This is the reason that we decided to show you only the
CL commands for this scenario and point out the differences between configuring Global
Mirror versus Metro Mirror with the PowerHA GUI. Each time that names for items such as
copy descriptions or sessions display, you have to provide the proper information. The main
differences are for the following panels:
 The panel shown in Figure 13-6 on page 273, where you select Global Mirror in place of
Metro Mirror.
 The panel shown in Figure 13-25 on page 285, where Global Mirror mirroring is selected
due to previous wizard selection.
 Provide information for consistency group logical volumes similar to steps 25 through 30 in
“Setting up a Metro Mirror environment” on page 266.

Chapter 13. Configuring and managing DS8000 Copy Services 311


Environment overview
As for Metro Mirror scenarios, we need to create several cluster items. Cluster, cluster
monitors, IASP, and administrative domain setup are covered in Chapter 11, “Creating a
PowerHA base environment” on page 199. They are common to all scenarios.

These are Global Mirror specific cluster items:


 Cluster resource group
 Device domain, which is the same as a Metro Mirror environment
 Global Mirror copy descriptions and session

Table 13-4 and Figure 13-56 on page 313 describe and show our environment.

Table 13-4 Global Mirror scenario settings


Preferred production Preferred backup

System name DEMOPROD DEMOHA

Cluster name PWRHA_CLU

Cluster resource group PWRHA_CRG2

Device domain PWRHA_DMN

Administrative domain PWRHA_CAD

IASP name, number IASP2, 182 IASP2, 182

Cluster resource group site name SITE1 SITE2

ASP copy description IASP2_GM1 IASP2_GM2

ASP session IASP2_GM

Takeover IP 10.0.0.1

Heartbeat cluster IP 10.10.10.1 10.10.10.2

Management access IP 9.5.168.129 9.5.168.130

DS8000 device ID IBM.2107-13ABGAA IBM.2107-1375210

DS8000 IP addressa 9.5.168.32 9.5.168.32

Volume group ID V15 V13

Volumes IDsb 0200-0201, 0300-0301 0200-0201, 0300-0301

Consistency group volumes IDs 0250-0251, 0350-0351 0250-0251, 0350-0351

FC IO adapter resource namec DC05 DC05

Host connection ID 07 01

Host connection WWPN 10000000C94802CB 10000000C94122A2


a. IP addresses are identical because in our environment we use a single management console
for both DS8000s.
b. Volumes IDs are not needed to be the same on the source and the target Global Mirror
relationships.
c. FC stands for Fibre Channel.

312 PowerHA SystemMirror for IBM i Cookbook


User access network
Heartbeat network

10.10.10.1

10.10.10.2
IBM Power
Systems server

DEMOPROD DEMOHA
IBM Power IBM i partition IBM i partition
Systems server

Host conn. id 7 for 10000000C94802CB on DC05

Host conn. Id 1 for 10000000C94122A2on DC05


9.5.168.129

9.5.168.130
Production Site Backup Site

Management network

9.5.168.129
9.5.168.129

IBM.2107-13ABGAA IBM.2107-1375210

IBM System Storage IBM System Storage

Global Copy pair


02
00

02
01

02
00

02
01

Global Copy pair


03
00

03
01

03
00

03
01

Volumes group V15 Volumes group V13


production mode
failback mode
Flashcopy in

Flashcopy in

A volumes B volumes
normal
02
50

02
51

02
50

02
51
03
50

03
51

03
50

03
51

D volumes C volumes

Figure 13-56 Global Mirror scenario settings

Chapter 13. Configuring and managing DS8000 Copy Services 313


Setting up a Global Mirror environment
Global Mirror environment setup is not handled by IBM PowerHA SystemMirror for i. You
have to set up all items directly to the DS8000. Refer to 6.3, “Global Mirror” on page 98, for
more information.

Creating the PPRC path


The path management activities are exactly the same for the Global Copy part of Global
Mirror as for Metro Mirror. Therefore, see “Creating the PPRC paths” on page 267 for detailed
information about path commands.

Note: Make sure that the paths exist on both DS8000s from one to the other for the
volumes’ LSS.

Creating the Global Copy relationships


After the path exists from Production DS8000 to Backup DS8000, we can establish the
Global Copy relationships. For that, we use mkpprc with the parameter type set to gcp in place
of mmir (Example 13-7). In our Global Mirror environment, Global Copy relationships are to
be established between 0200 - 0201 and 0200 - 0201 volumes for the first set, and 0300 -
0301 and 0300 - 0301 volumes for the second set. The relationship is properly established
when the first pass status is True.

Example 13-7 Creating Global Copy relationships


dscli> mkpprc -dev IBM.2107-13ABGAA -remotedev IBM.2107-1375210 -type gcp -mode full 0200-0201:0200-0201
0300-0301:0300-0301
Date/Time: October 3, 2011 11:48:40 AM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
CMUC00153I mkpprc: Remote Mirror and Copy volume pair relationship 0200:0200 successfully created.
CMUC00153I mkpprc: Remote Mirror and Copy volume pair relationship 0201:0201 successfully created.
CMUC00153I mkpprc: Remote Mirror and Copy volume pair relationship 0300:0300 successfully created.
CMUC00153I mkpprc: Remote Mirror and Copy volume pair relationship 0301:0301 successfully created.
dscli> lspprc 0200-03FF
Date/Time: October 3, 2011 11:49:11 AM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
==================================================================================================
0200:0200 Copy Pending - Global Copy 02 60 Disabled False
0201:0201 Copy Pending - Global Copy 02 60 Disabled False
0300:0300 Copy Pending - Global Copy 03 60 Disabled False
0301:0301 Copy Pending - Global Copy 03 60 Disabled False
dscli> lspprc 0200-03FF
Date/Time: October 3, 2011 11:52:22 AM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
==================================================================================================
0200:0200 Copy Pending - Global Copy 02 60 Disabled True
0201:0201 Copy Pending - Global Copy 02 60 Disabled True
0300:0300 Copy Pending - Global Copy 03 60 Disabled True
0301:0301 Copy Pending - Global Copy 03 60 Disabled True

Note: In a Global Copy relationship, the source volumes remain at copy pending status
and the target volumes at target copy pending status. They never reach full duplex and
target full duplex, because of the asynchronous process.

Creating the FlashCopy relationships on the backup site


After the Global Copy is established between the production DS8000 and the backup
DS8000, we can establish the FlashCopy relationships (between B and C volumes) on the
backup DS8000. For that, we use mkflash (Example 13-8 on page 315). In our Global Mirror

314 PowerHA SystemMirror for IBM i Cookbook


environment, Flash Copy relationships are to be established between 0200 - 0201 and 0250 -
0251 volumes for a first set, and 0300 - 0301 and 0350 - 0351 volumes for a second set. The
relationship is properly established as soon as the command returns. Specific parameters for
a Flash Copy relationship used in a Global Mirror session are as follows:
 persist (persistent)
 tgtinhibit (target write inhibit)
 record (record changes)
 nocp (no full copy)

Example 13-8 Global Mirror FlashCopy relationships


dscli> mkflash -persist -nocp -tgtinhibit -record 0200-0201:0250-0251 0300-0301:0350-0351
Date/Time: October 3, 2011 11:54:00 AM CDT IBM DSCLI Version: 7.6.10.530 DS:
IBM.2107-13ABGAA
CMUC00137I mkflash: FlashCopy pair 0200:0250 successfully created.
CMUC00137I mkflash: FlashCopy pair 0201:0251 successfully created.
CMUC00137I mkflash: FlashCopy pair 0300:0350 successfully created.
CMUC00137I mkflash: FlashCopy pair 0301:0351 successfully created.

Making a Global Mirror session


After FlashCopy relationships exist on the backup system, we can create the Global Mirror
session. For automatic switchback to be managed by PowerHA, the session must be created
on each DS8000 (in our case, IBM.2107-13ABGAA is the Production DS8000 and
IBM.2107-1375210 is the Backup DS8000). To do that, we use mksession (Example 13-9).
mksession applies to an LSS. At the same time, we include our volumes in the session on the
production DS8000 only. No volume must be included in the session on the backup DS8000.

Example 13-9 Global Mirror session creation


dscli> mksession -dev IBM.2107-13ABGAA -lss 02 -volume 0200-0201 2
Date/Time: October 3, 2011 12:02:04 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
CMUC00145I mksession: Session 2 opened successfully.
dscli> mksession -dev IBM.2107-13ABGAA -lss 03 -volume 0300-0301 2
Date/Time: October 3, 2011 12:02:37 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
CMUC00145I mksession: Session 2 opened successfully.
dscli> lssession 02-03
Date/Time: October 3, 2011 12:02:44 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
LSS ID Session Status Volume VolumeStatus PrimaryStatus SecondaryStatus FirstPassComplete
AllowCascading
===========================================================================================================
======
02 02 Normal 0200 Join Pending Primary Copy Pending Secondary Simplex True Disable
02 02 Normal 0201 Join Pending Primary Copy Pending Secondary Simplex True Disable
03 02 Normal 0300 Join Pending Primary Copy Pending Secondary Simplex True Disable
03 02 Normal 0301 Join Pending Primary Copy Pending Secondary Simplex True Disable
dscli> mksession -dev IBM.2107-1375210 -lss 02 1
Date/Time: October 3, 2011 12:05:25 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-1375210
CMUC00145I mksession: Session 1 opened successfully.
dscli> mksession -dev IBM.2107-1375210 -lss 03 1
Date/Time: October 3, 2011 12:05:37 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-1375210
CMUC00145I mksession: Session 1 opened successfully.
dscli> lssession 02-03
Date/Time: October 3, 2011 12:36:59 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-1375210
LSS ID Session Status Volume VolumeStatus PrimaryStatus SecondaryStatus FirstPassComplete AllowCascading
========================================================================================================
02 01 - - - - - - -
03 01 - - - - - - -

Chapter 13. Configuring and managing DS8000 Copy Services 315


Starting Global Mirror session
After A volumes have been added to the session on the production system, we can start the
session. For that, we use mkgmir (Example 13-10). mkgmir applies to an LSS. You use this
command on only one LSS.

Example 13-10 Starting Global Mirror session


dscli> mkgmir -dev IBM.2107-13ABGAA -lss 02 -session 2
Date/Time: October 3, 2011 12:28:24 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
CMUC00162I mkgmir: Global Mirror for session 2 successfully started.
dscli> lssession 02-03
Date/Time: October 3, 2011 12:28:33 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
LSS ID Session Status Volume VolumeStatus PrimaryStatus SecondaryStatus FirstPassComplete
AllowCascading
========================================================================================================================
=
02 02 CG In Progress 0200 Active Primary Copy Pending Secondary Simplex True Disable
02 02 CG In Progress 0201 Active Primary Copy Pending Secondary Simplex True Disable
03 02 CG In Progress 0300 Active Primary Copy Pending Secondary Simplex True Disable
03 02 CG In Progress 0301 Active Primary Copy Pending Secondary Simplex True Disable

Using CL commands
The corresponding CL commands for creating a device domain and a cluster resource group
for a Global Mirror scenario are exactly the same as the ones used for the Metro Mirror
scenario. Refer to “Using CL commands” on page 293 for more information.

The related command for adding a Global Mirror ASP copy description is ADDASPCPYD
(Example 13-11). You have to create one for each IASP copy (that is, one for the Global Mirror
source and one for the target site). The difference with the Metro Mirror ASP copy description
creation command resides in the consistency group volume specification. In our case,
volumes 0250 - 0251 and 0350 - 0351 are those volumes on both source and target Global
Mirror copy descriptions.

Example 13-11 Adding Global Mirror ASP copy descriptions


ADDASPCPYD ASPCPY(IASP2_GM1) ASPDEV(IASP2) CRG(PWRHA_CRG2) LOCATION(*DEFAULT)
SITE(SITE1) STGHOST('qlpar' 'password' ('9.5.168.32'))
LUN('IBM.2107-13ABGAA' ('0200-0201' '0300-0301') ('0250-0251' '0350-0351'))
ADDASPCPYD ASPCPY(IASP2_GM2) ASPDEV(IASP2) CRG(PWRHA_CRG2) LOCATION(*DEFAULT)
SITE(SITE2) STGHOST('qlpar' 'password' ('9.5.168.32'))
LUN('IBM.2107-1375210' ('0200-0201' '0300-0301') ('0250-0251' '0350-0351'))

316 PowerHA SystemMirror for IBM i Cookbook


Starting ASP session and cluster resource group using the CL
commands
When creating the ASP copy descriptions through the GUI, at the end of the procedure the
session is automatically started. However, if you need to start it again, use STRASPSSN. As shown
in Example 13-12, STRCRG can be used to start the cluster resourcegGroup when needed.

Note: STRASPSSN fails if at least one of the consistency group volumes is included in a
volume group. PowerHA assumes that there is a DS8000 configuration error in this case
because the consistency group volumes must not be accessed by any host, which is
supposed to be the case if they are members of a volume group.

Example 13-12 Starting the ASP session and cluster resource group
STRASPSSN SSN(IASP2_GM) TYPE(*GLOBALMIR) ASPCPY((IASP2_GM1 IASP2_GM2))
STRCRG CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG2)

13.1.4 Configuring IBM i DS8000 LUN-level switching


This section describes the scenario of DS8000 LUN-level switching between two IBM i
servers or partitions. For this scenario, we need to create several cluster items. Clusters,
cluster monitors, and the IASP setup are covered in Chapter 11, “Creating a PowerHA base
environment” on page 199. They are common to all scenarios of this chapter.

Cluster items that we specifically configure for our LUN-level switching environment are
cluster resource groups.

Environment overview
In this scenario, we use and switch DS8000 LUNs with volume group ID V48 between
partition DEMOFC and partition DEMOHA. Table 13-5 provides more detailed information
about the configuration.

Table 13-5 Environment overview


Production Backup

System name/node name DEMOFC DEMOHA

Cluster name PWRHA_CLU

Device CRG PWRHA_CRG3

Device domain PWRHA_DMN

IASP IASP_LUN

Site name SITE1 SITE1

Heartbeat cluster IP 10.0.0.3 10.0.0.2

DS8000 device ID IBM.2107-75AY032

DS8000 IP 9.5.168.55

Host connection ID 0002 0005

Volume group ID V48

Volumes IDs A020-A023

Chapter 13. Configuring and managing DS8000 Copy Services 317


Production Backup

FC IOA resource name DC04 DC05

Host connection WWPN 10000000C9523E9D 10000000C94122A2

IBM Power Systems

IBM Storage Systems DS8000


Production Site Device ID: IBM.2107-75AY032
WWPN : 10000000C9523E9D
DEMOFC
Volume Group ID: V48
*SYSBAS
LUNs : A020~A023

IASP_LUN

Backup Site
WWPN : 10000000C94122A2
DEMOHA
*SYSBAS

Figure 13-57 The schematics of our scenario

Configuration on the DS8000


For DS8000 LUN-level switching, you need to prepare host identifiers, volume groups, and
LUNs on DS8000. In this scenario, A020-A023 LUNs are assigned into volume group V48,
and host connection IDs 0002 and 0005 are used for DEMOFC and DEMOHA
(Example 13-13).

Note: When configuring LUN-level switching, a volume group is assigned only to the IBM i
production node. The host connection for the backup node is not allowed to have a volume
group assigned, as it will be automatically assigned when LUNs are switched by PowerHA.

Example 13-13 Information about volume groups and host connection on DS8000
dscli> showvolgrp V48
Date/Time: September 29, 2011 9:28:20 AM CDT IBM DSCLI Version: 7.6.10.530 DS:
IBM.2107-75AY032
Name V48fc
ID V48
Type OS400 Mask
Vols A020 A021 A022 A023
dscli> lshostconnect 2 5
Date/Time: September 28, 2011 11:27:31 AM CDT IBM DSCLI Version: 7.6.10.530 DS:
IBM.2107-75AY032
Name ID WWPN HostType Profile portgrp volgrpID
ESSIOport

318 PowerHA SystemMirror for IBM i Cookbook


===========================================================================================
==
demofc_powerha 0002 10000000C9523E9D iSeries IBM iSeries - OS/400 0 V48 all
DEMOHA_FC 0005 10000000C94122A2 iSeries IBM iSeries - OS/400 0 - all

Configuration on IBM i
Create a basic cluster environment for LUN-level switching, as described in Chapter 11,
“Creating a PowerHA base environment” on page 199:
1. Create a cluster with two nodes, DEMOFC and DEMOHA, and a device domain
(Example 13-14). The cluster nodes are automatically started as the default setting.

Example 13-14 Creating a cluster environment


CRTCLU CLUSTER(PWRHA_CLU)
NODE((DEMOFC ('10.0.0.3')) (DEMOHA ('10.0.0.2')))
START(*YES) VERSION(*CUR) HAVERSION(*CUR)
CLUMSGQ(QSYS/QSYSOPR) FLVWAITTIM(*NOMAX) FLVDFTACN(*PROCEED)

ADDDEVDMNE CLUSTER(PWRHA_CLU) DEVDMN(PWRHA_DMN) NODE(DEMOFC)


ADDDEVDMNE CLUSTER(PWRHA_CLU) DEVDMN(PWRHA_DMN) NODE(DEMOHA)

2. For production node DEMOFC, create an IASP using LUNs previously assigned by
DS8000. For backup node DEMOHA we create an IASP device description with the same
name that we used on the production node (Example 13-15).

Example 13-15 Creating IASP


On Production site
CFGDEVASP ASPDEV(IASP_LUN) ACTION(*CREATE) UNITS(DD026 DD025 DD024 DD023)

On Backup site
CRTDEVASP DEVD(IASP_LUN) RSRCNAME(IASP_LUN)

3. After creating a cluster environment and IASP, we create a device cluster resource group
using CRTCRG (Example 13-16). At this time do not list the IASP as a configuration object
for the CRG.

Example 13-16 Create cluster resource group


CRTCRG CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG3)
CRGTYPE(*DEV) EXITPGM(*NONE) USRPRF(*NONE)
RCYDMN((DEMOFC *PRIMARY *LAST SITE1) (DEMOHA *BACKUP *LAST SITE1))

Note: Make sure that you set the same site name for each recovery domain node. In
our scenario we use the site name SITE1 for node DEMOFC and node DEMOHA.

Chapter 13. Configuring and managing DS8000 Copy Services 319


4. Create the ASP copy description for the IASP to be used with LUN-level switching using
ADDASPCPYD (Example 13-17).

Example 13-17 Create ASP Copy Description


ADDASPCPYD ASPCPY(LUN_SWITCH) ASPDEV(IASP_LUN) CRG(PWRHA_CRG3) SITE(SITE1)
STGHOST(powerha () ('9.5.168.55')) LOCATION(*DEFAULT)
LUN('IBM.2107-75AY032' ('A020-A023') ())
RCYDMN((DEMOFC (0002) (V48)) (DEMOHA (0005) (V48)))

On this command, specify the host identifiers and volume groups for each node in the
recovery domain parameter (RCYDMN), even though we do not make the volume group
assign a host connection ID for the backup node DEMOHA (Example 13-13 on page 318).
5. Add the IASP to the configuration object list of device CRGs created in step 3 using
ADDCRGDEVE (Example 13-18). The reason why we did not add the configuration object list
when we created the CRG is that we migth get misleading error messages in the job log.
This happens when the IASP is listed in the configuration object list of CRGs before the
copy description is created. After the copy description is created, PowerHA tells the CRG
that PowerHA is in control of the switching, which causes the CRG to skip its IOP and
tower switching checks. If the IASP is listed in the configuration object list of CRGs before
the copy description is created, then we see misleading error messages in the job log.
After the copy description is created, PowerHA tells the CRG that PowerHA is in control of
the switching, which causes the CRG to skip its IOP and tower switching checks.

Example 13-18 Add the configuration object to CRG


ADDCRGDEVE CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG3) CFGOBJ((IASP_LUN))

6. After adding the IASP to the configuration object list of the device CRGs, we start the CRG
for LUN-level switching between the production site and the backup site using STRCRG
(Example 13-19).

Example 13-19 Start CRG


STRCRG CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG3)

Now your IBM PowerHA SystemMirror for i DS8000 LUN-level switching configuration is
complete and your IASP is highly available for planned and unplanned switches, as described
in 13.2.2, “Using CL commands for DS8000 LUN-level switching” on page 346.

13.2 Managing IBM i DS8000 Copy Services


In this section we describe how to manage the DS8000 environment for switchover and
switchback scenarios for Metro and Global Mirror.

13.2.1 Switchover and switchback for a Metro Mirror or Global Mirror planned
outage
In this section we describe the procedure for a planned site switch for Metro Mirror using the
PowerHA GUI, and for Global Mirror using CL commands.

From the user standpoint, the steps for performing a switch for both scenarios are the same.
The difference is only related to DS8000 commands done under-the-covers by PowerHA to
perform the actions.

320 PowerHA SystemMirror for IBM i Cookbook


Operations to perform the switchback are also exactly the same as the switch except in the
case of asymmetrical Global Mirror configuration. Switchback for this specific one will be
described in detail using CL commands.

Using the PowerHA GUI for a Metro Mirror planned switchover

Note: Before starting any switchover make sure that no jobs are using the IASP any longer.

Before starting the switch, we look at our current configuration. DEMOPROD is the current
production system, and DEMOHA is the current first backup system.

In Figure 13-58 and Figure 13-59 on page 322, you can see the IP address configuration on
both DEMOPROD and DEMOHA systems. On DEMOPROD the takeover IP address 10.0.0.1
is active. On DEMOHA, this IP address is inactive. The takeover IP address is usually the one
to which all users and systems connect.

Work with TCP/IP Interfaces


System: DEMOPROD
Type options, press Enter.
1=Add 2=Change 4=Remove 5=Display 9=Start 10=End

Internet Subnet Interface Alias


Opt Address Mask Status Name

9.5.168.129 255.255.255.0 Active *NONE


9.5.168.133 255.255.255.0 Failed *NONE
9.5.168.134 255.255.255.0 Failed *NONE
10.0.0.1 255.255.255.0 Active POWERHA_TAKEOVER_IP
10.10.10.1 255.255.255.0 Active POWERHA
127.0.0.1 255.0.0.0 Active LOCALHOST
192.168.86.10 255.255.255.0 Active *NONE
192.168.87.10 255.255.255.0 Active *NONE

Bottom
F3=Exit F5=Refresh F6=Print list F11=Display line information
F12=Cancel F17=Top F18=Bottom
Figure 13-58 Metro Mirror: IP configuration on preferred production before switchover

Chapter 13. Configuring and managing DS8000 Copy Services 321


Work with TCP/IP Interfaces
System: DEMOHA
Type options, press Enter.
1=Add 2=Change 4=Remove 5=Display 9=Start 10=End

Internet Subnet Interface Alias


Opt Address Mask Status Name

9.5.168.130 255.255.255.0 Active *NONE


9.5.168.133 255.255.255.0 Active *NONE
9.5.168.134 255.255.255.0 Active *NONE
10.0.0.1 255.255.255.0 Inactive POWERHA_TAKEOVER_IP
10.10.10.2 255.255.255.0 Active POWERHA
127.0.0.1 255.0.0.0 Active LOCALHOST
192.168.86.11 255.255.255.0 Active *NONE
192.168.87.11 255.255.255.0 Active *NONE

Bottom
F3=Exit F5=Refresh F6=Print list F11=Display line information
F12=Cancel F17=Top F18=Bottom
Figure 13-59 Metro Mirror: IP configuration on preferred backup before switchover

Figure 13-60 shows the current status for DS volume members of the Metro Mirror
relationships. With lspprc on the current production DS8000 75AY031, all volumes are full
duplex, which is the normal status for source volumes. On the current backup DS8000
75AY032, all volumes are target full duplex, which is the normal status for target volumes.

dscli> lspprc -dev IBM.2107-75AY031 6000-61FF


Date/Time: September 29, 2011 10:18:16 AM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
==================================================================================================
6000:6000 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6001:6001 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6002:6002 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6100:6100 Full Duplex - Metro Mirror 61 60 Disabled Invalid
6101:6101 Full Duplex - Metro Mirror 61 60 Disabled Invalid
6102:6102 Full Duplex - Metro Mirror 61 60 Disabled Invalid

dscli> lspprc -dev IBM.2107-75AY032 6000-61FF


Date/Time: September 29, 2011 10:28:06 AM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
=========================================================================================================
6000:6000 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6001:6001 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6002:6002 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6100:6100 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid
6101:6101 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid
6102:6102 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid

Figure 13-60 Metro Mirror volumes status before switchover

322 PowerHA SystemMirror for IBM i Cookbook


In this section we describe how to manage cluster resource groups and how to do
a switchover.

Note: Any cluster node can be selected to perform a switchover as long as it is in active status.

1. Switchover activity is related to the cluster resource group. Therefore, to enter the
switchover wizard (Figure 13-61), click Cluster Resource Groups.

Figure 13-61 Metro Mirror switchover starting point

Chapter 13. Configuring and managing DS8000 Copy Services 323


2. Select Switchover against the cluster resource group that you are working on. In our
case, this is PWRHA_CRG1 (Figure 13-62).

Figure 13-62 Metro Mirror launch switchover

3. Before starting the switchover, we have the opportunity to review the CRG and recovery
domain current status, and what will be the result of the operation. In our case
(Figure 13-63), all statuses are correct, and after the switchover DEMOPROD will be the
first backup site and DEMOHA will be the production site. If everything is correct, click OK.

Figure 13-63 Metro Mirror switchover confirmation

324 PowerHA SystemMirror for IBM i Cookbook


Note: Even if an item is not in the appropriate status, the wizard gives you the ability to
bypass its warning by clicking OK (Figure 13-64 on page 325) and tries to perform the
switchover if you request it. However it might fail, in which case you will have to
manually fix it.

Figure 13-64 Metro Mirror confirm switchover

4. The wizard displays the switchover progress. There are seven steps (Figure 13-65):
a. The first step ends all jobs using the IASP.
b. The second step varies off the IASP on the production system.

Figure 13-65 Metro Mirror switchover first steps

Chapter 13. Configuring and managing DS8000 Copy Services 325


We might want to monitor the independent ASP vary status through a 5250 session
using DSPASPSTS on the production system. Figure 13-66 shows the result. At this
moment, the IASP is not yet varied off, but no job can access it anymore.

Display ASP Vary Status

ASP Device . . . . : IASP1 Step . . . . . . . : 5 / 5


ASP Number . . . . : 33 Current time . . . : 00:00:13
ASP State . . . . : VARIED ON Previous time . . : 00:00:58

Step Elapsed
time
Cluster vary job submission
Ending jobs using the ASP 00:00:00
Waiting for jobs to end 00:00:04
Image catalog synchronization
> Writing changes to disk 00:00:08

Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=End automatic refresh


Figure 13-66 Metro Mirror: Independent ASP varying off on preferred production

326 PowerHA SystemMirror for IBM i Cookbook


Figure 13-67 shows the next switchover steps.
c. The third step ends the takeover IP interface on the production system.
d. The fourth step performs DS8000 activity for the Metro Mirror relationships to be
reversed. failoverpprc is issued on the backup DS8000 for the target volumes to
become source suspended, followed by failbackpprc for the volumes to be
synchronized back from the backup DS8000 to the production DS8000.
e. The fifth steps ensures that the cluster resource group is started.
f. The sixth steps varies on the independent ASP on the backup system.

Figure 13-67 Metro Mirror switchover almost complete

Chapter 13. Configuring and managing DS8000 Copy Services 327


We might want to monitor the independent ASP vary status through a 5250 session
using DSPASPSTS on the backup system. Figure 13-68 shows the result. At this moment,
the IASP is not yet available. The database cross-reference file merge is in progress.

Display ASP Vary Status

ASP Device . . . . : IASP1 Step . . . . . . . : 30 / 34


ASP Number . . . . : 33 Current time . . . : 00:00:07
ASP State . . . . : ACTIVE Previous time . . : 00:01:47

Step Elapsed
time
UID/GID mismatch correction 00:00:00
Database access path recovery 00:00:01
> Database cross-reference file merge 00:00:01
SPOOL initialization
Image catalog synchronization
Command analyzer recovery
Catalog validation

Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=End automatic refresh

Figure 13-68 Metro Mirror, Independent ASP varying on preferred backup

g. After the seventh step has completed, which is starting the takeover IP address on the
backup system, the overall switchover is complete (Figure 13-69).

Figure 13-69 Metro Mirror switchover completed

328 PowerHA SystemMirror for IBM i Cookbook


5. After switchover completion, when coming back to the Cluster Resource Group panel
(Figure 13-70), a refresh is needed to get the appropriate status. Click Refresh.

Figure 13-70 Cluster Resource Group panel after completed Metro Mirror switchover

6. We now have the correct status for the cluster resource group that we just switched over
from production to backup, from the DEMOPROD to the DEMOHA node (Figure 13-71).

Figure 13-71 Metro Mirror cluster resource group status after switchover

After the switchover is complete, we can compare the situation to the one before the
switchover. DEMOPROD is the current first backup system, and DEMOHA the current
production system.

Chapter 13. Configuring and managing DS8000 Copy Services 329


Figure 13-72 and Figure 13-73 on page 331 show the IP address configuration on both
DEMOPROD and DEMOHA systems. On DEMOHA, takeover IP address 10.0.0.1 is active.
On DEMOPROD, this IP address is now inactive.

Work with TCP/IP Interfaces


System: DEMOPROD
Type options, press Enter.
1=Add 2=Change 4=Remove 5=Display 9=Start 10=End

Internet Subnet Interface Alias


Opt Address Mask Status Name

9.5.168.129 255.255.255.0 Active *NONE


9.5.168.133 255.255.255.0 Failed *NONE
9.5.168.134 255.255.255.0 Failed *NONE
10.0.0.1 255.255.255.0 Inactive POWERHA_TAKEOVER_IP
10.10.10.1 255.255.255.0 Active POWERHA
127.0.0.1 255.0.0.0 Active LOCALHOST
192.168.86.10 255.255.255.0 Active *NONE
192.168.87.10 255.255.255.0 Active *NONE

Bottom
F3=Exit F5=Refresh F6=Print list F11=Display line information
F12=Cancel F17=Top F18=Bottom
Figure 13-72 Metro Mirror: IP configuration on preferred production after switchover

330 PowerHA SystemMirror for IBM i Cookbook


Work with TCP/IP Interfaces
System: DEMOHA
Type options, press Enter.
1=Add 2=Change 4=Remove 5=Display 9=Start 10=End

Internet Subnet Interface Alias


Opt Address Mask Status Name

9.5.168.130 255.255.255.0 Active *NONE


9.5.168.133 255.255.255.0 Active *NONE
9.5.168.134 255.255.255.0 Active *NONE
10.0.0.1 255.255.255.0 Active POWERHA_TAKEOVER_IP
10.10.10.2 255.255.255.0 Active POWERHA
127.0.0.1 255.0.0.0 Active LOCALHOST
192.168.86.11 255.255.255.0 Active *NONE
192.168.87.11 255.255.255.0 Active *NONE

Bottom
F3=Exit F5=Refresh F6=Print list F11=Display line information
F12=Cancel F17=Top F18=Bottom
Figure 13-73 Metro Mirror: IP configuration on preferred backup after switchover

Figure 13-74 shows the current situation for DS volume members of the Metro Mirror
relationships. With lspprc on the current backup DS8000 75AY031, all volumes are target full
duplex, which is the normal status for target volumes. On the current production DS8000
75AY032, all volumes are full duplex, which is the normal status for source volumes.

dscli> lspprc -dev IBM.2107-75AY031 6000-61FF


Date/Time: September 29, 2011 9:50:13 AM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
=========================================================================================================
6000:6000 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6001:6001 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6002:6002 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6100:6100 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid
6101:6101 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid
6102:6102 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid

dscli> lspprc -dev IBM.2107-75AY032 6000-61FF


Date/Time: September 29, 2011 9:53:23 AM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
==================================================================================================
6000:6000 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6001:6001 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6002:6002 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6100:6100 Full Duplex - Metro Mirror 61 60 Disabled Invalid
6101:6101 Full Duplex - Metro Mirror 61 60 Disabled Invalid
6102:6102 Full Duplex - Metro Mirror 61 60 Disabled Invalid

Figure 13-74 Metro Mirror volumes status after switchover

Chapter 13. Configuring and managing DS8000 Copy Services 331


When it is scheduled to perform a switchback, all the operations detailed above are to be
done, but on the other way. Switchback and switch invoke exactly the same operations.

Using CL commands for a Global Mirror planned switchover

Note: In contrast to the GUI, the CL command will not give you any possibility to switch
over or switch back if there is one incorrect item.

Before activating the switch, we can check whether everything is fine from a clustering state
perspective and that nothing will prevent the switchover from succeeding:
1. The first items that we check are the cluster consistency and the cluster resource group
status. Use WRKCLU, then option 9 (Work with cluster resource groups) (Figure 13-75).

Work with Cluster


System: DEMOPROD
Cluster . . . . . . . . . . . . . . . . : PWRHA_CLU

Select one of the following:

1. Display cluster information


2. Display cluster configuration information

6. Work with cluster nodes


7. Work with device domains
8. Work with administrative domains
9. Work with cluster resource groups
10. Work with ASP copy descriptions

20. Dump cluster trace

Selection or command
===> 9

F1=Help F3=Exit F4=Prompt F9=Retrieve F12=Cancel

Figure 13-75 Global Mirror switchback WRKCLU command panel

332 PowerHA SystemMirror for IBM i Cookbook


As shown in Figure 13-76, “Consistent information in cluster” must be Yes, and CRG
status must be Active. To continue checking, select option 6 for Recovery domain against
the cluster resource group, in our case PWRHA_CRG2.

Work with Cluster Resource Groups

Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Create 2=Change 3=Change primary 4=Delete 5=Display
6=Recovery domain 7=Configuration objects 8=Start 9=End
20=Dump trace

Cluster Primary
Opt Resource Group Type Status Node

PWRHA_CRG1 *DEV Active DEMOPROD


6 PWRHA_CRG2 *DEV Active DEMOPROD

Bottom
Parameters for options 1, 2, 3, 8, 9 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-76 Global Mirror switch over Cluster Resource Group panel

Chapter 13. Configuring and managing DS8000 Copy Services 333


2. As shown in Figure 13-77, the recovery domain status of both nodes must be active. We
can also see their current roles. In our case, DEMOPROD is the production node and
DEMOHA is the first backup. After the switch over, DEMOHA will be the production node
and DEMOPROD the first backup node.

Work with Recovery Domain

Cluster resource group . . . . . . . . . : PWRHA_CRG2


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add node 4=Remove node 5=Display more details 20=Dump trace

Current Preferred Site


Opt Node Status Node Role Node Role Name

DEMOHA Active *BACKUP 1 *BACKUP 1 SITE2


DEMOPROD Active *PRIMARY *PRIMARY SITE1

Bottom
Parameters for options 1 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-77 Global Mirror switch version Recovery Domain panel

334 PowerHA SystemMirror for IBM i Cookbook


3. The CL command DSPCRGINF summarizes various information and statuses. In our case
we run the following command:
DSPCRGINF CLUSTER(*) CRG(PWRHA_CRG2)
As shown in Figure 13-78 and Figure 13-79 on page 336, we must find the following
information:
– Consistent information in cluster is set to Yes.
– Cluster resource group status is set to Active.
– Both Recovery Domain nodes status is set to Active.
– Both nodes roles are set to the expected values:
• Primary
• Backup

Display CRG Information

Cluster . . . . . . . . . . . . : PWRHA_CLU
Cluster resource group . . . . . : PWRHA_CRG2
Reporting node . . . . . . . . . : DEMOHA
Consistent information in cluster: Yes

Cluster resource group type . . . : *DEV


Cluster resource group status . . : Active
Previous CRG status . . . . . . . : Switchover Pending
Exit program . . . . . . . . . . . : *NONE
Library . . . . . . . . . . . . : *NONE
Exit program job name . . . . . . : *NONE
Exit program format . . . . . . . : *NONE
Exit program data . . . . . . . . : *NONE

User profile . . . . . . . . . . . : *NONE


Text . . . . . . . . . . . . . . . : DS8000 Cluster Resource Group

More...
Cluster . . . . . . . . . . . . : PWRHA_CLU
Cluster resource group . . . . . : PWRHA_CRG2
Reporting node . . . . . . . . . : DEMOHA
Consistent information in cluster: Yes

Distribute information queue . . . : *NONE


Library . . . . . . . . . . . . : *NONE
Failover message queue . . . . . . : *NONE
Library . . . . . . . . . . . . : *NONE
Failover wait time . . . . . . . . : *NOWAIT
Failover default action . . . . . : *PROCEED
CRG extended attribute . . . . . . : *NONE
Application identifier . . . . . . : *NONE
Bottom
F1=Help F3=Exit F5=Refresh F12=Cancel Enter=Continue

Figure 13-78 Global Mirror switchover DSPCRGINF command first and second panel

Chapter 13. Configuring and managing DS8000 Copy Services 335


Display CRG Information

Cluster . . . . . . . . . . . . : PWRHA_CLU
Cluster resource group . . . . . : PWRHA_CRG2
Reporting node . . . . . . . . . : DEMOHA
Consistent information in cluster: Yes

Recovery Domain Information

Current Preferred Site


Node Status Node Role Node Role Name
DEMOPROD Active *PRIMARY *PRIMARY SITE1
DEMOHA Active *BACKUP 1 *BACKUP 1 SITE2

Bottom
Number of recovery domain nodes : 2

F1=Help F3=Exit F5=Refresh F12=Cancel Enter=Continue

Figure 13-79 Global Mirror switchover DSPCRGINF command fifth panel

336 PowerHA SystemMirror for IBM i Cookbook


4. After checking the cluster resources status, if you proceed with a switchover, end all
applications using the IASP, as they will automatically be varied off by PowerHA when
switching. There are two ways to initiate an IASP switchover. As shown in Figure 13-80,
you can choose option 3 (Change primary against the cluster resource group), in our case
PWRHA_CRG2. Or you can use CHGCRGPRI CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG2). The
same command runs under the covers when selecting option 3.

Work with Cluster Resource Groups

Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Create 2=Change 3=Change primary 4=Delete 5=Display
6=Recovery domain 7=Configuration objects 8=Start 9=End
20=Dump trace

Cluster Primary
Opt Resource Group Type Status Node

PWRHA_CRG1 *DEV Active DEMOPROD


3 PWRHA_CRG2 *DEV Active DEMOPROD

Bottom
Parameters for options 1, 2, 3, 8, 9 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-80 Global Mirror switch over from Cluster Resource Group panel

5. After the switchover is complete, the following message is shown:


Cluster Resource Services API QcstInitiateSwitchOver completed
There is no progress indicator when running the command, but it performs the same steps
as described in “Using the PowerHA GUI for a Metro Mirror planned switchover” on
page 321.

Chapter 13. Configuring and managing DS8000 Copy Services 337


6. You can verify that the recovery domain nodes are in the expected active state and that
their role is not their preferred one anymore. In our case, DEMOHA is the current
production node and DEMOPROD is currently the first backup node (Figure 13-81).

Work with Recovery Domain

Cluster resource group . . . . . . . . . : PWRHA_CRG2


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add node 4=Remove node 5=Display more details 20=Dump trace

Current Preferred Site


Opt Node Status Node Role Node Role Name

DEMOHA Active *PRIMARY *BACKUP 1 SITE2


DEMOPROD Active *BACKUP 1 *PRIMARY SITE1

Bottom
Parameters for options 1 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-81 Global Mirror switchover Cluster Resource Group panel

During the switchover, on both DS8000s the following actions are submitted by PowerHA
during step 4 (Changing Recovery Domain) reported by the GUI:
a. Check the Global Copy status on the preferred source DS8000.
b. Check the Global Mirror LSS status on the preferred source DS8000.
c. Pause the Global Mirror session on the preferred source DS8000.
d. Remove volumes from the Global Mirror session on the preferred source DS8000.
e. Resume the Global Mirror session on the preferred source DS8000.
f. Fail over the Global Copy relationships on the preferred target DS8000.
g. Fast reverse restore FlashCopy relationships on the preferred target DS8000 to put
consistent data on newly Global Copy source volumes.
The persistent option is not set in the DS8000 command, which means that the
FlashCopy relationships no longer exist at the end of reverseflash.
h. Fail back Global Copy relationships on the preferred target DS8000 to synchronize
the current production volumes with the current backup volumes on the preferred
source DS8000.
i. Create the FlashCopy relationships on the current target DS8000 to build a
consistency group.
j. Update the Global Mirror session on the current source DS8000 to include the current
source volumes.
k. Start the Global Mirror session on the current source DS8000.

338 PowerHA SystemMirror for IBM i Cookbook


Using CL commands for an asymmetrical Global Mirror switchback after
a planned outage
In this section we focus on describing how to switch back to the preferred production node in
an asymmetrical Global Mirror configuration. Manual steps are needed on the DS storage
server to re-establish Global Mirror in the original direction from the preferred source to the
preferred target DS8000.

The main difference is that D volumes do not exist when compared to our scenario setup in
Figure 13-56 on page 313. Therefore, when creating the ASP copy description for the
preferred source IASP, we do not specify the consistency group volumes (Example 13-20).

Example 13-20 Source ASP copy description creation command for an asymmetrical Global Mirror
ADDASPCPYD ASPCPY(IASP2_GM1) ASPDEV(IASP2) CRG(PWRHA_CRG2) LOCATION(*DEFAULT)
SITE(SITE1) STGHOST('qlpar' 'password' ('9.5.168.32'))
LUN('IBM.2107-13ABGAA' ('0200-0201' '0300-0301') ())

For a switchover to the preferred backup site in an asymmetrical Global Mirror configuration,
PowerHA cannot reverse Global Mirror because there have been no FlashCopy consistency
group volumes configured on the preferred source DS8000. Therefore, at the end of the
switchover the Global Copy relationship volume state on the preferred backup DS8000,
which is the current production DS8000, is source suspended. They are not members of the
Global Mirror session on the preferred backup DS8000 and the FlashCopy relationships no
longer exist.

Chapter 13. Configuring and managing DS8000 Copy Services 339


After the switchover, if you do not wait for the next occurrence of the QYASSTEHCK checking
job (which checks that replication is running properly), and although all PowerHA
configuration checks (such as cluster node is active, cluster resource group is active, nodes
in recovery domain are active) show that everything is correct for running a switchback with
CHGCRGPRI, you are not allowed to do so (Figure 13-82).

Command Entry DEMOHA


Request level: 7
Previous commands and messages:
> CHGCRGPRI CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG2)
Cluster resource group exit program QYASSTEHCK in library QHASM on node
DEMOHA failed.
Error invoking exit programs on node DEMOHA.
Nodes in cluster resource group PWRHA_CRG2 at mirror copy or target site
are not eligible to become a primary node.
Cluster resource group exit program QYASSTEHCK in library QHASM on node
DEMOPROD failed.
Error invoking exit programs on node DEMOPROD.
A switch over can not be done for cluster resource group PWRHA_CRG2.
A switch over can not be done for cluster resource group PWRHA_CRG2.
Cluster Resource Services API QcstInitiateSwitchOver completed.
Primary node of cluster resource group PWRHA_CRG2 not changed.
Type command, press Enter.
===> CHGCRGPRI CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG2)

F3=Exit F4=Prompt F9=Retrieve F10=Include detailed messages


F11=Display full F12=Cancel F13=Information Assistant F24=More keys

Figure 13-82 Asymmetrical Global Mirror switchback fails

340 PowerHA SystemMirror for IBM i Cookbook


The reason for PowerHA preventing a switchback to the preferred source DS8000 in an
asymmetrical Global Mirror configuration is that we miss the consistency group volumes,
which are intended to ensure consistency within a Global Mirror relationship. Without this
consistency group volume, we cannot ensure that consistency exists, at least when source
independent ASP is varied on, or Global Copy has tracks not yet written on target. After the
command fails, the preferred production node status changes from Active to Ineligible
(Figure 13-83).

Work with Recovery Domain

Cluster resource group . . . . . . . . . : PWRHA_CRG2


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add node 4=Remove node 5=Display more details 20=Dump trace

Current Preferred Site


Opt Node Status Node Role Node Role Name

DEMOHA Active *PRIMARY *BACKUP 1 SITE2


DEMOPROD Ineligible *BACKUP 1 *PRIMARY SITE1

Bottom
Parameters for options 1 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-83 Asymmetrical Global Mirror node ineligible status

Chapter 13. Configuring and managing DS8000 Copy Services 341


Figure 13-84 shows details about the CPDBB0D message ID.

Additional Message Information

Message ID . . . . . . : CPDBB0D Severity . . . . . . . : 30


Message type . . . . . : Diagnostic
Date sent . . . . . . : 10/04/11 Time sent . . . . . . : 14:03:04

Message . . . . : Nodes in cluster resource group PWRHA_CRG2 at mirror copy


or target site are not eligible to become a primary node.
Cause . . . . . : The backup nodes in cluster resource group PWRHA_CRG2 at
mirror copy or target site are not eligible to become a primary node, for
one or more of the following reasons:
1 -- One or more auxiliary storage pools are missing.
2 -- Cross site mirroring has not been configured for auxiliary storage
pools in cluster resource group PWRHA_CRG2.
3 -- Cross site mirroring is being synchronized.
4 -- One or more mirror copies or targets of auxiliary storage pools are
not usable and offline.
5 -- One or more mirror copies or targets of auxiliary storage pools are
not usable and cross site mirroring is suspended.
6 -- Cross site mirroring is suspended for one or more mirror copies or
targets of auxiliary storage pools.
Recovery . . . : Recovery actions for each reason code are:
1 -- All auxiliary storage pools must be present. Determine the cause for
the missing ones and fix the problem.
2 -- Configure cross site mirroring for auxiliary storage pools.
3 -- Wait for the synchronization to complete.
4, 5 and 6 -- Resume cross site mirroring if it is suspended. Vary on the
auxiliary storage pool and wait for the synchronization to complete.

Press Enter to continue.

F3=Exit F6=Print F9=Display message details F12=Cancel


F21=Select assistance level
Figure 13-84 Asymmetrical Global Mirror switchback error detail

342 PowerHA SystemMirror for IBM i Cookbook


At the same time, the following message ID HAI2001 is sent to the cluster resource group or
cluster message queue (Figure 13-85):
Cross-site mirroring (XSM) for ASP IASP2 is not active.

Additional Message Information

Message ID . . . . . . : HAI2001 Severity . . . . . . . : 10


Message type . . . . . : Information
Date sent . . . . . . : 10/04/11 Time sent . . . . . . : 14:03:04

Message . . . . : Cross-site mirroring (XSM) for ASP IASP2 is not active.


Cause . . . . . : The cross-site mirroring copy state for auxiliary storage
pool (ASP) IASP2 is not active. The copy state is 12. The source
clustering node is DEMOHA and the target clustering node is DEMOPROD.
The copy states are:
-1 -- The ASP copy state is not known.
11 -- The ASP copy is being synchronized.
12 -- The ASP copy is suspended.
Recovery . . . : Recovery actions for the copy states are:
-1 -- Verify that cross site mirroring is correctly configured and correct
any configuration errors.
11 -- Wait for the synchronization to complete.
12 -- Perform a resume or reattach operation on the ASP.

Press Enter to continue.

F3=Exit F6=Print F9=Display message details F12=Cancel


F21=Select assistance level
Figure 13-85 Asymmetrical Global Mirror switchback

The message suggests that you either perform a resume or a reattach operation on the ASP.
In our case, the mirroring is not only suspended, but it is also in a failover status. This means
that we need to run a reattach operation for the ASP session. However, we also need to
establish the Global Mirror session in the proper direction again. This is the reason why in this
special case, we have to run the following operations:
1. Make sure that independent ASPs are varied off on both sides.
2. On the current production DS, perform failbackpprc. This synchronizes data from the
current production DS to the current backup DS (Example 13-21).

Example 13-21 Asymmetrical Global Mirror switch back, synchronize preferred DS with current DS
dscli> failbackpprc -dev IBM.2107-1375210 -remotedev IBM.2107-13ABGAA -type gcp 0200-0201:0200-0201
0300-0301:0300-0301
Date/Time: October 4, 2011 3:48:19 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-1375210
CMUC00197I failbackpprc: Remote Mirror and Copy pair 0200:0200 successfully failed back.
CMUC00197I failbackpprc: Remote Mirror and Copy pair 0201:0201 successfully failed back.
CMUC00197I failbackpprc: Remote Mirror and Copy pair 0300:0300 successfully failed back.
CMUC00197I failbackpprc: Remote Mirror and Copy pair 0301:0301 successfully failed back.
dscli> lspprc 0200-03FF
Date/Time: October 4, 2011 3:48:23 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-1375210
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
==================================================================================================
0200:0200 Copy Pending - Global Copy 02 60 Disabled True
0201:0201 Copy Pending - Global Copy 02 60 Disabled True

Chapter 13. Configuring and managing DS8000 Copy Services 343


0300:0300 Copy Pending - Global Copy 03 60 Disabled True
0301:0301 Copy Pending - Global Copy 03 60 Disabled True

3. On the current production node make sure that the replication is finished, as indicated by a
synchronization progress of 100 in the DSPASPSSN command panel (Figure 13-86).

Display ASP Session DEMOHA


10/04/11 15:48:52
Session . . . . . . . . . . . . . . . . . . : IASP2_GM
Type . . . . . . . . . . . . . . . . . . : *GLOBALMIR
Synchronization progress . . . . . . . . : 100

Copy Descriptions

ASP
device Name Role State Node
IASP2 IASP2_GM2 SOURCE VARYOFF DEMOHA
IASP2 IASP2_GM1 TARGET RESUMING DEMOPROD

Bottom
Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=Automatic refresh

Figure 13-86 Asymmetrical Global Mirror switchback, synchronizing back current to preferred
production DS8000

4. Make the preferred source DS8000 the source of the Global Copy replication. On the
preferred source DS8000, run the pair of failoverpprc/failbackpprc commands
(Example 13-22).

Example 13-22 Asymmetrical Global Mirror switchback, making the preferred source DS8000 the current source
dscli> lspprc 0200-03FF
Date/Time: October 4, 2011 4:02:03 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
=========================================================================================================
0200:0200 Target Copy Pending - Global Copy 02 unknown Disabled Invalid
0201:0201 Target Copy Pending - Global Copy 02 unknown Disabled Invalid
0300:0300 Target Copy Pending - Global Copy 03 unknown Disabled Invalid
0301:0301 Target Copy Pending - Global Copy 03 unknown Disabled Invalid
dscli> failoverpprc -dev IBM.2107-13ABGAA -remotedev IBM.2107-1375210 -type gcp 0200-0201:0200-0201
0300-0301:0300-0301
Date/Time: October 4, 2011 4:09:18 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
CMUC00196I failoverpprc: Remote Mirror and Copy pair 0200:0200 successfully reversed.
CMUC00196I failoverpprc: Remote Mirror and Copy pair 0201:0201 successfully reversed.
CMUC00196I failoverpprc: Remote Mirror and Copy pair 0300:0300 successfully reversed.
CMUC00196I failoverpprc: Remote Mirror and Copy pair 0301:0301 successfully reversed.
dscli> failbackpprc -dev IBM.2107-13ABGAA -remotedev IBM.2107-1375210 -type gcp 0200-0201:0200-0201
0300-0301:0300-0301
Date/Time: October 4, 2011 4:09:30 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
CMUC00197I failbackpprc: Remote Mirror and Copy pair 0200:0200 successfully failed back.

344 PowerHA SystemMirror for IBM i Cookbook


CMUC00197I failbackpprc: Remote Mirror and Copy pair 0201:0201 successfully failed back.
CMUC00197I failbackpprc: Remote Mirror and Copy pair 0300:0300 successfully failed back.
CMUC00197I failbackpprc: Remote Mirror and Copy pair 0301:0301 successfully failed back.
dscli> lspprc 0200-03FF
Date/Time: October 4, 2011 4:09:33 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
==================================================================================================
0200:0200 Copy Pending - Global Copy 02 60 Disabled True
0201:0201 Copy Pending - Global Copy 02 60 Disabled True
0300:0300 Copy Pending - Global Copy 03 60 Disabled True
0301:0301 Copy Pending - Global Copy 03 60 Disabled True

5. The Global Mirror configuration also needs to be built again on the preferred source
DS8000. Refer to the following sections:
– “Making a Global Mirror session” on page 315
The sessions should already exist. Just replace the mksession commands with
chsession ones.
– “Creating the FlashCopy relationships on the backup site” on page 314
– “Starting Global Mirror session” on page 316.

Note: You might need to remove the Global Mirror session using the rmgmir DS8000
command before being allowed to create it again.

6. Reverse IBM i cluster node roles:


a. End the cluster resource group.
b. Change the current role for the current backup node to the primary and vice versa with
CHGCRG (Example 13-23).

Example 13-23 Asymmetrical Global Mirror failback: Change CRG role command
CHGCRG CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG2) CRGTYPE(*DEV) RCYDMNACN(*CHGCUR)
RCYDMN( (DEMOPROD *PRIMARY *SAME SITE1 *SAME *SAME)
(DEMOHA *BACKUP 1 SITE2 *SAME *SAME))

c. Restart the cluster resource group.


d. Wait for the backup node status to change from Ineligible to Active. This is done by the
PowerHA healthcheck job QYASSTEHCK, which runs every hour.

After completing these steps you have re-established Global Mirror in its original direction
from the preferred source to the preferred target DS8000.

Chapter 13. Configuring and managing DS8000 Copy Services 345


13.2.2 Using CL commands for DS8000 LUN-level switching
To perform a planned LUN-level switching, follow these steps:

Note: Before starting any switchover make sure that no jobs are using the IASP anymore.

1. Make sure that “Consistent information in cluster” is Yes. Set the CRG status (in our case
PWRHA_CRG3) to Active (Figure 13-87) by using WRKCLU, then option 9 (Work with cluster
resource groups).

Work with Cluster Resource Groups

Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Create 2=Change 3=Change primary 4=Delete 5=Display
6=Recovery domain 7=Configuration objects 8=Start 9=End
20=Dump trace

Cluster Primary
Opt Resource Group Type Status Node

PWRHA_CRG3 *DEV Active DEMOFC

Bottom
Parameters for options 1, 2, 3, 8, 9 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-87 Cluster Resource Group status

346 PowerHA SystemMirror for IBM i Cookbook


2. Using option 6 (Recovery Domain) on the Work with Cluster Resource Groups panel,
check that the recovery domain status of both nodes (DEMOFC as the production node
and DEMOHA as the backup node) are active and that their roles are the ones that you
expect (Figure 13-88).

Work with Recovery Domain

Cluster resource group . . . . . . . . . : PWRHA_CRG3


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add node 4=Remove node 5=Display more details 20=Dump trace

Current Preferred Site


Opt Node Status Node Role Node Role Name

DEMOFC Active *PRIMARY *PRIMARY SITE1


DEMOHA Active *BACKUP 1 *BACKUP 1 SITE1

Bottom
Parameters for options 1 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-88 Node status in recovery domain

Alternatively, use DSPCRGINF to summarize the information and status of the CRG.
3. Use CHGCRGPRI to switch the IASP from the primary node to the backup node
(Example 13-24).

Example 13-24 CHGCRGPRI command for switching the node


CHGCRGPRI CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG3)

Chapter 13. Configuring and managing DS8000 Copy Services 347


Also, you can switch the IASP by using option 3 (Change primary) on the Work with
Cluster Resource Group panel (Figure 13-87 on page 346).
While switching with the CL command there is no process indicator. Use DSPASPSTS to
check the status of IASP on another session from the production and backup nodes
(Figure 13-89).

Display ASP Vary Status

ASP Device . . . . : IASP_LUN Step . . . . . . . : 30 / 34


ASP Number . . . . : 71 Current time . . . : 00:00:07
ASP State . . . . : ACTIVE Previous time . . : 00:01:47

Step Elapsed time


UID/GID mismatch correction 00:00:00
Database access path recovery 00:00:01
> Database cross-reference file merge 00:00:01
SPOOL initialization
Image catalog synchronization
Command analyzer recovery
Catalog validation

Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=End automatic refresh


Figure 13-89 IASP status on the backup node during switching

348 PowerHA SystemMirror for IBM i Cookbook


4. After the switchover is complete, the following message is shown on the bottom of panel
and you can also see the change of the primary node for CRG PWRHA_CR3 from
DEMOFC to DEMOHA on the Work with Cluster Resource Groups panel (Figure 13-90)
Cluster Resource Services API QcstInitiateSwitchOver completed

Work with Cluster Resource Groups

Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Create 2=Change 3=Change primary 4=Delete 5=Display
6=Recovery domain 7=Configuration objects 8=Start 9=End
20=Dump trace

Cluster Primary
Opt Resource Group Type Status Node

PWRHA_CRG3 *DEV Active DEMOHA

Bottom
Parameters for options 1, 2, 3, 8, 9 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu
Cluster Resource Services API QcstInitiateSwitchOver completed. +

Figure 13-90 CRG status after switching

You can verify that the Recovery Domain nodes are in the expected active status and that
their role is the preferred backup status. In our case, DEMOHA is the production node and
DEMOFC is the backup node (Figure 13-91).

Work with Recovery Domain

Cluster resource group . . . . . . . . . : PWRHA_CRG3


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add node 4=Remove node 5=Display more details 20=Dump trace

Current Preferred Site


Opt Node Status Node Role Node Role Name

DEMOFC Active *BACKUP 1 *PRIMARY SITE1


DEMOHA Active *PRIMARY *BACKUP 1 SITE1

Bottom
Parameters for options 1 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-91 Node status in recovery domain after switching

Chapter 13. Configuring and managing DS8000 Copy Services 349


On DS8000 you can also verify that the host connection for the backup node is
automatically assigned to the volume group that was assigned to the production node
(Example 13-25).

Example 13-25 Host connection on DS8000 after switching


dscli> lshostconnect 2 5
Date/Time: September 28, 2011 4:31:36 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
Name ID WWPN HostType Profile portgrp volgrpID ESSIOport
=============================================================================================
demofc_powerha 0002 10000000C9523E9D iSeries IBM iSeries - OS/400 0 - all
DEMOHA_FC 0005 10000000C94122A2 iSeries IBM iSeries - OS/400 0 V48 all

5. Verify whether the IASP has been varied on (that is, that it is in available status) at the
backup node using the following command:
WRKCFGSTS CFGTYPE(*DEV) CFGD(*ASP)
If the configuration object online parameter in the device CRG is set to *ONLINE,
PowerHA automatically varies on the IASP on node at a switch over.

13.2.3 Failing over and back for an unplanned outage


In this section we describe the actions being performed for an unplanned failover and switchback
in a Metro Mirror or Global Mirror environment and a LUN-level switching environment.

Metro Mirror and Global Mirror failover and failback


Regarding unplanned outages, we distinguish three cases that apply to both Metro Mirror and
Global Mirror configurations:
 A failover for a primary node failure with a panic message
In this case, the production node is able to send a specific message to the backup node
stating that the production node is being terminated in an abnormal way.
 A failover for a primary node without the panic message
In this case, the backup node might receive primary node status information from the
HMC or VIOS.
 A primary node partition condition
In this case, the backup node does not find any way to determine the primary node status.

350 PowerHA SystemMirror for IBM i Cookbook


Failover events and actions you will take for each of these cases are detailed here:
 To simulate a primary node failure with a panic message, we force the primary node to go
into a restricted stated using ENDSBS SBS(*ALL) OPTION(*IMMED). (Using PWRDWNSYS would
lead to the same results.) The primary node will be able to send a panic message to the
backup node. After receiving this message, if the backup node is again able to establish a
heartbeat confirmation, the backup node invokes the automatic failover operations.
As shown in Figure 13-92, an inquiry the following message is logged in the defined
cluster message queue on the backup node (in our case QSYSOPR):
CPABB02 “Cluster resource groups are failing over to node DEMOHA. (G C)”
By using the cluster FLVDFTACN parameter default setting *PROCEED, the failover
continues automatically after expiration of the cluster failover time (the FLVWAITTIM
parameter, in our case *NOMAX). Also, the following informational message is logged,
indicating the name of the CRG that is being failed over:
CPIBB18 “Cluster resource group PWRHA_CRG1 is failing over from node DEMOPROD
to node DEMO.”
Replying G for go initiates the failover procedure.

Display Messages
System: DEMOHA
Queue . . . . . : QSYSOPR Program . . . . : *DSPMSG
Library . . . : QSYS Library . . . :
Severity . . . : 99 Delivery . . . : *HOLD

Type reply (if required), press Enter.


Cluster resource group PWRHA_CRG1 is failing over from node DEMOPROD to
node DEMOHA.
Cluster resource groups are failing over to node DEMOHA. (G C)
Reply . . .

Bottom
F3=Exit F11=Remove a message F12=Cancel
F13=Remove all F16=Remove all except unanswered F24=More keys

Figure 13-92 Failover inquiry message in the cluster message queue

Note: If using the cluster default parameter settings FLVDFTACT=*PROCEED and


FLVWAITTIM=*NOWAIT, an automatic failover occurs instantly. Whereas, when
using FLVWAITTIM=*NOMAX, the cluster waits forever for the user to reply on the
cluster message queue’s CPABB02 inquiry message to cancel or proceed with the
automatic failover.

Chapter 13. Configuring and managing DS8000 Copy Services 351


Metro Mirror (or Global Mirror) IASP volume relationships are failed over to the secondary.
That is, the Metro Mirror (or Global Mirror) target volumes become suspended source
volumes (Figure 13-93). In our case, on backup DS8000 75AY032, volumes 6000 - 6002
and 6100 - 6102 are suspended.

dscli> lspprc -dev IBM.2107-75AY032 6000-61FF


Date/Time: September 30, 2011 9:38:57 AM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
=====================================================================================================
6000:6000 Suspended Host Source Metro Mirror 60 60 Disabled Invalid
6001:6001 Suspended Host Source Metro Mirror 60 60 Disabled Invalid
6002:6002 Suspended Host Source Metro Mirror 60 60 Disabled Invalid
6100:6100 Suspended Host Source Metro Mirror 61 60 Disabled Invalid
6101:6101 Suspended Host Source Metro Mirror 61 60 Disabled Invalid
6102:6102 Suspended Host Source Metro Mirror 61 60 Disabled Invalid

Figure 13-93 Suspended source volumes

The IASP (in our case IASP1) is varied on for the node at the secondary site in our
example, as we used the CRG configuration object *ONLINE setting. Also, the takeover IP
address is started on the secondary site.
Cluster node roles are changed so that the node at the secondary site becomes the
primary and the node at the failed primary site becomes the backup (Figure 13-94), which
is the recovery domain for our cluster resource group PWRHA_CRG1. In our case,
DEMOHA becomes the primary node and DEMOPROD becomes the first backup node
with an inactive status.

Work with Recovery Domain

Cluster resource group . . . . . . . . . : PWRHA_CRG1


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add node 4=Remove node 5=Display more details 20=Dump trace

Current Preferred Site


Opt Node Status Node Role Node Role Name

DEMOHA Active *PRIMARY *BACKUP 1 SITE2


DEMOPROD Inactive *BACKUP 1 *PRIMARY SITE1

Bottom
Parameters for options 1 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-94 Metro Mirror failover recovery domain

352 PowerHA SystemMirror for IBM i Cookbook


When we are ready to fail back, the following steps needs to be done:
a. Make sure that the IASP on the preferred primary node is varied off.
b. Make sure that the cluster node on the preferred primary node is started. If it is not, use
STRCLUNOD. In our case:
STRCLUNOD CLUSTER(PWRHA_CLU) NODE(DEMOPROD)
The status of the preferred primary node (in our case DEMOPROD) within the recovery
domain becomes Ineligible (Figure 13-95).

Work with Recovery Domain

Cluster resource group . . . . . . . . . : PWRHA_CRG1


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add node 4=Remove node 5=Display more details 20=Dump trace

Current Preferred Site


Opt Node Status Node Role Node Role Name

DEMOHA Active *PRIMARY *BACKUP 1 SITE2


DEMOPROD Ineligible *BACKUP 1 *PRIMARY SITE1

Bottom
Parameters for options 1 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-95 Metro Mirror failback recovery domain

Chapter 13. Configuring and managing DS8000 Copy Services 353


c. As shown in Figure 13-96, the current status of the ASP session (in our case
IASP1_MM) can be displayed with DSPASPSSN. The replication status is suspended and
the preferred primary role is detached.

Display ASP Session DEMOPROD


09/30/11 10:47:41
Session . . . . . . . . . . . . . . . . . . : IASP1_MM
Type . . . . . . . . . . . . . . . . . . : *METROMIR

Copy Descriptions

ASP
device Name Role State Node
IASP1 IASP1_MM2 SOURCE UNKNOWN DEMOHA
IASP1 IASP1_MM1 DETACHED SUSPENDED DEMOPROD

Bottom
Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=Automatic refresh

Figure 13-96 Metro Mirror failback session status

d. Establish back the replication from the preferred backup to the preferred production DS.
i. For Metro Mirror or symmetrical Global Mirror, re-attach the detached ASP session
on the preferred primary node. That is, start the Metro Mirror (or symmetrical Global
Mirror) replication from the preferred secondary DS8000 to the preferred primary
DS8000, with the following command:
CHGASPSSN SSN(IASP1_MM) OPTION(*REATTACH) ASPCPY((IASP1_MM1 IASP1_MM2))
By default, this command requires an answer to the following message in the
QSYSOPR message queue:
HAA2000 “Reattach of ASP session IASP1_MM was requested. (C G)”

354 PowerHA SystemMirror for IBM i Cookbook


The second level of the message provides all the necessary information
(Figure 13-97).

Additional Message Information

Message ID . . . . . . : HAA2000 Severity . . . . . . . : 99


Message type . . . . . : Inquiry
Date sent . . . . . . : 09/30/11 Time sent . . . . . . : 11:03:16

Message . . . . : Reattach of ASP session IASP1_MM was requested. (C G)


Cause . . . . . : A request was made to reattach auxiliary storage pool
(ASP) session IASP1_MM. This operation will start replication from cluster
node DEMOHA to cluster node DEMOPROD.
Recovery . . . : Do one of the following:
-- Type G to continue the reattach.
-- Type C to cancel the reattach.
Possible choices for replying to message . . . . . . . . . . . . . . . :
C -- Processing is terminated.
G -- Processing is continued.

Bottom
Type reply below, then press Enter.
Reply . . . .

F3=Exit F6=Print F9=Display message details F12=Cancel


F21=Select assistance level

Figure 13-97 Metro Mirror failback re-attach confirmation

ii. For asymmetrical Global Mirror, start the replication from the preferred secondary
DS8000 to the preferred primary DS8000. This must be done with failbackpprc on
the preferred secondary DS8000 (Example 13-21 on page 343).

Chapter 13. Configuring and managing DS8000 Copy Services 355


e. After establishing back the replication from the preferred backup to the preferred
production DS, we can see the ASP session status on the preferred backup node with
DSPASPSSN as show in Figure 13-98 for Metro Mirror or as shown in Figure 13-99 for
Global Mirror. We can compare with the PPRC relationships on both the preferred
backup DS8000 and the preferred production DS8000 as shown in Figure 13-100 on
page 357 for Metro Mirror and as shown in Figure 13-101 on page 357 for Global
Mirror.

Display ASP Session DEMOHA


09/30/11 11:10:45
Session . . . . . . . . . . . . . . . . . . : IASP1_MM
Type . . . . . . . . . . . . . . . . . . : *METROMIR

Copy Descriptions

ASP
device Name Role State Node
IASP1 IASP1_MM2 SOURCE AVAILABLE DEMOHA
IASP1 IASP1_MM1 TARGET ACTIVE DEMOPROD

Bottom
Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=Automatic refresh

Figure 13-98 Metro Mirror failback ASP session status

Display ASP Session DEMOHA


10/05/11 10:32:42
Session . . . . . . . . . . . . . . . . . . : IASP2_GM
Type . . . . . . . . . . . . . . . . . . : *GLOBALMIR
Synchronization progress . . . . . . . . : 100

Copy Descriptions

ASP
device Name Role State Node
IASP2 IASP2_GM2 SOURCE AVAILABLE DEMOHA
IASP2 IASP2_GM1 TARGET RESUMING DEMOPROD

Bottom
Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=Automatic refresh

Figure 13-99 Global Mirror failback ASP session status

356 PowerHA SystemMirror for IBM i Cookbook


dscli> lspprc -dev IBM.2107-75AY032 6000-61FF
Date/Time: September 30, 2011 11:16:55 AM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
==================================================================================================
6000:6000 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6001:6001 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6002:6002 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6100:6100 Full Duplex - Metro Mirror 61 60 Disabled Invalid
6101:6101 Full Duplex - Metro Mirror 61 60 Disabled Invalid
6102:6102 Full Duplex - Metro Mirror 61 60 Disabled Invalid

dscli> lspprc -dev IBM.2107-75AY031 6000-61FF


Date/Time: September 30, 2011 11:17:00 AM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
=========================================================================================================
6000:6000 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6001:6001 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6002:6002 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6100:6100 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid
6101:6101 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid
6102:6102 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid

Figure 13-100 Metro Mirror failback PPRC relationships status

dscli> lspprc -dev IBM.2107-1375210 0200-0300


Date/Time: October 5, 2011 10:37:29 AM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-1375210
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
==================================================================================================
0200:0200 Copy Pending - Global Copy 02 60 Disabled True
0201:0201 Copy Pending - Global Copy 02 60 Disabled True
0300:0300 Copy Pending - Global Copy 03 60 Disabled True
0301:0301 Copy Pending - Global Copy 03 60 Disabled True
dscli> lspprc -dev IBM.2107-13ABGAA 0200-0300
Date/Time: October 5, 2011 10:38:15 AM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-13ABGAA
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
=========================================================================================================
0200:0200 Target Copy Pending - Global Copy 02 unknown Disabled Invalid
0201:0201 Target Copy Pending - Global Copy 02 unknown Disabled Invalid
0300:0300 Target Copy Pending - Global Copy 03 unknown Disabled Invalid
0301:0301 Target Copy Pending - Global Copy 03 unknown Disabled Invalid

Figure 13-101 Global Mirror failback PPRC relationships status

Now we can perform a switchback to the preferred primary node, as described in section
13.2.1, “Switchover and switchback for a Metro Mirror or Global Mirror planned outage” on
page 320.
 To simulate a primary node failure without a panic message, we force an immediate
power off from the HMC on primary node partition. This action simulates a server failure.
In this case, the primary node is not able to send a panic message to the backup node.

Chapter 13. Configuring and managing DS8000 Copy Services 357


If using the advanced node failure detection feature, as we do in our example f(or which
the configuration steps are described in 11.2, “Setting up cluster monitors” on page 208),
the HMC CIMOM server is able to send information about partition failures to the
registered IBM i cluster nodes. The preferred production node status becomes failed
(Figure 13-102). (DEMOPROD is the preferred production node in our example.)
Therefore, the backup node is able to invoke automatic failover operations just like our first
case above. It sends to the cluster message queue the same inquiry message CPABB02,
with the same considerations about wait time, then it performs the same failover
operations.

Work with Cluster Nodes

Local node . . . . . . . . . . . . . . . : DEMOHA


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add 2=Change 4=Remove 5=Display more details 6=Work with monitors
8=Start 9=End 20=Dump trace

Opt Node Status Device Domain

DEMOFC Active PWRHA_DMN


DEMOHA Active PWRHA_DMN
DEMOPROD Failed PWRHA_DMN

Bottom
Parameters for options 1, 2, 9 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve
F11=Order by status F12=Cancel F13=Work with cluster menu

Figure 13-102 Metro Mirror failover preferred production node status

After bringing back the preferred production node, he steps to fail back are the same as in
our first case above.
If you are not using the advanced node failure detection feature, the backup node is
unable to know the real production node status, and it puts the production node in the
partition status within the recovery domain. We describe this behavior in the third case,
below.
 To simulate a cluster partition condition, we break the heartbeat IP connection by ending
the heartbeat IP interface on the primary node.
If the issue is either a real production node issue, or, for example, a network issue that
prevents any user from connecting to the production node, you might decide to perform
a failover.
The node status for the primary cluster node needs to be changed from the partition to
failed with the following command:
CHGCLUNODE CLUSTER(PWRHA_CLU) NODE(DEMOPROD) OPTION(*CHGSTS)

358 PowerHA SystemMirror for IBM i Cookbook


The command handles only these failover events:
– It changes the cluster node status from partition to failed and the cluster resource
group to inactive, and it changes the cluster resource group node role (Figure 13-103).

Work with Recovery Domain

Cluster resource group . . . . . . . . . : PWRHA_CRG1


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add node 4=Remove node 5=Display more details 20=Dump trace

Current Preferred Site


Opt Node Status Node Role Node Role Name

DEMOHA Active *PRIMARY *BACKUP 1 SITE2


DEMOPROD Inactive *BACKUP 1 *PRIMARY SITE1

Bottom
Parameters for options 1 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-103 Metro Mirror failback: Partition status changed to inactive

– It performs the failover actions on the DS8000 (Figure 13-104).

dscli> lspprc -dev IBM.2107-75AY032 6000-61FF


Date/Time: September 30, 2011 2:34:39 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
=====================================================================================================
6000:6000 Suspended Host Source Metro Mirror 60 60 Disabled Invalid
6001:6001 Suspended Host Source Metro Mirror 60 60 Disabled Invalid
6002:6002 Suspended Host Source Metro Mirror 60 60 Disabled Invalid
6100:6100 Suspended Host Source Metro Mirror 61 60 Disabled Invalid
6101:6101 Suspended Host Source Metro Mirror 61 60 Disabled Invalid
6102:6102 Suspended Host Source Metro Mirror 61 60 Disabled Invalid

Figure 13-104 Metro Mirror failback: Preferred backup volumes status

The remaining actions, such as varying on the independent ASP and starting the takeover
IP address, must be done manually.
After recovery of the preferred primary node, the steps to fail back are the same as for our
first case above.

Chapter 13. Configuring and managing DS8000 Copy Services 359


When heartbeat communication is no longer possible, the backup node sets both the
production cluster node (Figure 13-105) and the cluster resource group (Figure 13-106 on
page 361) status to partition, and sends the following informational message to the
cluster message queue (Figure 13-107 on page 361):
CPFBB4F “Automatic fail over not started for cluster resource group PWRHA_CRG1
in cluster PWRHA_CLU. ”

Work with Cluster Nodes

Local node . . . . . . . . . . . . . . . : DEMOHA


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add 2=Change 4=Remove 5=Display more details 6=Work with monitors
8=Start 9=End 20=Dump trace

Opt Node Status Device Domain

DEMOFC Active
DEMOHA Active PWRHA_DMN
DEMOPROD Partition PWRHA_DMN

Bottom
Parameters for options 1, 2, 9 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve
F11=Order by status F12=Cancel F13=Work with cluster menu

Figure 13-105 Metro Mirror failback cluster node partition status

360 PowerHA SystemMirror for IBM i Cookbook


Work with Recovery Domain

Cluster resource group . . . . . . . . . : PWRHA_CRG1


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add node 4=Remove node 5=Display more details 20=Dump trace

Current Preferred Site


Opt Node Status Node Role Node Role Name

DEMOHA Active *BACKUP 1 *BACKUP 1 SITE2


DEMOPROD Partition *PRIMARY *PRIMARY SITE1

Bottom
Parameters for options 1 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-106 Metro Mirror failback CRG partition status

Additional Message Information

Message ID . . . . . . : CPFBB4F Severity . . . . . . . : 40


Message type . . . . . : Information
Date sent . . . . . . : 09/30/11 Time sent . . . . . . : 13:46:09

Message . . . . : Automatic fail over not started for cluster resource group
PWRHA_CRG1 in cluster PWRHA_CLU.
Cause . . . . . : A failure has occurred on primary node DEMOPROD or with
the job associated with cluster resource group PWRHA_CRG1 on node DEMOPROD.
Cluster Resource Services is unable to determine the state of the resources
being managed by cluster resource group PWRHA_CRG1 and is unable to perform
automatic fail over. The type of failure is reason code 1. Possible reason
codes are:
1 -- The cluster has become partitioned because the failure appears to be
a communication failure.
2 -- Either there is no backup node defined in the cluster resource group
or all backup nodes are not active. If this is a resilient device cluster
More...
Press Enter to continue.

F3=Exit F6=Print F9=Display message details F12=Cancel


F21=Select assistance level

Figure 13-107 Metro Mirror failback Heartbeat lost informational message

If the issue is really only about heartbeat communication and does not impact production
activity, and if it can be fixed quickly, after it is fixed, after a 15-minute interval (the default
value), the cluster status will automatically change back to active.

Chapter 13. Configuring and managing DS8000 Copy Services 361


DS8000 LUN-level switching failover and back
A system failure or other major outage at the production node might require an unplanned
switch of an IASP. This is handled in the same way as a planned switch in 13.2.2, “Using CL
commands for DS8000 LUN-level switching” on page 346. However, when the IASP is varied
on, there are added delay factors due to the same abnormal IPL considerations for rebuilding
database access paths that are encountered during a system IPL. Consider using
systems-managed access-path protection (SMAPP) and setting it to the shortest rebuild time
possible. For more information, see in 15.5, “Best practices for reducing IASP vary on times”
on page 409.

13.2.4 Detaching and reattaching a remote copy ASP session


If you want to get access to the IASP on your backup node, you have to detach the
IASP session associated with that IASP first. Reasons to do this might be short-term
testing on the backup node or major application changes on the production node that you
do not want to propagate to the backup system before doing final testing on the production
side. This scenario applies to Metro Mirror and both asymmetrical and symmetrical Global
Mirror configurations.

Note: To get a consistent status of your application data with DS8000 Remote Copy
Services, the IASP needs to be varied off before being detached.

Caution: While the ASP session is detached, it is not possible to perform a switch.

Before starting the operations, make sure that everything is correct for the cluster, cluster
resource groups, and recovery domain, as described in previous sections:
1. Vary off the independent ASP (in our case IASP1) on the production node using the
following command:
VRYCFG CFGOBJ(IASP1) CFGTYPE(*DEV) STATUS(*OFF)
2. Detach the ASP session, using the following command:
CHGASPSSN SSN(IASP1_MM) OPTION(*DETACH)

Note: The DETACH option is required to run on the backup node for a Global Mirror
configuration and on the production node for a Metro Mirror configuration.

362 PowerHA SystemMirror for IBM i Cookbook


The result of detaching an ASP session is that both source and target volumes are
suspended source so that volumes are available for host I/O (Figure 13-108).

dscli> lspprc -dev IBM.2107-75AY031 6000-61FF


Date/Time: September 30, 2011 3:34:12 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
=====================================================================================================
6000:6000 Suspended Host Source Metro Mirror 60 60 Disabled Invalid
6001:6001 Suspended Host Source Metro Mirror 60 60 Disabled Invalid
6002:6002 Suspended Host Source Metro Mirror 60 60 Disabled Invalid
6100:6100 Suspended Host Source Metro Mirror 61 60 Disabled Invalid
6101:6101 Suspended Host Source Metro Mirror 61 60 Disabled Invalid
6102:6102 Suspended Host Source Metro Mirror 61 60 Disabled Invalid

dscli> lspprc -dev IBM.2107-75AY032 6000-61FF


Date/Time: September 30, 2011 3:34:17 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
=====================================================================================================
6000:6000 Suspended Host Source Metro Mirror 60 60 Disabled Invalid
6001:6001 Suspended Host Source Metro Mirror 60 60 Disabled Invalid
6002:6002 Suspended Host Source Metro Mirror 60 60 Disabled Invalid
6100:6100 Suspended Host Source Metro Mirror 61 60 Disabled Invalid
6101:6101 Suspended Host Source Metro Mirror 61 60 Disabled Invalid
6102:6102 Suspended Host Source Metro Mirror 61 60 Disabled Invalid

Figure 13-108 PPRC relationship status after detaching session

3. Vary on the independent ASP (in our case IASP1) on the production node using the
following command:
VRYCFG CFGOBJ(IASP1) CFGTYPE(*DEV) STATUS(*ON)

Chapter 13. Configuring and managing DS8000 Copy Services 363


4. After the detach of the ASP session you can also vary on the independent ASP on the
backup node. As expected and shown on Figure 13-109, the DEMOHA node is in
ineligible status, meaning that it cannot be used for switchovers or failovers.

Work with Recovery Domain

Cluster resource group . . . . . . . . . : PWRHA_CRG1


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add node 4=Remove node 5=Display more details 20=Dump trace

Current Preferred Site


Opt Node Status Node Role Node Role Name

DEMOHA Ineligible *BACKUP 1 *BACKUP 1 SITE2


DEMOPROD Active *PRIMARY *PRIMARY SITE1

Bottom
Parameters for options 1 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-109 Backup mode ineligible

5. When you are ready to re-establish replication in the original direction from the preferred
primary to the preferred backup node, the next step is to vary off the independent ASP
on the backup node and re-attach the ASP session on the backup node using the
following command:
CHGASPSSN SSN(IASP1_MM) OPTION(*REATTACH)

364 PowerHA SystemMirror for IBM i Cookbook


Proper PPRC relationships are restored from the production node to the backup node
(Figure 13-110), and the recovery domain is active again (Figure 13-111).

dscli> lspprc -dev IBM.2107-75AY031 6000-61FF


Date/Time: September 30, 2011 4:00:02 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
==================================================================================================
6000:6000 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6001:6001 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6002:6002 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6100:6100 Full Duplex - Metro Mirror 61 60 Disabled Invalid
6101:6101 Full Duplex - Metro Mirror 61 60 Disabled Invalid
6102:6102 Full Duplex - Metro Mirror 61 60 Disabled Invalid
dscli> lspprc -dev IBM.2107-75AY032 6000-61FF
Date/Time: September 30, 2011 4:00:05 PM CDT IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
=========================================================================================================
6000:6000 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6001:6001 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6002:6002 Target Full Duplex - Metro Mirror 60 unknown Disabled Invalid
6100:6100 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid
6101:6101 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid
6102:6102 Target Full Duplex - Metro Mirror 61 unknown Disabled Invalid

Figure 13-110 PPRC relationships after re-attach

Work with Recovery Domain

Cluster resource group . . . . . . . . . : PWRHA_CRG1


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add node 4=Remove node 5=Display more details 20=Dump trace

Current Preferred Site


Opt Node Status Node Role Node Role Name

DEMOHA Active *BACKUP 1 *BACKUP 1 SITE2


DEMOPROD Active *PRIMARY *PRIMARY SITE1

Bottom
Parameters for options 1 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu

Figure 13-111 Recovery Domain active back

In the example above, the REATTACH option is run on the backup node, which means that
production-independent ASP data will copied to the backup independent ASP. This direction
is still consistent with the cluster resource group status.

Chapter 13. Configuring and managing DS8000 Copy Services 365


Switching while detached
For specific cases, you might need to copy current backup independent ASP data to the
current and preferred production independent ASP and perform a role swap, leading the
currently backup node to become the production node. To perform such actions, run the
following steps:
1. Make sure that independent ASPs are varied off on both sides.
2. Change the role of each cluster resource group node with CHGCRG (Example 13-26). The
node that was previously the primary one becomes the backup one, and the node that was
previously the backup one becomes the production one. This command requires the
cluster resource group to be inactive.

Example 13-26 Changing cluster resource group node role


CHGCRG CLUSTER(PWRHA_CLU) CRG(PWRHA_CRG1) CRGTYPE(*DEV) RCYDMNACN(*CHGCUR)
RCYDMN( (DEMOHA *PRIMARY *SAME *SAME *SAME *SAME)
(DEMOPROD *BACKUP 1 *SAME *SAME *SAME))

3. You can now run the REATTACH option in the proper way, which is from the current
production site to the current backup site (that is, on the current backup node)
(Example 13-27) with CHGASPSSN.

Note: Do not use CHGASPSSN for asymmetrical Global Mirror. Instead use the DS8000
failbackpprc command. Refer to specific operations for asymmetrical Global Mirror.

Example 13-27 Reattach ASP session


CHGASPSSN SSN(IASP1_MM) OPTION(*REATTACH)

13.2.5 Managing FlashCopy


In this section we describe the management of DS8000 FlashCopy with IBM PowerHA
SystemMirror for i including FlashCopy reverse, incremental FlashCopy, and detaching and
reattaching a FlashCopy ASP session.

Reversing FlashCopy

Note: To make the FlashCopy reversible you need to start the session with the options
FLASHTYPE(*COPY) and PERSISTENT(*YES).

Reverse operation of the FlashCopy causes the source volume data to be overwritten by
target volume data. This operation might be helpful if you want to remove the changes done
to the source IASP since the FlashCopy was started. In this example we use the same
environment as in “Environment overview” on page 295. The only changes we made was
removing system names in the ASP copy description, so the IASP target copy will not be
accessed by any other system. To do this we used the following command:
CHGASPCPYD ASPCPY(ASPCPYD2) LOCATION(*NONE)

In addition, we used target disks not included in any volume group on the D8000.

366 PowerHA SystemMirror for IBM i Cookbook


Now we need to start the FlashCopy session in the normal way:
1. Vary off the IASP or quiesce the database activity for the IASP on the production system.
2. Start the FlashCopy session with the following command:
STRASPSSN SSN(SESFC1) TYPE(*FLASHCOPY) ASPCPY((ASPCPYD1 ASPCPYD2))
FLASHTYPE(*COPY) PERSISTENT(*YES)
Figure 13-112 show the status of the session.

Display ASP Session DEMOHA


10/05/11 18:16:12
Session . . . . . . . . . . . . . . . . . . : SESFC1
Type . . . . . . . . . . . . . . . . . . : *FLASHCOPY
Persistent . . . . . . . . . . . . . . . : *NO
FlashCopy type . . . . . . . . . . . . . : *COPY
Number sectors copied . . . . . . . . . . : 3450496
Number sectors remaining to be copied . . : 63658368

Copy Descriptions

ASP
device Name Role State Node
IASP3 ASPCPYD1 SOURCE AVAILABLE DEMOHA
IASP3 ASPCPYD2 TARGET UNKNOWN *NONE

Bottom
Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=Automatic refresh

Figure 13-112 ASP session for FlashCopy reverse

3. Reverse the FlashCopy session.


When you want to remove the changes done to the source IASP3 and bring it to the state
that it was in when the FlashCopy was started, take the following steps:
a. Vary off the IASP3.
b. On the source system use the following command:
CHGASPSSN SSN(SESFC1) OPTION(*REVERSE).
c. Vary on the IASP3.
After this procedure is done you should see in the IASP3 the data that was there when the
FlashCopy was started. When the FlashCopy session finishes copying the data in the
reverse direction you might want to end the FlashCopy or reverse it back so that you have
another point-in-time copy for the recovery point.

Note: It is impossible to reverse a FlashCopy when its source is in a Metro Mirror or a


Global Mirror relationship.

Chapter 13. Configuring and managing DS8000 Copy Services 367


Figure 13-113 shows FlashCopy after the reverse.

Display ASP Session DEMOHA


10/05/11 22:18:05
Session . . . . . . . . . . . . . . . . . . : SESFC1
Type . . . . . . . . . . . . . . . . . . : *FLASHCOPY
Persistent . . . . . . . . . . . . . . . : *YES
FlashCopy type . . . . . . . . . . . . . : *COPY
Number sectors copied . . . . . . . . . . : 67108864
Number sectors remaining to be copied . . : 0

Copy Descriptions

ASP
device Name Role State Node
IASP3 ASPCPYD2 SOURCE UNKNOWN *NONE
IASP3 ASPCPYD1 TARGET AVAILABLE DEMOHA

Bottom
Press Enter to continue

F3=Exit F5=Refresh F12=Cancel F19=Automatic refresh

Figure 13-113 FlashCopy after reverse

Incremental FlashCopy
When you have a FlashCopy target that is used for a longer period of time (for example, as a
source for your data warehouse), you might want to have the FlashCopy session active all the
time and do the incremental updates of the target IASP. To do this you need to configure the
ASP copy descriptions pointing to your source and target volumes and allowing the IASPs to
be used by the source and target nodes. The ASP copy descriptions that we use for this
example is the same as in “Environment overview” on page 295. To do a incremental
FlashCopy we need to take the following steps:
1. Quiesce the source IASP with the following command:
CHGASPACT ASPDEV(IASP3) OPTION(*SUSPEND) SSPTIMO(30)
2. Start the FlashCopy session with the PERSIST(*YES) option, as we do here:
STRASPSSN SSN(SESFC1) TYPE(*FLASHCOPY) ASPCPY((ASPCPYD1 ASPCPYD2))
FLASHTYPE(*NOCOPY) PERSISTENT(*YES)
3. Resume the source IASP operation with the following command:
CHGASPACT ASPDEV(IASP3) OPTION(*RESUME)
4. Vary on the target IASP on the target node (DEMOFC) and use it.

Any time that you want to update the content of the target IASP take the following steps:
1. Quiesce the source IASP with the following command:
CHGASPACT ASPDEV(IASP3) OPTION(*SUSPEND) SSPTIMO(30)
2. Do the FlashCopy increment operation with the following command:
CHGASPSSN SSN(SESFC1) OPTION(*INCR)

368 PowerHA SystemMirror for IBM i Cookbook


3. Resume the source IASP operation with the following command:
CHGASPACT ASPDEV(IASP3) OPTION(*RESUME).
4. Vary on the target IASP on the target node (DEMOFC) and use it.

Detaching and reattaching FlashCopy


To detach the FlashCopy session you need to run following commands on the target system:
VRYCFG CFGOBJ(IASP3) CFGTYPE(*DEV) STATUS(*OFF)
CHGASPSSN SSN(SESFC1) OPTION(*DETACH) TRACK(*YES)

Specifying TRACK(*YES) enables the DS8000 to track the changes that are done to the
source volumes, and allows the reattach operation to be done faster. When the session is
detached, the target volumes are not usable. In this state you can do the system maintenance
of the target. In the DS8000, the FlashCopy session is not present.

When you want to bring the session back, you need to issue the following command:
CHGASPSSN SSN(SESFC1) OPTION(*REATTACH)

Chapter 13. Configuring and managing DS8000 Copy Services 369


370 PowerHA SystemMirror for IBM i Cookbook
14

Chapter 14. Configuring and managing


CSVC/V7000 Copy Services
In this chapter we provide you with step-by-step instructions for setting up a SVC/V7000
environment using PowerHA SystemMirror together with Metro Mirror, Global Mirror,
and FlashCopy.

In addition, each section contains recommendations for daily operation in the


respective environment.

© Copyright IBM Corp. 2012. All rights reserved. 371


14.1 SVC/V7000 Copy Services
In this section, we provide you with step-by-step setup instructions for an SVC/V7000
environment using PowerHA System Mirror for i together with Metro Mirror, Global Mirror,
and FlashCopy. In addition, we provide recommendations for the daily operation in the
respective environment.

Environment overview
For this scenario, we need to create several cluster items. Cluster, cluster monitors, IASP,
and administrative domain setup are covered in Chapter 11, “Creating a PowerHA base
environment” on page 199. Creating a PoweHA base environment is common to all
the scenarios.

These are the specific cluster items for this scenario:


 Cluster resource group
– Device domain
 ASP copy descriptions and sessions

Table 14-1 and Figure 14-1 on page 373 describe and show our environment.

Table 14-1 Settings for the SVC/V7000 scenarios


Preferred primary Preferred backup FlashCopy

System name CTCIHA9V CTCIHA9W CTCIHA9X

Cluster name PWRHA_CLU

Cluster resource group SVC_MM_CRG / SVC_GM_CRG

Device domain PWRHA_DMN

Administrative domain PWRHA_CAD

IASP name, number IASP1, 144 / IASP2, 145 IASP1, 144 / IASP2, 145

Cluster resource group site name SITE1 SITE2

Takeover IP 10.0.0.1

Heartbeat cluster IP 10.10.10.1 10.10.10.2 10.10.10.3

Management access IP 9.5.167.53 9.5.167.54 9.5.167.60

SVC IP address 9.5.168.218 9.5.168.220 9.5.168.218 or


9.5.168.220

ssh key file location /QIBM/UserData/HASM/hads/.ssh/id_rsa

Volumes IDsa 15-18 / 19-22 0-3 / 4-7 23-26 or 8-11

SCSI adapter resource name DC07 / DC09 DC08 / DC09 DC07 or DC01
a. Volumes IDs do not need to be the same on the source and target of remote copy relationships.

372 PowerHA SystemMirror for IBM i Cookbook


Figure 14-1 shows a graphical overview of the setup used in the SVC scenarios.

IBM Power Systems IBM Power Systems


VIOS Primary VIOS Backup
IBM i Primary IBM i Backup
IBM i FlashCopy IBM i FlashCopy

Power 75 0
Power 75 0
D3 D4 D5 D6 D7 D8 D9 D1 0
D3 D4 D5 D6 D7 D8 D9 D10

2 paths per VIOS 2 paths per VIOS

IBM SAN Switch IBM SAN Switch


20 0 55 KB 20 0 5 5 KB

ISL
0 4 1 5 2 6 3 7 8 12 9 13 10 14 1 15 16 20 17 21 18 2 19 23 24 82 25 29 26 30 27 31 0 4 1 5 2 6 3 7 8 12 9 13 10 14 1 15 16 20 17 21 18 2 19 23 24 28 25 29 26 30 27 31

IBM SVC 4 paths per IBM SVC 4 paths per


Primary Cluster SVC node Backup Cluster SVC node
aux
MetroMirror or
Syst emSt orage

master
Syst emSt orage

FlashCopy
FlashCopy GlobalMirror
Syst emSt orage
Syst emSt orage

target
target

DS5300
DS5300

System Storage
System Storage

IBM System Storage IBM System Storage


Primary Backup

Figure 14-1 Overview of SVC setup for scenarios

14.1.1 Setting up an IBM i SVC/V7000 Copy Services environment


In this section we discuss Setting up an IBM i SVC/V7000 Copy Services environment.

Preparing for SSH connection between IBM i and SVC/V7000


Communication between PowerHA and SVC or V7000 is done using SSH. Therefore, you
must create SSH key pairs and attach the SSH public key to a user on the SVC or V7000.
The corresponding private key file to be used is specified in the creation of the ASP copy
descriptions. It has to be distributed to all nodes in the cluster.

Generation of the SSH key pair is done on IBM i from QSHELL. Example 14-1 shows the
command to do this and a list of the files created.

Example 14-1 Generation of ssh keys on IBM i


> cd /QIBM/UserData/HASM/hads/.ssh/
$
> ssh-keygen -t rsa -f id_rsa -N ''
Generating public/private rsa key pair.
Your identification has been saved in id_rsa.
Your public key has been saved in id_rsa.pub.

Chapter 14. Configuring and managing CSVC/V7000 Copy Services 373


The key fingerprint is:
76:49:e1:9e:04:0a:c5:e2:68:3a:d0:6b:0b:4b:e2:2e [email protected]
$
> ls -la
total 64
drwxrwsrwx 2 powerha 0 8192 Sep 21 17:02 .
drwx--Sr-x 3 qsys 0 8192 Sep 21 15:19 ..
-rw------- 1 powerha 0 1679 Sep 21 17:02 id_rsa
-rw-r--r-- 1 powerha 0 414 Sep 21 17:02 id_rsa.pub
$

Note: The ssh key pair generated here and used by PowerHA is in OpenSSH key format.
It cannot be used by PuTTY, as PuTTY expects SSH2 keys.

You then have to import the id_rsa.pub file as a key into a user on your SVCs. This user
must have administrator role to perform the functions used by PowerHA. Transfer the file to
your PC and import it to the SVC user (Figure 14-2).

Figure 14-2 Import SSH public key to SVC user

Make sure to distribute the id_rsa file to all nodes in your cluster to the same directory.

Initializing IBM i disk units on the backup nodes


Prior to setting up Copy Services on SVC/V7000, you have to initialize and format the
read-protected DPHxxx disk units to become usable for the IASP on the IBM i backup nodes.
This can be done in SST by choosing option 3 (Working with disk units), then selecting option

374 PowerHA SystemMirror for IBM i Cookbook


3 (Work with disk unit recovery), then selecting option 2 (Disk unit problem recovery
procedure), and finally selecting option 1 (Initialize and format disk unit). Failing to do so
can result in IASP disk units not showing up properly after a switchover/failover to the
secondary system.

14.1.2 Configuring IBM i SVC/V7000 remote Copy Services


This section describes the configuration of IBM PowerHA SystemMirror for i IASP replication
using SVC/V7000 Metro Mirror or Global Mirror.

Setting up SVC/V7000 remote Copy Services


Take the following steps to set up SVC/V7000 remote Copy Services:
1. To set up a remote Copy Services partnership between two SVC clusters, both clusters
need to be zoned in the SAN switch configuration such that they can see each other. This
can be verified using svcinfo lsclustercandidate, as shown for each SVC cluster in
Example 14-2.

Example 14-2 svcinfo lsclustercandidate


IBM_2145:ctcsvcclu1:admin>svcinfo lsclustercandidate
id configured name
0000020065FFFFFE no ctcsvcclu2

IBM_2145:ctcsvcclu2:admin>svcinfo lsclustercandidate
id configured name
0000020065005ADE no ctcsvcclu1

2. After the SVC inter-cluster zoning has been configured, we create SVC cluster
partnerships between the two clusters using svctask mkpartnership on each cluster
(Example 14-3). Note that when the partnership has been configured only on one cluster it
is in a partially_configured state.

Example 14-3 svctask mkpartnership


IBM_2145:ctcsvcclu1:admin>svctask mkpartnership -bandwidth 200 ctcsvcclu2

IBM_2145:ctcsvcclu2:admin>svctask mkpartnership -bandwidth 200 ctcsvcclu1

IBM_2145:ctcsvcclu1:admin>svcinfo lscluster
id name location partnership bandwidth id_alias
0000020065005ADE ctcsvcclu1 local
0000020065005ADE
0000020065FFFFFE ctcsvcclu2 remote fully_configured 200
0000020065FFFFFE

3. Before creating the remote copy volume relationships we create a remote copy
consistency group on each SVC cluster, which is required by PowerHA (Example 14-4).

Example 14-4 svctask mkrcconsistgrp


IBM_2145:ctcsvcclu1:admin>svctask mkrcconsistgrp -cluster ctcsvcclu2 -name IASP1_MM
RC Consistency Group, id [0], successfully created

IBM_2145:ctcsvcclu1:admin>svcinfo lsrcconsistgrp
id name master_cluster_id master_cluster_name aux_cluster_id aux_cluster_name primary state relationship_count copy_type
0 IASP1_MM 0000020065005ADE ctcsvcclu1 0000020065FFFFFE ctcsvcclu2 empty 0 empty_group

Chapter 14. Configuring and managing CSVC/V7000 Copy Services 375


4. After creating the remote copy consistency group we create the remote copy volume
relationships for our IBM i IASP volumes:
– CTCIHA9V_MM_Vx for the IBM i production node
– CTCIHA9W_MM_Vx for the backup node (Example 14-5)

Note: svctask mkrcrelationship creates Metro Mirror remote copy relationships by


default. If Global Mirror relationships are desired, add the -global parameter.

Example 14-5 Creating SVC Metro Mirror relationships


IBM_2145:ctcsvcclu1:admin>svcinfo lsvdisk -filtervalue name=CTCIHA9V_MM*
id name IO_group_id IO_group_name status mdisk_grp_id mdisk_grp_name capacity type FC_id FC_name RC_id RC_name vdisk_UID
fc_map_count copy_count fast_write_state se_copy_count
15 CTCIHA9V_MM_V0 0 io_grp0 online 1 PowerHA 20.00GB striped
600507680194016B7800000000000012 0 1 empty 0
16 CTCIHA9V_MM_V1 0 io_grp0 online 1 PowerHA 20.00GB striped
600507680194016B7800000000000013 0 1 empty 0
17 CTCIHA9V_MM_V2 0 io_grp0 online 1 PowerHA 20.00GB striped
600507680194016B7800000000000014 0 1 empty 0
18 CTCIHA9V_MM_V3 0 io_grp0 online 1 PowerHA 20.00GB striped
600507680194016B7800000000000015 0 1 empty 0

IBM_2145:ctcsvcclu2:admin>svcinfo lsvdisk -filtervalue name=CTCIHA9W_MM*


id name IO_group_id IO_group_name status mdisk_grp_id mdisk_grp_name capacity type FC_id FC_name RC_id RC_name vdisk_UID
fc_map_count copy_count fast_write_state se_copy_count
0 CTCIHA9W_MM_V0 0 io_grp0 online 0 PowerHA 20.00GB striped
600507680197FFFFF800000000000000 0 1 empty 0
1 CTCIHA9W_MM_V1 0 io_grp0 online 0 PowerHA 20.00GB striped
600507680197FFFFF800000000000001 0 1 empty 0
2 CTCIHA9W_MM_V2 0 io_grp0 online 0 PowerHA 20.00GB striped
600507680197FFFFF800000000000002 0 1 empty 0
3 CTCIHA9W_MM_V3 0 io_grp0 online 0 PowerHA 20.00GB striped
600507680197FFFFF800000000000003 0 1 empty 0

IBM_2145:ctcsvcclu1:admin>svctask mkrcrelationship -master CTCIHA9V_MM_V0 -aux CTCIHA9W_MM_V0 -cluster ctcsvcclu2 -consistgrp IASP1_MM
-name CTCIHA9_MM_V0
RC Relationship, id [15], successfully created
IBM_2145:ctcsvcclu1:admin>svctask mkrcrelationship -master CTCIHA9V_MM_V1 -aux CTCIHA9W_MM_V1 -cluster ctcsvcclu2 -consistgrp IASP1_MM
-name CTCIHA9_MM_V1
RC Relationship, id [16], successfully created
IBM_2145:ctcsvcclu1:admin>svctask mkrcrelationship -master CTCIHA9V_MM_V2 -aux CTCIHA9W_MM_V2 -cluster ctcsvcclu2 -consistgrp IASP1_MM
-name CTCIHA9_MM_V2
RC Relationship, id [17], successfully created
IBM_2145:ctcsvcclu1:admin>svctask mkrcrelationship -master CTCIHA9V_MM_V3 -aux CTCIHA9W_MM_V3 -cluster ctcsvcclu2 -consistgrp IASP1_MM
-name CTCIHA9_MM_V3
RC Relationship, id [18], successfully created

IBM_2145:ctcsvcclu1:admin>svcinfo lsrcrelationship -filtervalue name=CTCIHA9_MM*


id name master_cluster_id master_cluster_name master_vdisk_id master_vdisk_name aux_cluster_id aux_cluster_name aux_vdisk_id
aux_vdisk_name primary consistency_group_id consistency_group_name state bg_copy_priority progress copy_type
15 CTCIHA9_MM_V0 0000020065005ADE ctcsvcclu1 15 CTCIHA9V_MM_V0 0000020065FFFFFE ctcsvcclu2 0
CTCIHA9W_MM_V0 master 0 IASP1_MM inconsistent_stopped 50 0 metro
16 CTCIHA9_MM_V1 0000020065005ADE ctcsvcclu1 16 CTCIHA9V_MM_V1 0000020065FFFFFE ctcsvcclu2 1
CTCIHA9W_MM_V1 master 0 IASP1_MM inconsistent_stopped 50 0 metro
17 CTCIHA9_MM_V2 0000020065005ADE ctcsvcclu1 17 CTCIHA9V_MM_V2 0000020065FFFFFE ctcsvcclu2 2
CTCIHA9W_MM_V2 master 0 IASP1_MM inconsistent_stopped 50 0 metro
18 CTCIHA9_MM_V3 0000020065005ADE ctcsvcclu1 18 CTCIHA9V_MM_V3 0000020065FFFFFE ctcsvcclu2 3
CTCIHA9W_MM_V3 master 0 IASP1_MM inconsistent_stopped 50 0 metro

5. After the remote copy relationships have been created and added to the consistency
group we start the consistency group using svctask startrcconsistgrp (Example 14-6).
Note that after starting the consistency group its status changes from
inconsistent_stopped to inconsistent_copying.

Example 14-6 Starting the SVC remote copy consistency group


IBM_2145:ctcsvcclu1:admin>svctask startrcconsistgrp IASP1_MM

IBM_2145:ctcsvcclu1:admin>svcinfo lsrcconsistgrp IASP1_MM


id 0
name IASP1_MM

376 PowerHA SystemMirror for IBM i Cookbook


master_cluster_id 0000020065005ADE
master_cluster_name ctcsvcclu1
aux_cluster_id 0000020065FFFFFE
aux_cluster_name ctcsvcclu2
primary master
state inconsistent_copying
relationship_count 4
freeze_time
status
sync
copy_type metro
RC_rel_id 15
RC_rel_name CTCIHA9_MM_V0
RC_rel_id 16
RC_rel_name CTCIHA9_MM_V1
RC_rel_id 17
RC_rel_name CTCIHA9_MM_V2
RC_rel_id 18
RC_rel_name CTCIHA9_MM_V3

6. We can now observe the progress of the SVC remote copy background synchronization
process using svcinfo lsrcrelationship (Example 14-7).

Note: As long as the SVC remote copy background synchronization progress has not
completed (that is, it has not reached consistent_synchronized), the remote copy
relationships should not be switched, as the data on the remote site is still inconsistent.

Example 14-7 Viewing the SVC remote copy relationship status


IBM_2145:ctcsvcclu1:admin>svcinfo lsrcrelationship CTCIHA9_MM_V0
id 15
name CTCIHA9_MM_V0
master_cluster_id 0000020065005ADE
master_cluster_name ctcsvcclu1
master_vdisk_id 15
master_vdisk_name CTCIHA9V_MM_V0
aux_cluster_id 0000020065FFFFFE
aux_cluster_name ctcsvcclu2
aux_vdisk_id 0
aux_vdisk_name CTCIHA9W_MM_V0
primary master
consistency_group_id 0
consistency_group_name IASP1_MM
state inconsistent_copying
bg_copy_priority 50
progress 39
freeze_time
status online
sync
copy_type metro

Chapter 14. Configuring and managing CSVC/V7000 Copy Services 377


Setting up PowerHA for SVC/V7000 remote Copy Services
Assuming that you have already created a cluster environment, an administrative domain (if
needed), and an IASP on your production site, here are the steps required to set up the
PowerHA configuration for SVC/V7000 Metro Mirror:
1. On the backup system, create a device description for the IASP with the same IASP name
that you use on your production site:
CRTDEVASP DEVD(IASP1) RSRCNAME(IASP1) RDB(*GEN)
If you choose a different RDB name than the IASP name on the production site, make
sure to specify the same value on the backup site.

378 PowerHA SystemMirror for IBM i Cookbook


2. Create the cluster resource group (CRG) using the command shown in Figure 14-3.

Create Cluster Resource Group (CRTCRG)

Type choices, press Enter.

Cluster . . . . . . . . . . . . > PWRHA_CLU Name


Cluster resource group . . . . . > SVC_MM_CRG Name
Cluster resource group type . . > *DEV *DATA, *APP, *DEV, *PEER
CRG exit program . . . . . . . . > *NONE Name, *NONE
Library . . . . . . . . . . . Name
User profile . . . . . . . . . . > *NONE Name, *NONE

Recovery domain node list:


Node identifier . . . . . . . > CTCIHA9V Name
Node role . . . . . . . . . . > *PRIMARY *CRGTYPE, *PRIMARY...
Backup sequence number . . . . *LAST 1-127, *LAST
Site name . . . . . . . . . . > SITE1 Name, *NONE
Data port IP address . . . . . *NONE

Node identifier . . . . . . . > CTCIHA9W Name


Node role . . . . . . . . . . > *BACKUP *CRGTYPE, *PRIMARY...
Backup sequence number . . . . *LAST 1-127, *LAST
Site name . . . . . . . . . . > SITE2 Name, *NONE
Data port IP address . . . . . *NONE

Exit program format name . . . . EXTP0100 EXTP0100, EXTP0101...


Exit program data . . . . . . . *NONE
Distribute info user queue . . . *NONE Name, *NONE
Library . . . . . . . . . . . Name
Job . . . . . . . . . . . . . . *JOBD Name, *JOBD, *CRG

Configuration object list:


Configuration object . . . . . > IASP1 Name, *NONE
Configuration object type . . *DEVD *DEVD, *CTLD, *LIND, *NWSD
Configuration object online . > *ONLINE *OFFLINE, *ONLINE,
Server takeover IP address . . > '10.0.0.1'

Text description . . . . . . . . *BLANK


Failover message queue . . . . . *NONE Name, *NONE
Library . . . . . . . . . . . Name

F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display


F24=More keys
Figure 14-3 Create cluster resource group for SVC/V7000 Metro Mirror

3. Start the CRG using the following command:


STRCRG CLUSTER(PWRHA_CLU) CRG(SVC_MM_CRG)

Chapter 14. Configuring and managing CSVC/V7000 Copy Services 379


4. Add the ASP copy descriptions using ADDSVCCPYD (Figure 14-4). This has to be done for
the IASP configuration on the production site and for the IASP configuration on the backup
site, so create two ASP copy descriptions.

Add SVC ASP Copy Description (ADDSVCCPYD)

Type choices, press Enter.

ASP copy . . . . . . . . . . . . > SVC_MM_S Name


ASP device . . . . . . . . . . . > IASP1 Name
Cluster resource group . . . . . > SVC_MM_CRG Name, *NONE
Cluster resource group site . . > SITE1 Name, *NONE
Node identifier . . . . . . . . > *CRG Name, *CRG, *NONE
Storage host:
User name . . . . . . . . . . > admin
Secure shell key file . . . . > '/QIBM/UserData/HASM/hads/.ssh/id_rsa'

Internet address . . . . . . . > '9.5.168.218'

Virtual disk range:


Range start . . . . . . . . . > 15 0-8191
Range end . . . . . . . . . . > 18 0-8191
+ for more values

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 14-4 Add SVC ASP Copy Description

Unless you are on SVC/V7000 6.2 or later, the user name that you specify here has to be
admin. This profile has no relationship to the user that is used on the SVC. The actual user
is chosen based on the ssh key file pair only. With SVC6.2 you can either specify admin
as the user or the name of the SVC/V7000 user that the ssh keyfile actually belongs to.
The virtual disk range is the SVC volume IDs of the disks in your IASP.
5. Start the ASP session using STRSVCSSN (Figure 14-5).

Start SVC Session (STRSVCSSN)

Type choices, press Enter.

Session . . . . . . . . . . . . > SVC_MM Name


Session type . . . . . . . . . . > *METROMIR *METROMIR, *GLOBALMIR...
Cluster resource group . . . . . > SVC_MM_CRG Name
Switchover reverse replication *YES *YES, *NO
Failover reverse replication . . *YES *YES, *NO

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 14-5 Start SVC Session

380 PowerHA SystemMirror for IBM i Cookbook


For the session type you must specify whether the underlying remote copy function used
in the SVC setup is Metro Mirror (specify session type as *METROMIR) or Global Mirror
(specify session type as *GLOBALMIR).
The two parameters, Switchover reverse replication and Failover reverse replication,
determine whether reverse replication in the SVC environment should be started
automatically after a switchover or failover has occurred or whether the SVC session
should stay in suspended mode.
Notice that you do not have to explicitly specify the ASP copy descriptions here. They are
chosen by PowerHA from the CRG name provided in the session description and the
information about IASP name and site name in the CRG.

Your IBM PowerHA SystemMirror i SVC/V7000 remote copy configuration is now done and
your IASP is highly available for planned and unplanned site switches, as described in 14.2,
“Managing IBM i SVC/V7000 Copy Services” on page 386.

14.1.3 Configuring IBM i SVC/V7000 FlashCopy


PowerHA SystemMirror for i allows you to create a FlashCopy of an IASP in a non-replicated
IASP environment but also from a replicated IASP with Metro Mirror or Global Mirror. This
allows a point-in-time copy to be taken of either the primary or the secondary site of a remote
replication environment.

To create a FlashCopy point-in-time copy for an IASP in the IBM i cluster device domain,
perform the following steps:
1. Create ASP copy descriptions for both the FlashCopy source and the target volumes.
2. Start an ASP session with type *FLASHCOPY, which links the copy descriptions and
creates the FlashCopy relationships on the SVC/V7000 storage system.

Next we describe these steps for an example scenario of configuring FlashCopy on a Metro
Mirror secondary site based on our SVC/V7000 environment (as shown in 14.1, “SVC/V7000
Copy Services” on page 372), taking the FlashCopy from the Metro Mirror secondary volumes.

Because we have not added our IBM i partition CTCIHA9X to the cluster and device domain,
we must add it as the third node to the cluster and device domain first before being able to
create an ASP copy description referencing it (Example 14-8). We use the IBM i partition as
the FlashCopy partition accessing the FlashCopy target volumes for doing a backup to
physical tape.

Example 14-8 Adding the FlashCopy node into the cluster and device domain
ADDCLUNODE CLUSTER(PWRHA_CLU) NODE(CTCIHA9X ('10.10.10.3'))
ADDDEVDMNE CLUSTER(PWRHA_CLU) DEVDMN(PWRHA_DMN) NODE(CTCIHA9X)

We also need to create a device description for the IASP on the FlashCopy target node
(CTCIHA9X in our example) using CRTDEVASP (Example 14-9). Make sure to specify the same
database name here that you specified when creating your original IASP.

Example 14-9 Creating the ASP device description on the FlashCopy target node
CRTDEVASP DEVD(IASP1) RSRCNAME(IASP1)

Chapter 14. Configuring and managing CSVC/V7000 Copy Services 381


Creating ASP copy descriptions for FlashCopy
To create an ASP copy description for FlashCopy:
1. As we already have an ASP copy description for the Metro Mirror volumes on the
secondary site, we do not need to create a new one for FlashCopy. However, we can use
the existing ASP copy description from the Metro Mirror volumes to describe the
FlashCopy source volumes. If you do not have an existing ASP copy description already
from a remote copy relationship, use ADDSVCCPYD to add a corresponding SVC/V7000 copy
description for the FlashCopy source volumes as shown in the following step.
2. We create an ASP copy description for the FlashCopy target volumes (Figure 14-6) by
specifying information about the IASP being replicated and the details of the virtual disks
(VDisks) within the SVC/V7000 that will be used as the FlashCopy target volumes for
the IASP.

Add SVC ASP Copy Description (ADDSVCCPYD)

Type choices, press Enter.

ASP copy . . . . . . . . . . . . SVC_FLC_T Name


ASP device . . . . . . . . . . . IASP1 Name
Cluster resource group . . . . . *NONE Name, *NONE
Cluster resource group site . . *NONE Name, *NONE
Node identifier . . . . . . . . CTCIHA9W Name, *CRG, *NONE
Storage host:
User name . . . . . . . . . . admin
Secure shell key file . . . . '/QIBM/UserData/HASM/hads/.ssh/id_rsa'

Internet address . . . . . . . '9.5.168.220'

Virtual disk range:


Range start . . . . . . . . . 8 0-8191
Range end . . . . . . . . . . 11 0-8191
+ for more values

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 14-6 Add SVC/V7000 ASP Copy Description for FlashCopy target volumes

Note: For the target copy description that is to be used for FlashCopy, the cluster resource
group and cluster resource group site must be *NONE. The node identifier must be set to
the cluster node name that will own the target copy of the independent ASP.

382 PowerHA SystemMirror for IBM i Cookbook


Creating an ASP session for FlashCopy
To create an ASP session for FlashCopy, follow these steps:
1. After having created the two ASP copy descriptions for the FlashCopy source and target
volumes we are ready to create a FlashCopy for the IASP. However, before taking the
FlashCopy by starting an ASP session for FlashCopy (Figure 14-7), the IASP either needs
to be quiesced (using CHGASPACT to have as much modified data from main memory
flushed to disk for reaching a consistent database state) or varied off.

Start SVC Session (STRSVCSSN)

Type choices, press Enter.

Session . . . . . . . . . . . . > SVC_FLC Name


Session type . . . . . . . . . . > *FLASHCOPY *METROMIR, *GLOBALMIR...
ASP copy:
Preferred source . . . . . . . SVC_MM_T Name
Preferred target . . . . . . . SVC_FLC_T Name
+ for more values
Incremental flash . . . . . . . *NO *NO, *YES
Copy rate . . . . . . . . . . . 0 0-100
Cleaning rate . . . . . . . . . 0 0-100
Grain size . . . . . . . . . . . 256 256, 64
Consistency group . . . . . . . *NEW

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 14-7 Starting an SVC/V7000 FlashCopy session

Note: IBM PowerHA SystemMirror for i automatically creates the FlashCopy mappings
within the SVC/V7000 when starting an ASP session for type *FLASHCOPY. That is,
the FlashCopy mappings between the VDisks must not be created on the SVC/V7000
by a user before STRSVCSSN is executed.

STRSVCSSN allows the user to also create an incremental FlashCopy relationship and
specify a FlashCopy background copy rate and grain size. PowerHA requires the
FlashCopy relationships to be included in a consistency group that is by default newly
created by PowerHA on the SVC/V7000 when starting a FlashCopy session. Alternatively,
the user can specify the name for an existing FlashCopy consistency group. For further
information about these parameters see 7.4, “FlashCopy” on page 124. Information about
managing a FlashCopy environment with SVC/V7000 can be found in 14.2, “Managing
IBM i SVC/V7000 Copy Services” on page 386.

Chapter 14. Configuring and managing CSVC/V7000 Copy Services 383


Example 14-10 shows a CL script to be run from the FlashCopy target node for
automating a FlashCopy backup, including quiescing the IASP on production node
CTCIHA9V prior to starting the FlashCopy session, varying on the IASP on the FlashCopy
target node CTCIHA9X for doing the backup to tape before varying off the IASP on the
FlashCopy node, and removing the FlashCopy session again.

Example 14-10 CHGASPACT run from the FlashCopy target node for quiescing an IASP
PGM
RUNRMTCMD CMD('CHGASPACT ASPDEV(IASP1) +
OPTION(*SUSPEND) SSPTIMO(30)') +
RMTLOCNAME(CTCIHA9V *IP) RMTUSER(POWERHA) +
RMTPWD(REDB00K)
STRSVCSSN SSN(SVC_FLC) TYPE(*FLASHCOPY) +
ASPCPY((SVC_MM_T SVC_FLC_T))
RUNRMTCMD CMD('CHGASPACT ASPDEV(IASP1) +
OPTION(*RESUME)') RMTLOCNAME(CTCIHA9V +
*IP) RMTUSER(POWERHA) RMTPWD(REDB00K)
VRYCFG CFGOBJ(IASP1) CFGTYPE(*DEV) STATUS(*ON)
/* INSERT CALL OF YOUR BACKUP PROGRAMS HERE */
VRYCFG CFGOBJ(IASP1) CFGTYPE(*DEV) STATUS(*OFF)
ENDSVCSSN SSN(SVC_FLC)
ENDPGM

Notes: STRSVCSSN and ENDSVCSSN must be executed from the IBM i cluster node that
owns the target copy of the IASP.

The *DETACH and *REATTACH operations are not supported for a SVC/V7000
FlashCopy session.

384 PowerHA SystemMirror for IBM i Cookbook


Displaying an ASP session for FlashCopy
Using the DSPSVCSSN SSN(SVC_FLC), we can display our newly created ASP session for
FlashCopy (Figure 14-8). Note that the ASP status for the IASP FlashCopy source node is
always UNKNOWN, as the FlashCopy target node cannot determine the ASP state for the
FlashCopy source node. For the FlashCopy target node, the ASP status shows AVAILABLE
as we varied on the IASP on our target node CTCIHA9X.

Display SVC Session CTCIHA9X


09/27/11 17:40:26
Session . . . . . . . . . . . . . . . . . . : SVC_FLC
Type . . . . . . . . . . . . . . . . . . . : *FLASHCOPY

Incremental flash . . . . . . . . . . . . . : *NO


Copy rate . . . . . . . . . . . . . . . . . : 0
Cleaning rate . . . . . . . . . . . . . . . : 0
Grain size (KB) . . . . . . . . . . . . . . : 256
Consistency group . . . . . . . . . . . . . : fccstgrp2
More...
Storage cluster name . . . . . . . . . . . : ctcsvcclu2
Bottom
Copy Descriptions

ASP ASP copy ASP Replication


device name Role Status state Node
IASP1 SVC_MM_T SOURCE UNKNOWN ACTIVE CTCIHA9W
SVC_FLC_T TARGET AVAILABLE CTCIHA9X

Bottom
Copy Descriptions

ASP Copy
device Role Node progress Storage state
IASP1 SOURCE CTCIHA9W 0 Copying
TARGET CTCIHA9X
Press Enter to continue

F3=Exit F5=Refresh F11=View 2 F12=Cancel


Figure 14-8 Displaying an ASP session for FlashCopy using DSPSVCSSN

The PowerHA SystemMirror for i log file /QIBM/UserData/HASM/hads/xsm.log shows the


actions performed by PowerHA for starting the ASP session including the actual SVC/V7000
CLI commands that it executed for creating the FlashCopy consistency group and mappings
(Example 14-11).

Example 14-11 /QIBM/UserData/HASM/hads/xsm.log


09/27/2011 17:31:34 950F29B47F165001 <INFO> : YaspPluginSession : start : Start of plugin session start for ASP session: SVC_FLC
09/27/2011 17:31:34 950F29B47F165001 <INFO> : YaspSession : deleteIAspObjects : Start of delete IASP Objects (Perm: 0) : 144
09/27/2011 17:31:34 950F29B47F165001 <INFO> : YaspSession : deleteIAspObjects : delete IASP objects succeeded
09/27/2011 17:31:34 950F29B47F165001 <INFO> : YaspSession : iapEnlistReject : Start of IASP enlist reject: 5
144
09/27/2011 17:31:34 950F29B47F165001 <INFO> : YaspSession : iapEnlistReject : iaspEnlistReject succeeded
09/27/2011 17:31:34 950F29B4CA3B3001 <INFO> : YaspSVCSession : startFlashCopy : Start of FlashCopy start for session SVC_FLC
09/27/2011 17:31:35 950F29B4CA3B3001 /QOpenSys/usr/bin/ssh -i "/QIBM/UserData/HASM/hads/.ssh/id_rsa" -o UserKnownHostsFile=/dev/null -o
StrictHostKeyChecking=no [email protected] "svctask mkfcconsistgrp"

FlashCopy Consistency Group, id [2], successfully created

Chapter 14. Configuring and managing CSVC/V7000 Copy Services 385


09/27/2011 17:31:35 950F29B4CA3B3001 /QOpenSys/usr/bin/ssh -i "/QIBM/UserData/HASM/hads/.ssh/id_rsa" -o UserKnownHostsFile=/dev/null -o
StrictHostKeyChecking=no [email protected] "svcinfo lsfcconsistgrp -delim \"&\" -nohdr -filtervalue \"id=2\""

2&fccstgrp2&empty

09/27/2011 17:31:36 950F29B4CA3B3001 /QOpenSys/usr/bin/ssh -i "/QIBM/UserData/HASM/hads/.ssh/id_rsa" -o UserKnownHostsFile=/dev/null -o


StrictHostKeyChecking=no [email protected] "svctask mkfcmap -source 0 -target 8 -consistgrp fccstgrp2 -copyrate 0 -grainsize 256
-cleanrate 0"

FlashCopy Mapping, id [0], successfully created

09/27/2011 17:31:36 950F29B4CA3B3001 /QOpenSys/usr/bin/ssh -i "/QIBM/UserData/HASM/hads/.ssh/id_rsa" -o UserKnownHostsFile=/dev/null -o


StrictHostKeyChecking=no [email protected] "svctask mkfcmap -source 1 -target 9 -consistgrp fccstgrp2 -copyrate 0 -grainsize 256
-cleanrate 0"

FlashCopy Mapping, id [1], successfully created

09/27/2011 17:31:37 950F29B4CA3B3001 /QOpenSys/usr/bin/ssh -i "/QIBM/UserData/HASM/hads/.ssh/id_rsa" -o UserKnownHostsFile=/dev/null -o


StrictHostKeyChecking=no [email protected] "svctask mkfcmap -source 2 -target 10 -consistgrp fccstgrp2 -copyrate 0 -grainsize 256
-cleanrate 0"

FlashCopy Mapping, id [2], successfully created

09/27/2011 17:31:38 950F29B4CA3B3001 /QOpenSys/usr/bin/ssh -i "/QIBM/UserData/HASM/hads/.ssh/id_rsa" -o UserKnownHostsFile=/dev/null -o


StrictHostKeyChecking=no [email protected] "svctask mkfcmap -source 3 -target 11 -consistgrp fccstgrp2 -copyrate 0 -grainsize 256
-cleanrate 0"

FlashCopy Mapping, id [3], successfully created

09/27/2011 17:31:38 950F29B4CA3B3001 /QOpenSys/usr/bin/ssh -i "/QIBM/UserData/HASM/hads/.ssh/id_rsa" -o UserKnownHostsFile=/dev/null -o


StrictHostKeyChecking=no [email protected] "svctask startfcconsistgrp -prep fccstgrp2"

09/27/2011 17:31:39 950F29B4CA3B3001 <INFO> : YaspSVCSession : startFlashCopy : FlashCopy session SVC_FLC started successfully
09/27/2011 17:31:39 950F29B4CA3B3001 <INFO> : yaspSVCActionPgm : doSessionAction : Session action completed with return code: 1
09/27/2011 17:31:39 950F29B47F165001 <INFO> : YaspSession : iapEnlistReject : Start of IASP enlist reject: 6
144
09/27/2011 17:31:39 950F29B47F165001 <INFO> : YaspSession : iapEnlistReject : iaspEnlistReject succeeded
09/27/2011 17:31:39 950F29B47F165001 <INFO> : YaspSession : waitForUnitsToEnlist : Start of wait for units to enlist for session: SVC_FLC
09/27/2011 17:31:47 950F29B47F165001 <INFO> : YaspSession : waitForUnitsToEnlist : Disk units for all ASPs in session: SVC_FLC have
enlisted
09/27/2011 17:31:47 950F29B47F165001 <INFO> : YaspSession : resetMultipath : Start of reset multipath: 144
09/27/2011 17:31:47 950F29B47F165001 <INFO> : YaspSession : resetMultipath : Reset multipath succeeded
09/27/2011 17:31:47 950F29B47F165001 <INFO> : YaspPluginSession : start : Plugin session start for ASP Session: SVC_FLC completed
successfully.

14.2 Managing IBM i SVC/V7000 Copy Services


In this section we discuss managing IBM i SVC/V7000 Copy Services.

386 PowerHA SystemMirror for IBM i Cookbook


14.2.1 Displaying and changing a remote copy ASP session
An existing SVC/V7000 remote copy ASP session can be displayed using DSPSVCSSN, as
shown for our synchronized Metro Mirror IASP environment in Figure 14-9.

Display SVC Session CTCIHA9V


09/29/11 17:12:11
Session . . . . . . . . . . . . . . . . . . : SVC_MM
Type . . . . . . . . . . . . . . . . . . . : *METROMIR

Switchover reverse replication . . . . . . : *YES


Failover reverse replication . . . . . . . : *NO
Consistency group . . . . . . . . . . . . . : IASP1_MM
Source storage cluster name . . . . . . . . : ctcsvcclu1
Target storage cluster name . . . . . . . . : ctcsvcclu2
More...
CRG name . . . . . . . . . . . . . . . . . : SVC_MM_CRG
Source site name . . . . . . . . . . . . . : SITE1
Target site name . . . . . . . . . . . . . : SITE2

Bottom
Copy Descriptions

ASP ASP copy ASP Replication


device name Role Status state Node
IASP1 SVC_MM_S SOURCE AVAILABLE ACTIVE CTCIHA9V
SVC_MM_T TARGET UNKNOWN CTCIHA9W

Bottom
ASP Copy
device Role Node progress Storage state
IASP1 SOURCE CTCIHA9V 100 Consistent sync
TARGET CTCIHA9W

Bottom
Press Enter to continue

F3=Exit F5=Refresh F11=View 2 F12=Cancel


Figure 14-9 Displaying an SVC/V7000 Metro Mirror ASP session

Using CHGSVCSSN ... OPTION(*CHGATTR), only the switchover and failover reverse replication
parameters can be changed.

14.2.2 Suspending a remote copy ASP session


A remote copy ASP session for Metro Mirror or Global Mirror can be suspended to pause
replication, as for performing disruptive maintenance actions for the remote site storage
system. The remote copy secondary volumes remain inaccessible for the host.

An SVC/V7000 remote copy ASP session can be suspended from the IBM i cluster node that
owns the primary or secondary copy of the IASP, using CHGSVCSSN ... OPTION(*SUSPEND),
which puts the SVC/7000 remote copy relationships in consistent_stopped state.

Chapter 14. Configuring and managing CSVC/V7000 Copy Services 387


To resume replication with synchronization of tracked changes from the primary volumes to
the secondary volumes use CHGSVCSSN ... OPTION(*RESUME).

The backup node becomes ineligible while the remote copy replication is suspended or
resuming (Figure 14-10). That is, the CRG cannot be switched or failed over before the
remote copy status becomes consistent_synchronized again.

Work with Recovery Domain

Cluster resource group . . . . . . . . . : SVC_MM_CRG


Consistent information in cluster . . . : Yes

Type options, press Enter.


1=Add node 4=Remove node 5=Display more details 20=Dump trace

Current Preferred Site


Opt Node Status Node Role Node Role Name

CTCIHA9V Active *PRIMARY *PRIMARY SITE1


CTCIHA9W Ineligible *BACKUP 1 *BACKUP 1 SITE2

Bottom
Parameters for options 1 and 20 or command
===>
F1=Help F3=Exit F4=Prompt F5=Refresh F9=Retrieve F12=Cancel
F13=Work with cluster menu
Figure 14-10 Work with Recovery Domain panel after suspending an ASP session

A suspended SVC/V7000 Metro Mirror or Global Mirror ASP session (because data
consistency is ensured by using SVC/V7000 remote copy consistency groups) can also be
detached, which puts the remote copy relationship in idling state to allow access of the
secondary copy of the IASP from the IBM i backup node.

14.2.3 Detaching and reattaching a remote copy ASP session


If you want to get access to the IASP on your backup node, you have to detach the IASP
session associated with that IASP first. Reasons to do this might be short-term testing on the
backup node or major application changes on the production node that you do not want to
propagate to the backup system before doing final testing on the production side. To get a
consistent status of your application data, issue CHGASPACT on your production site first
(Example 14-12).

Example 14-12 CHGASPACT for quiescing an IASP


CHGASPACT ASPDEV(IASP1) OPTION(*SUSPEND) SSPTIMO(30)

You can then detach the session from your backup node using the command shown in
Example 14-13.

Example 14-13 CHGSVCSSN for detaching an IASP


CHGSVCSSN SSN(SVC_GM) OPTION(*DETACH)

388 PowerHA SystemMirror for IBM i Cookbook


After the detach is finished, the SVC/V7000 remote copy relationships are in idling status and
you can resume access to your IASP on the production system using CHGASPACT
OPTION(*RESUME). The IASP can be varied on on the backup node and can be used there. By
default, changes are tracked on the source side and on the target side so that a full
resynchronization is not required after a reattach.

The reattach again has to be done from the backup node. Any changes made to the
secondary copy IASP while detached will be overwritten by the data from the primary ASP
copy. Vary off the IASP on the backup node and issue the command shown in
Example 14-14.

Example 14-14 CHGSVCSSN for reattaching an IASP


CHGSVCSSN SSN(SVC_GM) OPTION(*REATTACH) SNDINQMSG(*YES)

By default, a message in the QSYSOPR message queue tells you which node will be acting
as the primary node after the reattach and waits for confirmation before starting remote copy
services on the SVC/V7000 again. If you want to avoid this behavior, you can specify the
SNDINQMSG(*NO) parameter.

Chapter 14. Configuring and managing CSVC/V7000 Copy Services 389


14.2.4 Planned switchover
Planned switchovers are usually done for maintenance purposes on the production system.
For example, you might need to install PTFs that require an IPL on your production system,
and you want to keep your users working during this timeframe.These are the steps for a
planned switchover:
1. Using DSPSVCSSN, make sure that remote copy services on your SVC/V7000 environment
work correctly and that data is in sync between production and backup. Figure 14-11 gives
an example of an ASP session for SVC Metro Mirror replication, where Metro Mirror is in a
consistent status. Notice that the replication status is reported as active, meaning that the
target IASP is current with the source IASP.

Display SVC Session CTCIHA9V

Session . . . . . . . . . . . . . . . . . . : SVC_MM
Type . . . . . . . . . . . . . . . . . . . : *METROMIR

Switchover reverse replication . . . . . . : *YES


Failover reverse replication . . . . . . . : *YES
Consistency group . . . . . . . . . . . . . : IASP1_MM
Source storage cluster name . . . . . . . . : ctcsvcclu1
Target storage cluster name . . . . . . . . : ctcsvcclu2
CRG name . . . . . . . . . . . . . . . . . : SVC_MM_CRG
Source site name . . . . . . . . . . . . . : SITE1
Target site name . . . . . . . . . . . . . : SITE2

Copy Descriptions

ASP ASP copy ASP Replication


device name Role Status state Node
IASP1 SVC_MM_S SOURCE AVAILABLE ACTIVE CTCIHA9V
SVC_MM_T TARGET UNKNOWN CTCIHA9W

ASP Copy
device Role Node progress Storage state
IASP1 SOURCE CTCIHA9V 100 Consistent sync
TARGET CTCIHA9W

Figure 14-11 Displaying an ASP session for SVC Metro Mirror replication

2. End all applications using the IASP on the production site.


3. Use CHGCRGPRI to switch from the current production node to the next node in the
recovery domain.

CHGCRGPRI performs the following tasks:


 It ends the switchable IP interface on the production site if one is defined.
 It varies off the IASP on the production site.
 In the cluster, it promotes the first backup node to primary node and moves the old primary
node to the last backup position.

390 PowerHA SystemMirror for IBM i Cookbook


 If switchover reverse replication is configured with the default value of *YES in the ASP
session, then the remote copy direction is switched on the SVC/V7000.
If switchover reverse replication is configured as *NO in the ASP session, then the remote
copy is detached on the SVC/V7000. When you want to restart the replication, you have to
issue CHGSVCSSN <session name> OPTION(*REATTACH) from the current backup system.
 It varies on the IASP on the backup site if the configuration object information in the CRG
specifies *ONLINE for the IASP.
 It starts the switchable IP interface on the backup site if one is defined.

14.2.5 Unplanned failover


For an unplanned failover we can distinguish between the following failure scenarios from the
cluster resource service perspective:
 Primary node failure triggering an automatic failover event
 Primary node failure without an automatic failover event
 Primary node partition status

Each scenario requires different failover/recovery actions, which we describe in further detail
in the following sections.

Primary node failure triggering an automatic failover event


An unplanned automatic failover event is triggered for a switchable IASP in a cluster resource
group for a primary node failure event detected by cluster resource services either by a panic
message sent by the failing primary node, as for an action of ending a cluster node or
powering down the system, or by a power state change event sent by the HMC CIMOM
server for a partition failure to the registered IBM i cluster nodes when using advanced node
failure detection, as described in 11.2, “Setting up cluster monitors” on page 208.

For an unplanned automatic failover event, a CPABB02 inquiry message is sent to the cluster
or CRG message queue on the backup node if a failover message queue is defined for either
the cluster or the CRG. If no failover message queue is defined, then the failover starts
immediately without any message being posted.

With the default cluster parameters FLVDFTACT=*PROCEED and FLVWAITTIM=*NOWAIT,


an automatic failover is triggered. Setting a failover wait time with the FLVWAITTIM cluster
parameter, either specified with a duration in minutes or *NOMAX, allows the user to respond
to the CPABB02 inquiry message to either proceed with or cancel the failover. The failover
default action FLVDFTACT parameter setting determines IBM PowerHA SystemMirror for its
behavior to either automatically proceed or cancel the failover processing after the specified
failover wait time expired without getting a response for the inquiry message from the user.
Note that in any case, the primary IASP is taken offline by IBM PowerHA SystemMirror for i
for a failover event regardless of these cluster failover parameter settings.

Chapter 14. Configuring and managing CSVC/V7000 Copy Services 391


When using the default ASP session setting parameter FLVRVSREPL=*NO for an
SVC/V7000 remote copy ASP session, for a failover event the session is only detached
(Figure 14-12) for our ASP session for SVC Metro Mirror after an unplanned failover event.
Also, the remote copy relationships are not reversed, which allows the user to preserve the
primary node IASP data for possible further failure analysis.

Display SVC Session CTCIHA9W


09/28/11 18:23:47
Session . . . . . . . . . . . . . . . . . . : SVC_MM
Type . . . . . . . . . . . . . . . . . . . : *METROMIR

Switchover reverse replication . . . . . . : *YES


Failover reverse replication . . . . . . . : *NO
Consistency group . . . . . . . . . . . . . : IASP1_MM
Source storage cluster name . . . . . . . . : ctcsvcclu2
Target storage cluster name . . . . . . . . : ctcsvcclu1
More...
CRG name . . . . . . . . . . . . . . . . . : SVC_MM_CRG
Source site name . . . . . . . . . . . . . : SITE2
Target site name . . . . . . . . . . . . . : SITE1

Bottom
Copy Descriptions

ASP ASP copy ASP Replication


device name Role Status state Node
IASP1 SVC_MM_T SOURCE AVAILABLE DETACHED CTCIHA9W
SVC_MM_S TARGET UNKNOWN CTCIHA9V

ASP Copy
device Role Node progress Storage state
IASP1 SOURCE CTCIHA9W 100 Idling
TARGET CTCIHA9V

Bottom
Press Enter to continue

F3=Exit F5=Refresh F11=View 2 F12=Cancel


Figure 14-12 Displaying an ASP session for SVC Metro Mirror after an unplanned failover event

After recovery of the preferred primary node, including restart of its cluster services, the
detached ASP session should be re-attached to resume remote copy replication and make
the IASP highly available again. There are options for re-attaching the ASP session:
 To resume remote copy data replication from the secondary site back to the primary use
CHGSVCSSN ... OPTION(*REATTACH) on the preferred primary site node, which is the
current backup node. A planned switchback to the preferred primary node can then be
done using CHGCRGPRI.
 Alternatively, if instead the data updates performed on the secondary site while the
primary site node was not available will be discarded, the direction of remote copy data
replication can be changed from the preferred primary node to the preferred backup node
by ending the CRG, changing the recovery domain primary/backup node roles via CHGCRG
before reattaching the session via CHGSVCSSN ... OPTION(*REATTACH) to resume
replication between the preferred primary and preferred backup node site.

392 PowerHA SystemMirror for IBM i Cookbook


Note: The re-attach operation always needs to be run on the node that is supposed to
become the target for the remote copy replication.

Primary node failure without an automatic failover event


A primary node failure without an automatic failover event being triggered could be caused,
for example, by an IASP or SYSBAS storage access loss.

In this case the user should carefully consider whether it is more appropriate to try to recover
from the access loss condition before invoking a failover to the backup cluster node, which
requires the following manual actions. If the primary node is still responsive, the ASP session
can be detached as follows to stop the remote copy relationships by allowing access to the
secondary volumes so that IASP can be varied on at the backup node:
CHGASPSSN SSN(<ASP_session>) OPTION(*DETACH)

The *DETACH operation for a failed IASP implies that the node roles still need to be changed
via ENDCRG and CHGCUR RCYDMNACT(*CHGCUR).

Otherwise, if the primary node is no longer responsive, a cluster partition condition and
failover can be enforced from the backup node as follows:
CHGCLUNODE CLUSTER(<cluster_name>) NODE(<primary_node_name>) OPTION(*CHGSTS)

Note: If CHGCLUNODE first fails with message CPFBB89, retry the command a short time
later after the cluster changed to a partition condition.

A CHGCLUNODE triggers a cluster failover without a failover inquiry message and still requires
the user to vary on the IASP on the new primary node and start the takeover IP interface.

After recovery of the preferred primary node, including a restart of its cluster services and of
the device CRG, the ASP session should be reattached on the preferred primary node, which
is the current backup node to resume replication. A switchback to the preferred primary node
can then be done using CHGCRGPRI.

Primary node partition status


A primary node partition status can be triggered by a sudden primary node failure or cluster
communication error. Using advanced node failure detection, as described in 11.2, “Setting
up cluster monitors” on page 208, can help to avoid cluster partition conditions for many
cases of sudden node failures, which the HMC or flexible service processor (FSP) are able to
detect by sending corresponding CIM server power state events to the registered IBM i
cluster nodes to invoke an automatic failover. However, there are still cases like cluster
heartbeat communication errors or sudden system power outages that lead to a cluster node
partition condition. The user should determine the reason for the partition condition of a
cluster node to decide whether to fix it or declare the node in partition status as failed to
invoke a cluster failover to a backup node.

A cluster partition condition with a failed cluster node indicated by message ID CPFBB20
requires manual actions to invoke a cluster failover, for example, by setting the primary
cluster node from partition to failed status using the following CL command:
CHGCLUNODE CLUSTER(<cluster_name>) NODE(<primary_node_name>) OPTION(*CHGSTS)

After this CHGCLUNODE command, cluster node roles are changed and remote copy
relationships are stopped so that the IASP can be manually varied on at the new primary
node on the backup site.

Chapter 14. Configuring and managing CSVC/V7000 Copy Services 393


After recovery of the preferred primary node, including restart of its cluster services and of the
device CRG, a switchback to the preferred primary node can be done using CHGCRGPRI.

14.2.6 Displaying and changing a FlashCopy ASP session


An existing SVC/V7000 FlashCopy ASP session can be displayed using DSPSVCSSN, as
shown in Figure 14-13 for our FlashCopy no-copy environment after varying on the IASP on
the FlashCopy target node CTCIHA9X.

Display SVC Session CTCIHA9X


09/29/11 18:27:08
Session . . . . . . . . . . . . . . . . . . : SVC_FLC
Type . . . . . . . . . . . . . . . . . . . : *FLASHCOPY

Incremental flash . . . . . . . . . . . . . : *NO


Copy rate . . . . . . . . . . . . . . . . . : 0
Cleaning rate . . . . . . . . . . . . . . . : 0
Grain size (KB) . . . . . . . . . . . . . . : 256
Consistency group . . . . . . . . . . . . . : fccstgrp1
More...
Storage cluster name . . . . . . . . . . . : ctcsvcclu2

Bottom
Copy Descriptions

ASP ASP copy ASP Replication


device name Role Status state Node
IASP1 SVC_MM_T SOURCE UNKNOWN ACTIVE CTCIHA9W
SVC_FLC_T TARGET AVAILABLE CTCIHA9X

Bottom
ASP Copy
device Role Node progress Storage state
IASP1 SOURCE CTCIHA9W 0 Copying
TARGET CTCIHA9X

Bottom
Press Enter to continue

F3=Exit F5=Refresh F11=View 2 F12=Cancel


Figure 14-13 Displaying an SVC/V7000 FlashCopy ASP session

When using FlashCopy with the background copy (that is, copy rate > 0), the background
copy progress is reported in the copy progress information with a storage state of copying
when the background copy has not finished yet or copied when the background copy has
completed, as also indicated by a copy progress of 100.

Using CHGSVCSSN ... OPTION(*CHGATTR), only the copy rate and cleaning rate parameters of
a FlashCopy ASP session can be changed. This operation can be executed from either the
IBM i cluster node that owns the source copy or the target copy of the IASP.

394 PowerHA SystemMirror for IBM i Cookbook


14.2.7 Reversing a FlashCopy ASP session
FlashCopy reverse for SVC/7000 is currently not supported by PowerHA SystemMirror for i
and is planned to be released with a future PTF.

14.2.8 Using incremental FlashCopy


Incremental FlashCopy copies data from the source to the target volumes that have been
modified since the initial creation of the FlashCopy or the last time an increment operation
was performed.

Using incremental FlashCopy requires that the ASP session for the initial FlashCopy is
created with the incremental flash option set to *YES using STRSVCSSN ... TYPE(*FLASHCOPY)
INCR(*YES). A background copy rate greater than 0 should be specified when starting the
ASP session for the initial FlashCopy because any subsequent incremental FlashCopy
operations can only be performed using CHGSVCSSN with the *INCR option after the initial
background copy has finished.

As with any other FlashCopy operation, the FlashCopy incremental operation must also be
executed from the IBM i cluster node that owns the target copy of the independent IASP.

Chapter 14. Configuring and managing CSVC/V7000 Copy Services 395


Figure 14-14 shows an existing FlashCopy incremental session displayed using DSPSVCSSN.

Display SVC Session CTCIHA9X


10/03/11 12:02:16
Session . . . . . . . . . . . . . . . . . . : SVC_FLC
Type . . . . . . . . . . . . . . . . . . . : *FLASHCOPY

Incremental flash . . . . . . . . . . . . . : *YES


Copy rate . . . . . . . . . . . . . . . . . : 100
Cleaning rate . . . . . . . . . . . . . . . : 0
Grain size (KB) . . . . . . . . . . . . . . : 256
Consistency group . . . . . . . . . . . . . : fccstgrp0
More...
Storage cluster name . . . . . . . . . . . : ctcsvcclu2

Bottom
Copy Descriptions

ASP ASP copy ASP Replication


device name Role Status state Node
IASP1 SVC_MM_T SOURCE UNKNOWN ACTIVE CTCIHA9W
SVC_FLC_T TARGET AVAILABLE CTCIHA9X

Bottom
ASP Copy
device Role Node progress Storage state
IASP1 SOURCE CTCIHA9W 100 Idle or copied
TARGET CTCIHA9X

Bottom
Press Enter to continue

F3=Exit F5=Refresh F11=View 1 F12=Cancel


Figure 14-14 Display SVC Session panel for an incremental FlashCopy session

If the background copy has not been enabled or finished, the incremental FlashCopy
operation is rejected with the following CMMVC5907E message:
The FlashCopy mapping or consistency group was not started because the mapping or
consistency group is already in the copying state.

In this case the user should either wait for an existing background copy process to complete,
which can be checked from the copy progress information displayed by DSPSVCSSN, or enable
the background copy by using CHGSVCSSN ... OPTION(*CHGATTR) CPYRATE(...).

396 PowerHA SystemMirror for IBM i Cookbook


15

Chapter 15. Best practices


In this chapter we describe best practices that can help you efficiently use IBM PowerHA
SystemMirror for i.

For further details about PowerHA functions and issues, you might also want to consult the
“IBM Software Knowledge Base” section related to high availability at the following URL:
http://www-912.ibm.com/s_dir/slkbase.NSF/wHighAv?OpenView&view=wHighAv

© Copyright IBM Corp. 2012. All rights reserved. 397


15.1 Clustering configuration
In this section we discuss clustering configuration.

15.1.1 PowerHA license consideration


The 5770HAS Licensed Program Product is not tier based, but processor based. This means
that on each node member of a cluster, you need a license key related to the number of
processors using the product on every partition of the same server.

To avoid product commands from becoming unusable in case of processor activation related
to temporary Capacity On Demand feature usage, make sure to apply PTF SI41735.

15.1.2 Requirements
When looking at the cluster properties with the GUI, there is an option to check the
requirements. Make sure to use this option and apply all of them. As shown in Figure 11-10
on page 207, these are the most common requirements to be applied:
 System value QRETSVRSEC must be set to 1. This value is necessary for the
administrative domain to retrieve user profiles passwords.

Note: Considerations apply regarding this system value, which are described in the
knowledge base document number 518525704 Admin Domain: What to Consider For
User Profiles When Implementing, at the following URL:
http://www-912.ibm.com/s_dir/slkbase.NSF/cd034e7d72c8b65c862570130058d69e/92
eb712dc067c1558625757b006a2a55?OpenDocument

 The QGPL/QBATCH job queue entry for QSYS/QBATCH must have *NOMAX or a value
greater than 1. This value is necessary because most of the cluster jobs are submitted to
this job queue, and they must not be prevented from running by user jobs.
 QIBM_PWRDWNSYS_CONFIRM and QIBM_ENDSYS_CONFIRM environment
variables should be set to *YES.

15.1.3 Independent ASP


For a device recovery configuration, independent ASPs are the main building block for
implementing a hardware-based replication solution with PowerHA. Therefore, most of the
best practices that you can use for the independent ASP can apply for clustering. See the
overview in Chapter 2, “Implementing an independent auxiliary storage pool” on page 13, and
IBM i 6.1 Independent ASPs: A Guide to Quick Implementation of Independent ASPs,
SG24-7811, for more information.

Caution: Any IASP existing on a system before adding this system to a device domain
must be deleted.

15.1.4 Communications
The inetd service must be started on all cluster nodes.

398 PowerHA SystemMirror for IBM i Cookbook


The network attribute Allow Add to Cluster (ALWADDCLU) must have the value *ANY.

The cluster makes use of several types of IP interfaces:


 Heartbeat interfaces
These are used to ensure efficient communication between all cluster nodes. They do not
need high bandwidth, as they exchange only a small amount of information.
To avoid situations in which the users can no longer connect to the node, and at the same
time nodes can no longer exchange heartbeat information (as described in “Metro Mirror
and Global Mirror failover and failback” on page 350), you might want to use separate
network devices for both user access and heartbeat access.
Conversely, to protect heartbeat exchanges you can also activate redundancy by using
virtual IP addressing or using the two available IP interfaces at the cluster definition level.
Of course, physical links must use separate network devices.
 Data port interfaces
These interfaces are used for data replication in case of geographic mirroring
implementation. For bandwidth and redundancy purposes, you can configure up to four IP
addresses. These IPs should use separate Ethernet adapters and network devices.
We recommend that you do not mix data port traffic and that users access traffics to avoid
performance degradation for the replication. Heartbeat interfaces and data port interfaces
can share the same IP addresses.
 Communication with HMC or VIOS

When using advanced node failure detection, the HMC or VIOS should be able to send
appropriate information to cluster nodes in case of partition failure. There is no specific
configuration step for this item, but the name or IP address of the HMC or VIOS should be
available. Therefore, this communication takes the IP routes available when needed. You
might want to make sure that no network device failure could have an impact on this
communication.
If there is a firewall between the HMC or VIOS and the cluster nodes, make sure that the
flow is authorized. Advanced node failure detection makes use of the CIMOM services,
which listen on TCP 5989 (unencrypted communication) and 5990 (ssl encrypted
communication) ports. Flow must be authorized in both ways.
 Communication with DS8000
When using DS8000 replication, the cluster nodes need to establish IP communication
with the DS8000 to check the replication status and start the appropriate actions, such as
failover or failback. There is no specific configuration step for this item, but the name or IP
address of the DS8000 should be available. Therefore, this communication will take the IP
routes available when needed. You might want to make sure that no network devices
failure can have an impact on this communication.
If there is a firewall between the cluster nodes and the DS8000, make sure that the flow is
authorized. DS8000 listens on the TCP 1750 port. The flow must be authorized from the
nodes to the DS8000.

Chapter 15. Best practices 399


15.1.5 Failover wait time and default action
The failover wait time and default action settings determine the cluster behavior in case of a
possible failover need detected by the backup node. Make sure that you understand how they
work together to avoid unexpected events:
 Wait time defines the time to wait for a reply to the failover inquiry message sent to the
cluster message queue:
– THe *NOWAIT value means that failover proceeds immediately without
any confirmation.
– The *NOMAX value means that there is no limit for the cluster waiting on a reply for the
message. If nobody answers, failover does not proceed.
– “a number of minutes” means that the cluster will wait this number of minutes before
proceeding as specified by the default action.
 The default action defines the action at the end of wait time if specified in minutes.
– The *PROCEED value means to proceed with the failover.
– The *CANCEL value means that failover occurs.

15.1.6 Administrative domain


One of the most common errors is related to a UID/GID user profile mismatch between the
source and target nodes. As detailed in 15.5.2, “Synchronizing UIDs/GIDs” on page 409, it
might have a huge impact when switching over or failing over. Using the cluster administrative
domain is a good way to maintain consistency regarding this parameter. However, the
administrative domain only takes care of resources that you have requested it to monitor. By
itself, it is not able to take care of new resources. For example, when a new user profile is
created, it is not added to the monitored resources list. A workaround is to use the
QIBM_QSY_CRT_PROFILE exit point, which is activated when a user profile is created. You
can write a CL program and register it against the exit point. This program should add the
newly created user profile to the administrative domain with ADDCADMRE. Note that the same
kind of considerations apply when you want to delete a user profile. It must be removed from
the administrative domain monitored resources with RMVCADMRE before being deleted.
The QIBM_QSY_DLT_PROFILE exit point can be used to automate the process. Care must
be taken when deleting a user profile that owns objects. You need to decide what to do with
the owned objects.

15.2 Journaling
It is difficult to achieve both Recovery Time Objective (RPO) and Recovery Point Objective
(RTO) with an hardware replication solution such as IBM PowerHA SystemMirror for i without
looking seriously to a journaling setup. This solution relies on disk replication. Therefore, it
copies, from source disks to target disks, only data that exists on disks. Everything that is still
in main memory and that has not been yet written to disk cannot be replicated. It is not a
matter for planned switches, for which you will use varying off the IASP, which flushes
memory content to disks. It becomes a matter for unplanned switches, which can occur at any
time, for any reason, and for which we have to apply techniques to prevent loosing data and
to reduce recovery time.

Let us see an example. Assume that an application makes use of a data area to keep track of
the next customer number of a database file to record registration information for the new
customer to be assigned and of an IFS file to capture a scanned image of the customer’s

400 PowerHA SystemMirror for IBM i Cookbook


signature. Three separate objects are used to store data. At the end of the transaction that
enrolls this new customer, everything is consistent, for those three objects in main memory.
But there is no assurance that the related disk image has received the updates. Most likely,
they have not been received yet. It might happen that the disk update order will be different
from the memory updates order, depending on main memory flushing algorithm.

This lack of consistency on disks drives clearly affects the ability to reach planned RPO.
Some objects, like database files, might also need time-consuming internal consistency
checks and recovery at IASP vary-on time. Detecting and correcting them affects planned
RTO also.

Using journal is the way to go for achieving a consistent state and reducing recovery times as
much as possible. Journal protection should be enabled for all critical recovery point
objectives, data areas, database files, data queues, and IFS files. Using journal does not
require an update to the application code. Starting, ending, and managing journal operations
are to be done outside the application. The best protection for data consistency is achieved
by changing the application to take care of transaction consistency with commitment control.
The application decides when, from a functional standpoint, any transaction is finished and
then it performs a commit. If an outage occurs during commitment control cycles, the system
ensures, at IASP vary-on time, that data is consistent regarding the transactions. It undoes
any update that is not committed.

Journaling has never been mandatory with the IBM i operating system. Therefore, numerous
users have never tried to implement it, even when using hardware replication solutions. Note
that software replication solutions are based on journaling, so considerations differ for them.

Note: It is a user responsibility to take care of journaling. IBM PowerHA SystemMirror for i
has no option to help you with it.

In the past, there were two main obstacles to using journaling:


 Performance impact
 Management effort

15.2.1 Journal performance impact


Performance impact still exists and always will because the way that journaling works implies
using more hardware resources, specifically increased disk writes operations. The objective
of journaling is to write to disk, in the journal receiver object, on a synchronous manner, all
updates to journaled objects before these updates become effective on disks.

However, in IBM i release after release, major improvements have been done on this subject:
 When using internal drives for IASP, write cache performance is a key factor for journal
performance. For more information about this topic, refer to Journaling - Configuring for
your Fair Share of Write Cache, TIPS0653, at:
http://publib-b.boulder.ibm.com/abstracts/tips0653.html
 If you decide to use a private ASP for receivers, do not skimp on the quantity of disk arms
available for use by the journal. For more information about this topic, refer to Journaling -
User ASPs Versus the System ASP, TIPS0602, at:
http://publib-b.boulder.ibm.com/abstracts/tips0602.html

Note: Do not forget that a private (user) ASP can be created inside an IASP as a
secondary ASP type and coexist with a primary type ASP.

Chapter 15. Best practices 401


 If you do not intend to use journal receiver entries for application purposes, or if you want
to do so and are ready for API programming, consider minimizing receiver entries data.
Journal parameters receiver size options, minimize entry specific data, and fixed length
data play a role in this optimization step. For example, if you are using those journals only
for recovery purposes, do you really need the entire database file record content in each
record entry? For more information about this topic, refer to Journaling: How to View and
More Easily Audit Minimized Journal Entries on the IBM System i Platform, TIPS0626, at:
http://www.redbooks.ibm.com/abstracts/tips0626.html
 Depending on application needs, you cannot start journaling for work files that are used
only to help data processing and do not contain any critical data.
 The number of disks arms used by a journal receiver is determined through the journal
threshold parameter. Using as many disks as possible is a good point for performance
improvement. For more information about this topic, refer to Journaling - How Can It
Contribute to Disk Usage Skew?, TIPS0603, and Journaling: Unraveling the mysteries of
sporadic growth of Journal receivers, TIPS0652, at:
http://www.redbooks.ibm.com/abstracts/tips0603.html
http://www.redbooks.ibm.com/abstracts/tips0652.html

Note: PTF MF51614 for IBM i 7.1 apply result is that journal receivers are spread
across all the disks just like any other object type, via normal storage management
technique. The journal threshold is no longer used to determine the number of disks
required for journaling.

 Make sure that your journal is employing the most recent defaults. Many journals created
prior to release IBM i 5.4 might still be locked into old settings, and these old settings
might be thwarting performance. One of the easiest ways to ensure that you remain in
lock-step with the best journal settings is to let the system apply its own latest defaults.
You can help ensure that such settings are employed by specifying the
RCVSIZOPT(*SYSDFT) parameter on CHGJRN.
 Consider installing and enabling the journal caching feature if journal performance
(especially during batch jobs) is a major concern in your shop. Make sure to understand
possible impacts on missing receiver writes in case of an outage. For more information
about this topic, refer to Journal Caching: Understanding the Risk of Data Loss,
TIPS0627, at:
http://www.redbooks.ibm.com/abstracts/tips0627.html
 The more actively being-modified objects (such as open files) that you have associated
with a journal, the higher you might want to set your journal recovery count. Failure to do
so slows your runtime performance and increases your housekeeping overhead on your
production system without adding much benefit to your RTO. Increasing this value might
make good sense, but do not get carried away. For more information about this topic, refer
to The Journal Recovery Count: Making It Count on IBM i5/OS, TIPS0625, at:
http://www.redbooks.ibm.com/abstracts/tips0625.html
 If you have physical files that employ a force-write-ratio (the FRCRATIO setting) and those
same files are also journaled, disable the force-write-ratio. Using both a force-write-ratio
and journal protection for the same file yields no extra recovery/survivability benefit and
only slows down your application. It is like paying for two health insurance policies when
one will suffice. The journal protection approach is the more efficient choice.

402 PowerHA SystemMirror for IBM i Cookbook


 If your applications tend to produce transactions that consist of fewer than 10 database
changes, you might want to give serious consideration to use of soft commitment control.
Just make sure to understand that the last committed transaction can not be written to the
journal. For more information about this topic, refer to Soft Commit: Worth a Try on IBM
i5/OS V5R4, TIPS0623, at:
http://www.redbooks.ibm.com/abstracts/tips0623.html
 Both before and after images of each updated record can be written to the journal receiver
depending on the way that the database file journaling is started. For recovery purposes,
we only need the after image. When the IASP is varied on, storage database recovery
applies after images if needed. Consider journaling only if after images are an option.
.

Note: Commitment control automatically activates before images if needed for a


database file involved in a transaction and for which only after images are journaled.

More in-depth treatments of a number of these best practices can be found in Striving for
Optimal Journal Performance on DB2 Universal Database for iSeries, SG24-6286, at:
http://www.redbooks.ibm.com/abstracts/sg246286.html

There are also numerous technotes that are probably more recent than the publication above,
which can be found on the IBM Redbooks publication website.

You might want to try this query URL to find information about Journaling and IBM i:
http://publib-b.boulder.ibm.com/Redbooks.nsf/searchdomain?SearchView&query=[subjec
ts]=AS400+and+Journaling&SearchOrder=1&category=systemi

15.2.2 Journal management effort


It is now much easier than in the past to manage the journal on a library-wide basis. New
commands exist, and certain restrictions existing in the previous releases have disappeared.

Chapter 15. Best practices 403


Automatic journaling for changed objects within a (set of) library
STRJRNLIB allows the system to automatically start journaling for any object in this library that
receives the operation create, move, restore, or all of these (Figure 15-1).

Start Journal Library (STRJRNLIB)

Type choices, press Enter.

Library . . . . . . . . . . . . Name, generic*


+ for more values
Journal . . . . . . . . . . . . Name
Library . . . . . . . . . . . *LIBL Name, *LIBL, *CURLIB
Inherit rules:
Object type . . . . . . . . . *ALL *ALL, *FILE, *DTAARA, *DTAQ
Operation . . . . . . . . . . *ALLOPR *ALLOPR, *CREATE, *MOVE...
Rule action . . . . . . . . . *INCLUDE *INCLUDE, *OMIT
Images . . . . . . . . . . . . *OBJDFT *OBJDFT, *AFTER, *BOTH
Omit journal entry . . . . . . *OBJDFT *OBJDFT, *NONE, *OPNCLO
Remote journal filter . . . . *OBJDFT *OBJDFT, *NO, *YES
Name filter . . . . . . . . . *ALL Name, generic*, *ALL
+ for more values
Logging level . . . . . . . . . *ERRORS *ERRORS, *ALL

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 15-1 STRJRNLIB

By using this command, all objects created in, moved to, and restored in the library will get
automatic journaling start to the same journal, with the same settings. There is no longer a
need to take care of them. However, you have to take care of those objects that you do not
want to journal. End journaling must be run for them.

Note: This command does not take care of existing objects in the library. It only applies to
changes to the library.

404 PowerHA SystemMirror for IBM i Cookbook


A new panel exists through the DSPLIBD command to display the current settings for automatic
journaling of a library. Figure 15-2 shows an example with a journaled library MYLIB.

Display Library Description

Library . . . . . . . : MYLIB Type . . . . . . . . . : TEST

Journaling information:
Currently journaled . . . . . . . . : YES
Current or last journal . . . . . . : MYJOURNAL
Library . . . . . . . . . . . . . : MYLIB
Journal images . . . . . . . . . . . : *AFTER
Omit journal entry . . . . . . . . . : *NONE
New objects inherit journaling . . . : *YES
Inherit rules overridden . . . . . . : NO
Journaling last started date/time . : 09/20/11 17:00:25
Starting receiver for apply . . . . :
Library . . . . . . . . . . . . . :
Library ASP device . . . . . . . . :
Library ASP group . . . . . . . . :

Bottom
F3=Exit F10=Display inherit rules F12=Cancel Enter=Continue
Figure 15-2 Journaling information for MYLIB library

Considerations apply with another method to automatically start journaling, introduced in i5/OS
V5R4, with QDFTJRN data area. Form more information about these considerations and
about STRJRNLIB, refer to Journaling at Object Creation with i5/OS V6R1M0, TIPS0662, at:
http://www.redbooks.ibm.com/abstracts/tips0662.html

Chapter 15. Best practices 405


Start journaling for all or generic files or objects for one or more library
STRJRNPF, STRJRNAP, and STRJRNOBJ have been enhanced to allow the start of journaling for all
or generic files or objects at one time. In the previous release, we had to write a program to
build a list of objects, read this list, and start journaling for each entry of this list. For example,
all database files included in a library, or a set of libraries, can be journaled with one
command (Figure 15-3).

Start Journal Physical File (STRJRNPF)

Type choices, press Enter.

Physical file to be journaled . Name, generic*, *ALL


Library . . . . . . . . . . . *LIBL Name, *LIBL, *CURLIB
+ for more values
*LIBL
Journal . . . . . . . . . . . . Name
Library . . . . . . . . . . . *LIBL Name, *LIBL, *CURLIB
Record images . . . . . . . . . *AFTER *AFTER, *BOTH
Journal entries to be omitted . *NONE *NONE, *OPNCLO
Logging level . . . . . . . . . *ERRORS *ERRORS, *ALL

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys
Figure 15-3 STRJRNPF command

Therefore, for an existing library with four commands, it is possible to start journaling for all
existing applicable objects and for all changes that apply in the future. Example 15-1 shows
an example with library name MYLIB.

Example 15-1 Automatic journaling starting for MYLIB library


STRJRNPF FILE(MYLIB/*ALL) JRN(MYLIB/MYJOURNAL) IMAGES(*AFTER) OMTJRNE(*OPNCLO)
STRJRNOBJ OBJ(MYLIB/*ALL) OBJTYPE(*DTAARA) JRN(MYLIB/MYJOURNAL) IMAGES(*AFTER)
STRJRNOBJ OBJ(MYLIB/*ALL) OBJTYPE(*DTAQ) JRN(MYLIB/MYJOURNAL) IMAGES(*AFTER)
STRJRNLIB LIB(MYLIB) JRN(MYLIB/MYJOURNAL) INHRULES((*ALL *ALLOPR *INCLUDE *AFTER
*OPNCLO))

End journaling locking restriction lifted


We can now end journaling of a physical file or an object, even if the object has been opened
by an application. There is no longer a need to shut down the application to perform this
action. It still remains impossible to do so for files in the midst of a commitment control cycle.
With this update, ENDJRNPF and ENDJRNOBJ now also support generic names and all objects.

Receiver name wrap


In the past, for a user-created journal, when the receiver name got its maximum sequence
(for example, JRNRCV9999), the system was unable to create a new one, and no more

406 PowerHA SystemMirror for IBM i Cookbook


entries could be written through the journal. All application programs were waiting for an
answer to an inquiry message in the QSYSOPR message queue requesting that the user
does something. Change Journal (CHGJRN) command processing has been enhanced to
support receiver name wrapping. This means that if your currently attached receiver name is
at the maximum value, the system will now generate a new receiver with the receiver number
value being wrapped (for example, JRNRCV0000). And if (often due to poor housekeeping
practices) a journal receiver with the name JRNRCV0000 lingers, the system simply
increments the numerical suffix to skip over potential lingering receivers from days gone by
and continues.

Pseudo journal tool


The pseudo journal tool is a standalone set of software. Its purpose is to assist in estimating
the quantity of journal traffic that will ensue that journal protection is enabled for a set of
designated physical files. It is useful to answer these questions:
 How many journals should I configure?
 Will the total quantity of journal/disk traffic justify use of more than one journal?
 Does it make sense for me to configure the journal caching feature on my production
system and, if I do so, how much benefit am I likely to gain for my particular applications?

The nice thing about the pseudo journal tool is that it not only helps answer these questions, it
does so without having a high impact on your system as it performs the analysis. Better yet, it
produces a customized analysis of projected additional disk traffic, tuned to your particular
application and its database reference pattern.

More information regarding the pseudo journal tool along with software to download and a
tutorial can be found on the database tools website:
http://www-03.ibm.com/systems/i/software/db2/journalperfutilities.html

15.3 Best practices for planned site switches


In this section we discuss best practices for planned site switches.

15.3.1 Regular tests


A high-availability solution has no value if it is not tested on a regular basis and if regular
production activities are not conducted on both production and backup sites.

At a minimum, we recommend a switchover every six months. This way, you can run half a
year on one site, switch over to the other site, and start again six months later. A six month
switch frequency is usually enough to make sure that the hardware is ready to handle
production workload when needed. Also make sure that all people are trained to handle an
unplanned failover. Running production on a six-month time window on the same node
makes you comfortable with events that occur with an higher frequency than the week or the
month.

15.3.2 Check cluster and replication health


Each time hat a planned switchover is required, cluster and replication health must
be checked.

Chapter 15. Best practices 407


Nodes included in the planned switchover must be active, the administrative domain must be
active, the related cluster resource group must be active, and recovery domain nodes must
be active.

Note: Cluster resource groups with an inactive status are not handled by the
failover procedure.

Normally, if there is an issue with the replication, the recovery domain node is ineligible
instead of active, which prevents you from switching. Using DSPASPSSN or DSPSVCSSN gives you
the opportunity to double-check the replication status, while using DSPCRGINF allows you to
check the recovery domain node status.

Messages are sent to the cluster message queue, which can be monitored to fix related issues.

An the following message is an example of such a message:


HAI2001 “Cross-site mirroring (XSM) for ASP &1 is not active.”

Fixing this event allows the related recovery node status to change from ineligible to active.

15.3.3 Reverse replication


When running a planned switch, make sure to understand that before switching the node
roles PowerHA reverses the replication. This means that all the updates that you will do on
the new production node will be replicated to the new backup node. For an SVC/V7000
remote Copy Services environment, the switchover reverse replication parameter
(SWTRVSREPL) of the ASP session allows you to determine whether reverse replication is
to be started after the switchover. For a geographic mirroring or DS8000 remote Copy
Services environment, a switchover without starting replication in the reverse direction can be
achieved by detaching the ASP session and manually changing cluster node roles in the
CRG using CHGCRG.

15.3.4 Ending applications using the IASP before a switchover


When a switchover is performed all jobs using the independent ASP are ended abnormally
during its vary-off. Therefore, it is important to properly end all applications using the IASP
before switching over to help reduce application recovery and vary-on times.

Make sure that the interactive job running CHGCRGPRI is not using the IASP anymore because
though the command succeeds, the interactive job gets disconnected otherwise.

15.4 Best practices for unplanned site switches


As for planned switchovers, all cluster resources and ASP sessions must be active. With
IBM i 7.1, new CL commands were introduced to better support CL programming for cluster
automation management by retrieving status information for diverse clustering objects that
you might want to query on a regular basis. These commands are:
 RTVCLU
 RTVCRG
 RTVASPCPYD
 RTVASPSSN
 PRTCADMRE

408 PowerHA SystemMirror for IBM i Cookbook


Note: Cluster resource groups with inactive status are not handled by a failover procedure.

When using a CRG or cluster message queue, the cluster parameters Failover wait time and
Failover default action can be used to either set up an automatic cluster failover or require a
user confirmation for the failover to proceed or be cancelled. If both a CRG and a cluster
message queue have been configured, the cluster message queue takes precedence over
the corresponding CRG settings.

Note: In any case, clustering varies off the IASP on the current production node first when
it detects a failover condition before issuing the failover message.

A cluster partition condition, as described in 15.7, “Resolving a cluster partition condition” on


page 412, requires the user to decide whether the reason for the partition condition should be
fixed or whether a cluster failover should be invoked by changing the cluster node from the
partition to failed status.

In contrast to a planned switch, after an unplanned switch, by default replication is not


automatically started from the new production to the new backup node to still allow for failure
analysis of the original production node data.

15.5 Best practices for reducing IASP vary on times


In this section we describe the main tips that you can follow to be able to reduce the IASP
vary-on times.

15.5.1 Keeping as few DB files in SYSBAS as possible


The system disk pool and basic user disk pools (SYSBAS) should primarily contain operating
system objects, licensed program libraries, and few-to-none user libraries. This structure
yields the best possible protection and performance. Application data is isolated from
unrelated faults and can also be processed independently of other system activity. IASP
vary-on and switchover times are optimized with this structure. Expect longer IASP vary-on
and switchover times if you have a large number of database objects residing in SYSBAS
because additional processing is required to merge database cross-reference information
into the disk pool group cross-reference table.

15.5.2 Synchronizing UIDs/GIDs


In a high-availability environment, a user profile is considered to be the same across
systems if the profile names are the same. The name is the unique identifier in the cluster.
However, a user profile also contains a user identification number (UID) and group
identification number (GID).

This UID and GID are used when looking at object ownership. To reduce the amount of
internal processing that occurs during a switchover, where the IASP is made unavailable on
one system and then made available on a different system, the UID and GID values should be
synchronized across the recovery domain of the device cluster resource group. If this is not
the case, then each object owned by a user profile with a non-matching UID needs to be
accessed, and the UID needs to be changed as part of the vary-on process of the IASP after

Chapter 15. Best practices 409


each switchover or failover. Synchronization of user profiles including IUD and GID can be
accomplished by using the administrative domain support.

15.5.3 Access path rebuild


To reduce the amount of time that an independent ASP vary-on waits for access path
rebuilds, consider having the rebuilds performed in the background after the vary-on has
completed. This is determined by the RECOVER attribute of a logical file. If RECOVER(*IPL)
is specified, the vary-on will wait for the rebuild to complete when one is necessary. If
RECOVER(*AFTIPL) is specified, the vary-on does not wait and the AP is built in the
background after the vary-on is complete.

Access paths are not available while they are being rebuilt. Therefore, if there are specific
logical files that an application requires to be valid shortly after a vary-on is complete, the user
should consider specifying RECOVER(*IPL) to avoid a situation in which jobs will try to use
the them before they are valid.

Another option to consider is the journaling of access paths so that they are recovered from
the journal during a vary-on and do not need to be rebuilt. Access paths are journaled
implicitly by SMAPP, as discussed in 15.5.4, “System Managed Access Path Protection” on
page 410. To ensure that specific, critical access paths are journaled, they should be explicitly
journaled.

When the system rebuilds an access path after IPL time, you can use EDTRBDAP to modify
rebuild sequences to select those access paths that need to be rebuilt now. But, by default,
this command applies only to SYSBASE. To allow it to work with independent ASP access
paths, do the following:
1. Enter the following command, where YY is the IASP number:
CRTDTAARA DTAARA(QTEMP/QDBIASPEDT) TYPE(*DEC) LEN(4) VALUE(YY)
2. Enter EDTRBDAP when the iASP becomes ACTIVE or AVAILABLE. It might be necessary to
invoke the command multiple times while the iASP is being varied on and is ACTIVE. A
CPF325C message will be sent until the command is allowed to be used.
3. Enter the following command to delete the data area:
DLTDTAARA DTAARA(QTEMP/QDBIASPEDT)

15.5.4 System Managed Access Path Protection


Analyze your System Managed Access Path Protection (SMAPP) setting and ensure that you
are not locked into an outdated setting inherited from the last decade. A SMAPP setting that
is too high can significantly increase your recovery duration and thereby cause you to miss
your RTO. SMAPP is a form of behind-the-scenes journaling. If you see a SMAPP setting
larger than, for example, 50 minutes, give serious consideration to lowering the value. (An
original default setting nearly a decade ago was 150 minutes, and many shops that have not
revisited this setting as hardware speeds have improved might still be operating with outdated
settings. Continuing to do so can make the vary-on duration for an IASP exceed your RTO.)

SMAPP applies to the IASP vary-on (or IPL for the SYSBASE) step, which is responsible for
rebuilding access paths in case of damages after an outage. For access paths (or indexes in
the SQL world) dependant on large physical files (or tables in the SQL world), it might take a
considerable amount of time, which can be several hours. To avoid rebuilding the access
paths, SMAPP uses existing journals for access path updates behind-the-scene recording,

410 PowerHA SystemMirror for IBM i Cookbook


just like if the access paths were journaled. They are recorded so that IASP vary-on step can
use them to update access paths in place of rebuilding them.

Access paths update recording occurs automatically at each ASP level (independent or not,
and including SYSBASE) depending on the following parameters:
 Access path recovery time target for each ASP.
 Access path recovery estimate for each ASP. Each access path for which the estimated
rebuild time is higher than the target will be protected by SMAPP.

SMAPP effect the overall system performance. The lower the target recovery time that you
specify for access paths, the greater this effect can be. Typically, the effect is not noticeable,
unless the processor is nearing its capacity.

In the Figure 15-4, you can see an example of an existing independent ASP installation.

There is no target for any Independent ASP. Therefore, the system, by itself, has started
access path journaling to achieve the overall system target, which is set to 50 minutes. Using
the F14 key, we can see which access paths are currently protected (Figure 15-4).

For this example, the user should review the targets to specify a better recovery time target,
for example, by specifying *MIN, which means the lower possible recovery time.

Display Recovery for Access Paths SYSTEMA


05/10/11 21:15:58
Estimated system access path recovery time . . . : 44 Minutes
Total not eligible recovery time . . . . . . . . : 0 Minutes
Total disk storage used . . . . . . . . . . . . . : 909,073 MB
% of disk storage used . . . . . . . . . . . . . : 0,035
System access path recovery time . . . . . . . . : 50
Include access paths . . . . . . . . . . . . . . : *ALL

------Access Path Recovery Time------ -----Disk Storage Used-----


ASP Target Estimated Megabytes ASP %
1 *NONE 1 34,611 0,009
IASP1 *NONE 9 265,965 0,022
IASP2 *NONE 33 608,497 0,059

Bottom
F3=Exit F5=Refresh F12=Cancel F13=Display not eligible access paths
F14=Display protected access paths F15=Display unprotected access paths
Figure 15-4 DSPRCYAP command result

Chapter 15. Best practices 411


Display Protected Access Paths SYSTEMA
05/10/11 21:21:06
Estimated
Recovery
File Library ASP If Not Protected
OSASTD10 M3EPRD IASP1 00:03:03
OSASTD90 M3EPRD IASP1 00:02:57
OSASTD00 M3EPRD IASP1 00:02:55
QADBIFLD QSYS00033 IASP1 00:02:50
MITTRA30 MVXCDTA800 IASP2 00:02:00
MITTRA35 MVXCDTA800 IASP2 00:01:55
OODOCU00 MVXCDTA800 IASP2 00:01:53
MITTRAZ9 MVXCDTA800 IASP2 00:01:50
MITTRA50 MVXCDTA800 IASP2 00:01:49
QADBIFLD QSYS00034 IASP2 00:01:49
MITTRA20 MVXCDTA800 IASP2 00:01:48
MITTRA70 MVXCDTA800 IASP2 00:01:48
MITTRA60 MVXCDTA800 IASP2 00:01:45
MITTRAU6 MVXCDTA800 IASP2 00:01:41
MITTRA90 MVXCDTA800 IASP2 00:01:41
MITTRA80 MVXCDTA800 IASP2 00:01:40
More...
F3=Exit F5=Refresh F12=Cancel F17=Top F18=Bottom

Figure 15-5 Protected access paths

15.6 Switching mirroring while detached


When mirroring is detached, you still have the opportunity to switch, which means changing
the node role. The usual CHGCRGPRI command cannot be used at this time because the
backup node has become ineligible.

This is the appropriate way to do a switch while detached:


1. Change cluster resource group node roles with CHGCRG.
2. Reattach the ASP session from the current backup node to re-establish remote mirroring
from the current production to the backup node.

Refer to “Switching while detached” on page 366.

15.7 Resolving a cluster partition condition


A cluster partition condition occurs when nodes in a cluster can no longer communicate with
each other (that is, they no longer receive heartbeat information). Therefore, clustering
cannot distinguish between a communication failure or a real node failure. You cannot switch
to a node while it is in partition status.

To avoid some of these partition conditions, the HMC managing the production node server
or the VIOS managing the production node partition can send information about the status of

412 PowerHA SystemMirror for IBM i Cookbook


the partition through the advanced node failure detection mechanism. When it is really in a
not running status, the backup node receives the information and automatically starts a
failover procedure. Refer to 11.2, “Setting up cluster monitors” on page 208, for more
information.

Make sure that you analyze the cause of any partition condition when deciding to either
resolve it or start a manual cluster failover. It could be due to a temporary network issue,
which can be easily corrected. As soon as the temporary failure is corrected, on a 15-minute
cycle the partition status is automatically solved and the node status changes from partition
back to active.

If you really need to start a failover use CHGCLUNODE to change the node status from partition
to failed.

See “Metro Mirror and Global Mirror failover and failback” on page 350, for examples of
these scenarios.

15.8 IBM i hosting environments


IBM i hosting environments can be used to deploy either IBM i, AIX, or Linux partitions, or
they can be used to provide storage for IBM Blade servers or IBM System x® servers
attached to IBM i using iSCSI. In all these cases, storage is provided to the hosted
environment using network server storage spaces and a network server description. As these
network server storage spaces from an IBM i perspective are simply objects located in the
IFS, it is possible to move these network server storage spaces into an IASP and therefore
use IBM PowerHA SystemMirror for i to provide high availability for these hosted
environments. This is not recommended for large IBM i, AIX, or Linux partitions, but is an
option for smaller environments.

There are, however, a few additional steps that need to be considered if using a
hosted environment:
 Although the network server storage space can reside in an IASP, the network server
description and network server configuration objects needed for iSCSI cannot. Therefore,
make sure to use the administrative domain to keep the network server descriptions
synchronized between all nodes in your cluster.
 For hosted IBM i, AIX, or Linux partitions, virtual SCSI adapters have to be defined on the
production system as well as on the backup system to connect a network storage space to
the hosted LPAR.
 For a planned switch, make sure to first end all activities in your hosted environment. Then
power down your hosted environment and end the network server description before
issuing CHGCRGPRI.
 After a planned switch or a failover, vary on the network server description on the backup
system and start your hosted environment on the backup system.
 The following white paper provides sample exit programs that can be used to automate
the vary-off and vary-on processes for network server descriptions:
http://www-03.ibm.com/systems/resources/systems_i_os_aix_pdf_linux_aix_xsm_final
.pdf

Perform a switchover test before moving the environment to production to make sure that the
setup does work on the backup site (that is, that you have done the necessary configuration
steps on the backup site correctly).

Chapter 15. Best practices 413


15.9 Upgrading IBM i and PowerHA release
If your business can afford to power down your system, then you can upgrade both systems
to the new version of the operating system at the same time.

If you need to run continuous operations and are confident with the release upgrade, you can
do a so-called rolling upgrade. During this process (described in 15.9.1, “Performing a rolling
upgrade in a clustering environment” on page 414) either the production or the backup
system is available to the user. Be aware though that there are periods of time when you do
not have a backup system that you could fail over to if your current production system fails.

If you like to test your applications on one system using the new release while keeping the
application environment intact on the other system (so that in case of problems you can
simply move back to using this environment with the old version of the operating system), do
the release upgrade on the backup system after detaching the IASP.

15.9.1 Performing a rolling upgrade in a clustering environment


When doing a rolling upgrade you first have to upgrade your backup system to the new
version of the operating system. The reason for this is that you can switch an independent
auxiliary storage pool (iASP) from a lower release to a higher release, but not from a higher
release down to a lower release, so you cannot switch back to lower release node after the
IASP has been varied on at the upgraded node.

The general order of steps when performing IBM i and PowerHA release upgrades is
as follows:
1. If you are using geographic mirroring, you can suspend it with tracking.
2. Upgrade IBM i and PowerHA on the current backup node and restart the backup
cluster node.
3. If you are using geographic mirroring and suspended it, resume it to synchronize the IASP
from the current production to the current backup node.
4. Switch from production to the upgraded backup node.
Independent ASP database conversion from one release to another occurs when the
IASP is first varied on.
5. If you are using geographic mirroring, you can suspend it with tracking.
6. Upgrade IBM i and PowerHA on the current backup node and restart the backup
cluster node.
7. If you are using geographic mirroring and suspended it, resume it to synchronize the IASP
from the current production to the current backup node.
8. When all cluster nodes are on the same release, update the PowerHA version
with CHGCLUVER.

Note: CHGCLUVER might need to be used twice to update the PowerHA version if you
skipped a release at the upgrade, for example, from V5R4 to i 7.1.

9. Optional: Switch back to your preferred production node.


10.You might want to review your cluster administrative domain monitored resource
entries for additional objects supported by the new release in a cluster administrative
domain or IASP.

414 PowerHA SystemMirror for IBM i Cookbook


15.9.2 Performing release upgrade while retaining current production system
Using the “switching while detached” mechanism allows you to use the backup node for
production activity after the release upgrade while preserving the production node data.

With this option, the following steps apply:


1. If you are using geographic mirroring, you can suspend it with tracking.
2. Upgrade IBM i and PowerHA on the current backup node and restart the backup
cluster node.
3. If you are using geographic mirroring and suspended it, resume it to synchronize the IASP
from the current production to the current backup node.
4. Perform a detach operation for your ASP session. At this time the independent ASP on
both sides can be varied on.
5. Allow the users to connect to the upgraded node, which is still the backup node from a
cluster standpoint. If problems exist, you can then route user connections to the normal
production node, with data current at the time of the detach operation.
6. If there are no or minor problems on the backup node, you can decide to make this node
the new production one in preparation for upgrading the previous production node. Use
CHGCRG ... RCYDMNACN(*CHGCUR) to switch node roles.
7. Upgrade IBM i and PowerHA on the current backup node and restart clustering. Perform a
reattach operation on the current backup node to synchronize current production data on
the current backup node. Note that this option submits a full synchronization process for
geographic mirroring as both IASP copies had been varied on.
8. When all cluster nodes run the same release, update the PowerHA version with
CHGCLUVER.

Note: CHGCLUVER might need to be used twice to update the PowerHA version by two if
you skipped a release at the upgrade, for example, from V5R4 to i 7.1.

9. Optional: Switch back to your preferred production node.


10.You might want to review your cluster administrative domain monitored resource entries
for additional objects supported by the new release in a cluster administrative domain
or IASP.

15.10 Integration with BRMS


This section describes one way to use BRMS for saving to tapes operations in order to keep
track of operations in the current BRMS production database, even they effectively run on a
backup node. We use the possibility for BRMS to have its own system name distinct from the
network attribute one. Refer to this Knowledge Base document for more information about
BRMS system name:

http://www-912.ibm.com/s_dir/SLKBase.nsf/a9b1a92c9292c2818625773800456abc/ff759abf
1f796e8e8625715d00167c07?OpenDocument

Chapter 15. Best practices 415


15.10.1 Normal BRMS usage
You might prefer running the save-to-tape operations on the backup node and keeping save
information here (that is, in the backup BRMS database) even for production IASP copies
saves. In this case, there is no difference from a standard BRMS implementation and,
therefore, this section does not apply.

However, you will have to deal with questions like “What was the system name for the last
backup tape for this file I need to restore now?” because saves to tapes can potentially run on
production and backup nodes.

BRMS network configuration can help you. Make sure to share the media information at the
library level. With this setting, to find the last backup for your file you can run the following
command on each system of your production node, which can run the save to tapes:
WRKMEDIBRM LIB(MYLIB) ASP(MYASP) FROMSYS(MYSYS)

To restore the desired object (assume that the last save was done on the MYSYS system),
set your job to use the appropriate IASP, then run the following command:
RSTOBJBRM OBJ(MYFILE) SAVLIB(MYLIB) DEV(*MEDCLS) SAVLVL(*CURRENT) FROMSYS(MYSYS)

There is another disadvantage with this usage. BRMS does not allow appending backups to
tape owned by another BRMS system, and retention calculation is done at the BRMS system
level. This means that you might need much more media than with the proposed
implementation discussed later.

15.10.2 Production IASP copy save-to-tapes on backup nodes


If you want BRMS to run save-to-tape on a backup node for production IASP copies and to
update the production BRMS database just like saves that were done on the PRODUCTION
system, there are considerations to take care of when using a BRMS network, using a BRMS
name, and sharing tapes and locations.

In any case, when using a BRMS network, we recommend using a dedicated IP addresses
for BRMS.

Suppose that we have three systems in a BRMS network.


 PRODUCTION is the system name that runs production activities as a preferred role. The
PRODUCTION name is longer than allowed (eight characters maximum), but this name is
only for description purposes.
 BACKUP is the system name of the current and preferred backup node with an active
ASP mirroring. BACKUP is also used for save-to-tape operations, for example, after
detaching the ASP session to make the target ASP available. Or BACKUP can also be
another node using a FlashCopy target IASP.
 OTHER is a third system.

In this specific case, when a save operation is scheduled to run on a system with a BRMS
name different from the system name, other BRMS network members must be aware of the
save operation. They need to know the current tape usage by the system running the save, in
place of the usual system and they need to update the BRMS database of the system running
the save.

When there is no save operation for the usual production system, the BRMS OTHER system
must connect to the BRMS PRODUCTION system to be aware of real-availability tapes, and it
must update the PRODUCTION BRMS database when it writes to available tapes.

416 PowerHA SystemMirror for IBM i Cookbook


Figure 15-6 shows the BRMS communication flow. The BRMS OTHER system
communicates with the BRMS PRODUCTION system.

ASP Mirroring active

Role
Role BACKUP
PRODUCTION
Po wer 7 5 0
D3 D4 D5 D6 D7 D8 D9 D1 0
Po wer 7 5 0
D3 D4 D5 D6 D7 D8 D9 D1 0

Node name Node name


PRODUCTION BACKUP

BRMS network

Po wer 7 5 0
D3 D4 D5 D6 D7 D8 D9 D1 0

OTHER

Figure 15-6 BRMS network when no save operation is running and roles are the preferred ones

Chapter 15. Best practices 417


When running a save-to-tape for PRODUCTION ASP copies, the BRMS OTHER system
must no longer communicate with the BRMS PRODUCTION system but instead with the
BACKUP system because it is now acting like it is PRODUCTION for the BRMS activity. As
shown in Figure 15-7, the easiest way to perform this swap, without changing anything on the
OTHER system, is to swap the BRMS PRODUCTION IP address on the BACKUP system.

ASP session detached


Or flashcopy usage

Role
Role BACKUP
PRODUCTION
Po wer 7 5 0
D3 D4 D5 D6 D7 D8 D9 D1 0
Po wer 7 5 0
D3 D4 D5 D6 D7 D8 D9 D1 0

Node name Node name BRMS name


PRODUCTION BACKUP PRODUCTION

BRMS network

Po wer 7 5 0
D3 D4 D5 D6 D7 D8 D9 D1 0

OTHER

Tape Library

Figure 15-7 BRMS network when running a save-to-tape operation on the backup node and the roles are preferred ones

418 PowerHA SystemMirror for IBM i Cookbook


These are the steps to update the PRODUCTION BRMS database when running this ASP
save on the BACKUP system:
1. Make independent ASP available on the BACKUP system by using the ASP session with
the detach option or FlashCopy procedures.
2. On the PRODUCTION system, we need to send BRMS information to the
BACKUP system:
a. End the Q1ABRMNET subsystem.
b. End the BRMS IP interface.
c. Save the QUSRBRM library to a save file and send it to the BACKUP system.
3. Make sure to understand that, from this point, any update to the PRODUCTION BRMS
database will be lost. This means that you should not run any save or media operation on
PRODUCTION from this time on. The restore operation can be run, but related BRMS
history will be lost.
4. On the BACKUP system, we need to save the current BRMS information:
a. End the Q1ABRMNET subsystem.
b. End the BACKUP BRMS IP interface.
c. Save the QUSRBRM library to a save file.
d. Clear the QUSRBRM library.
5. On the BACKUP system, we want the BRMS database to be the PRODUCTION one:
a. Restore QUSRBRM objects from the PRODUCTION save file.
b. Change the BRMS name to PRODUCTION:
i. Using the BRMS Q1AOLD API with this command:
QSYS/CALL QBRM/Q1AOLD PARM('BRMSYSNAME' '*SET ' 'PRODUCTION').
This command creates the QUSRBRM/Q1ABRMSYS data-area.
ii. The BACKUP system acts now as the PRODUCTION BRMS system.
c. Start the PRODUCTION BRMS IP interface.
d. Start the Q1ABRMNET subsystem.

Chapter 15. Best practices 419


6. On the BACKUP system, submit the save-to-tape for independent ASP libraries and
directories. Figure 15-8 shows an example of a control group.

Create Backup Control Group Entries DEMOHA

Group . . . . . . . . . . : IASP1
Default activity . . . . . *BKUPCY
Text . . . . . . . . . . . *NONE

Type information, press Enter.

Weekly Retain Save SWA


Backup List ASP Activity Object While Message Sync
Seq Items Type Device SMTWTFS Detail Active Queue ID

10 *ALLUSR IASP1 *DFTACT *ERR *NO


20 *LINK IASP1 *DFTACT *NO *NO

Bottom
F3=Exit F5=Refresh F10=Change item F11=Display exits
F12=Cancel F14=Display client omit status F24=More keys

Figure 15-8 BRMS control group for saving IASP1 objects only

Note: SYSBASE objects can also be included in this backup if they are included in the
administrative domain and, therefore, are identical on both PRODUCTION and
BACKUP nodes.

7. On the BACKUP system, we prepare the comeback for the BRMS environment:
a. End the Q1ABRMNET subsystem.
b. End the PRODUCTION BRMS IP interface.
c. Delete the QUSRBRM/Q1ABRMSYS data area.
8. On the BACKUP system, we need to send the BRMS database to the
PRODUCTION system:
a. Save the QUSRBRM library to a save file.
b. Send it to the PRODUCTION system.
9. On the PRODUCTION system, we need to restore the BRMS database updated with the
save-o-tape operation run on the BACKUP system. If any update has been done to the
PRODUCTION BRMS database before this point, it is lost. Take these steps:
a. Clear the QUSRBRM library.
b. Restore objects from the save done in step 8.
10.On the PRODUCTION system, start the BRMS IP interface and the Q1ABRMNET
subsystem. Any BRMS save and tape information regarding IASP is now on the
PRODUCTION system, and from now on this one is used within the BRMS network by
other members.

420 PowerHA SystemMirror for IBM i Cookbook


11.On the BACKUP system, we need to set back the usual BRMS BACKUP database so that
we can run SYSBASE BACKUP-specific saves to tape:
a. Clear the QUSRBRM library.
b. Restore objects from the save taken in step 4 on page 419.
c. Start the BACKUP BRMS IP interface.
d. Start the Q1ABRMNET subsystem.
12.If the IASP was made available by changing an ASP session with a detach option, it can
be reattached at this time.

Note: BRMS can have only one system name at any time. This means that you cannot run
several saves to tapes at the same time for IASP copies from several nodes. You can still
use a single node for saves to tape of IASP copies from several nodes, but they must be
done in a serial way. You have to run the steps described in this section for each production
system, one after the other.

15.10.3 Run production IASP copy saves to tapes after a roles switch
The same concern exists after a role switch (that is, when the preferred backup node
becomes the production node and the preferred production node becomes that backup node,
but they keep their system name). You will probably want to run the save-to-tape operations
on the actual backup node.

Chapter 15. Best practices 421


After roles switch, when there is no save in progress, the BRMS network is as shown in
Figure 15-9.

ASP Mirroring active

Role
Role BACKUP
Po wer 7 5 0
D3 D4 D5 D6 D7 D8 D9 D1 0
PRODUCTION
Po wer 7 5 0
D3 D4 D5 D6 D7 D8 D9 D1 0

Node name BRMS name Node name BRMS name


PRODUCTION BACKUP BACKUP PRODUCTION

BRMS network

Po wer 7 5 0
D3 D4 D5 D6 D7 D8 D9 D1 0

OTHER

Figure 15-9 BRMS network when no save operation is running and roles are switched ones

422 PowerHA SystemMirror for IBM i Cookbook


When running a save-to-tape from preferred production node while roles are switched ones,
the network is as shown in Figure 15-10.

ASP session detached


Or flashcopy usage

Role
Role BACKUP
Po wer 7 5 0
D3 D4 D5 D6 D7 D8 D9 D1 0
PRODUCTION
Po wer 7 5 0
D3 D4 D5 D6 D7 D8 D9 D1 0

Node name BRMS name Node name


PRODUCTION PRODUCTION BACKUP

BRMS network

Po wer 7 5 0
D3 D4 D5 D6 D7 D8 D9 D1 0

OTHER

Tape Library

Figure 15-10 BRMS network when running a save-to-tape operation on the backup node and the roles are switched ones

After switching roles, when there is no BRMS activity, BRMS names must be swapped, and
related IP interfaces as well:
 On a production role node:
a. Start the BRMS production IP interface.
b. Change the BRMS system name to PRODUCTION with Q1AOLD API, as
shown above.
 On backup role node:
a. Start the BRMS backup IP interface.
b. Change the BRMS system name to BACKUP with the Q1AOLD API, as shown above.

Chapter 15. Best practices 423


Take these steps to update the PRODUCTION BRMS database when running this ASP save
on the current backup node (preferred production node):
1. Make the independent ASP available on the backup by using the ASP session with detach
option or FlashCopy procedures.
2. On the production role node, we need to send BRMS information to backup role node:
a. End the Q1ABRMNET subsystem.
b. End the production BRMS IP interface.
a. Save the QUSRBRM library to a save file and send it to the backup role node.
3. Make sure to understand that, from this point, any update to the PRODUCTION BRMS
database will be lost. This means that you should not run any save or media operation on
the production role node from this time. A restore operation can be run, but any related
BRMS history will be lost.
4. On the backup role node, we need to save the current BRMS information:
a. End the Q1ABRMNET subsystem.
b. End the backup BRMS IP interface.
c. Save the QUSRBRM library to a save file.
d. Clear the QUSRBRM library.
5. On the backup role node, we want the BRMS database to be the PRODUCTION one:
a. Restore QUSRBRM objects from the production save file.
b. The backup role node acts now as production BRMS system because of
QUSRBRM/Q1ABRMSYS data area content.
c. Start the production BRMS IP interface.
d. Start the Q1ABRMNET subsystem.
6. On the backup role node, submit the save-to-tape for independent ASP libraries and
directories. Figure 15-8 on page 420 shows an example of a control group.
7. On the backup role node, we prepare the comeback for the BRMS environment:
a. End the Q1ABRMNET subsystem.
b. End the production BRMS IP interface.
8. On the backup role node, we need to send the BRMS database to the production
role node:
a. Save the QUSRBRM library to a save file.
b. Send it to the production role node.
9. On the production role node, we need to restore the BRMS database updated with the
save-to-tape operation run on the backup role node. If any update has been done to the
PRODUCTION BRMS database before this point, it is lost. Take these steps:
a. Clear the QUSRBRM library.
b. Restore objects from the save taken in step 8.
10.On the production role node, start production the BRMS IP interface and the
Q1ABRMNET subsystem. Any BRMS save and tape information regarding IASP is now
on the production role node, and from now on this one is used again within the BRMS
network by other members.

424 PowerHA SystemMirror for IBM i Cookbook


11.On the backup role node, we now need to set back the usual BRMS database to be able
to run SYSBASE-specific saves to tape:
a. Clear the QUSRBRM library.
b. Restore objects from the save taken in step 4 on page 424.
c. Start the backup BRMS IP interface.
d. Start the Q1ABRMNET subsystem.
12.If the IASP was made available with changing an ASP session with the detach option, it
can be reattached at this time.

15.10.4 Specific system synchronization


Another way to run saves to tape on a backup system and make the production system own
the backup history just like it did it is to use specific system synchronization. More information
can be found on the BRMS website:
http://www-03.ibm.com/systems/i/support/brms/new.html#foo7

We decide on the backup system (that is, the one that runs the saves to tapes), which the
production system will look like it did it itself. Configuration steps are done either through the
BRMS GUI or with the QBRM/Q1AOLD API.

To add a specific system synchronization, change the system name to make it look like the
backup was done by this system and synchronize the reference date/time using this command:
CALL QBRM/Q1AOLD PARM('HSTUPDSYNC' '*ADD' ‘SYSTEM’ ‘NETWORK ID' 'IASPNAME'
'*CHGSYSNAM')

To add a specific system synchronization, keep the name of who did the backup and
synchronize the reference date/time:
CALL QBRM/Q1AOLD PARM('HSTUPDSYNC' '*ADD' ‘SYSTEM' ‘NETWORK ID' 'IASPNAME '
'*NORMAL')

15.10.5 User-defined IASP timestamps


When running saves to tapes on backup systems, after a FlashCopy or after detaching an
ASP session, it might be interesting to supply a user-defined timestamp for the backup. A
FlashCopy or detach operation can occur at one time, and specific save-to-tape operations
later, but you want to keep the FlashCopy or detach operation time for the save reference,
and the reference point for future incremental backups. Still using the QBRM/Q1AOLD API, it
is possible to define a timestamp for IASP save to tape.

To add a timestamp, use this command:


CALL QBRM/Q1AOLD PARM('FLASHTIME' '*ENABLE' 'IASPNAME' 'FILESYSTEM TYPE'
'timestamp')

Refer to the same website as given in 15.10.4, “Specific system synchronization” on


page 425, for more details.

15.10.6 SYSBASE save-to-tape considerations


We can consider SYSBASE objects included in the administrative domain in the same way
as IASP copies. This means that saving them to tape can be run at the same time that the
IASP is saved to tape and that they belong to the production role node wherever they are
saves to tape.

Chapter 15. Best practices 425


All other SYSBASE objects not included in the administrative domain belong to the node on
which they are installed, whatever its role. This means that they have to be saved while the
BRMS system name is not overridden, when it is the same as the network attribute system
name. This means, for example, that when you are running the preferred configuration, you
have to run SYSBASE-specific backup node saves-to-tape outside of a regular
production-related control group.

15.11 Hardware replacement in a PowerHA environment


In this section we describe the considerations for replacing either an IBM i server or storage
system in an IBM PowerHA SystemMirror for i environment.

15.11.1 Server replacement


Depending on the PowerHA System Mirror solution that you implemented, replacing one or
two of the servers in your high availability environment involves a number of steps, which we
discuss in this section.

Geographic mirroring
When replacing the backup system, including its disks, in an environment using geographic
mirroring with internal disks you cannot simply do a save and restore operation. Perform
these steps:
1. To preserve your old backup system in case migration to the new backup system fails,
perform a detach of the ASP session using CHGASPSSN OPTION(*DETACH).
2. Power down the old backup system.
3. Deconfigure geographic mirroring from the production system using either the GUI
interfaces or CFGGEOMIR. This results in an error message stating that the backup system
cannot be found. You can tell the system to ignore this status and to proceed with the
deconfiguration. Be aware that your production IASP needs to be varied off to perform
this step.
4. Start the new backup system. The system should have unconfigured drives available to
become the new backup IASP. Make sure that clustering works between the production
system and the new backup system. It might be necessary to remove and add the backup
cluster node to and from the cluster and recovery domain.
5. Configure geographic mirroring from the production system to the new backup system
using either the GUI interfaces or CFGGEOMIR. Be aware that your production IASP needs
to be varied off to perform this step.
6. When the configuration of geographic mirroring is finished, vary on the production IASP
and make sure that geographic mirroring is active. A full resynchronization is required.

Exchanging the production system without first deconfiguring geographic mirroring and
reconfiguring it afterwards is also not possible. Consider doing a CHGCRGPRI to switch over to
the backup system, and then follow the steps described above.

To replace a backup server only in a geographic mirroring environment using external


storage, make sure to suspend geographic mirroring from the production site first. Then
power down the old backup server, attach the new server to the existing external storage, and
restart the new backup server. Finally, resume geographic mirroring to run a partial
synchronziation.

426 PowerHA SystemMirror for IBM i Cookbook


When replacing the production server only in a geographic mirroring environment using
external storage, either switch to the backup system first or make sure to properly end your
production system (vary off the IASP, ENDCRG, ENDCLUNOD, before PWRDWNSYS)
before exchanging the server hardware.

Remote Copy Services environments


When using IBM PowerHA System Mirror for i with remote Copy Services for either DS6000,
DS8000, or SVC/V7000, consider this:
 Consider exporting your old server LPAR configuration to a system plan to have current
documentation available for either manual or automatic redeployment by importing the
(modified) system plan. Make sure to set up your partitions in the same way that they were
set up on the old server. Use specific care for virtual adapter numbers when using VIOS.
 Consider exchanging the server results in new WWPNs for server Fibre Channel
adapters. On DS6000 and DS8000, host connections have to be deleted and recreated.
On SVC/V7000, first add the new host ports and then delete the old ones, preserving
existing vdisk mappings to the host.

Make sure to change your SAN switch zoning for the new server WWPNs.

15.11.2 Storage system replacement


If a remote copy secondary storage system needs to be replaced, plan for these actions:
 Make sure that the IBM i cluster node accessing the secondary storage system is the
current backup node.
 If this IBM i backup node has its SYSBAS configured on the secondary storage system,
perform an entire system save for this IBM i node before powering it down. Otherwise, just
end the remote copy ASP sessions.
 Stop remote mirroring on the storage system.
 Perform the secondary storage system replacement procedure.
 Change any SAN switch zoning information for the new WWPNs from the new secondary
storage system.
 Re-create the required logical storage configuration for host ports/connections and
volumes on the new secondary storage system.
 If the IBM i backup node had its SYSBAS configured on the secondary storage system,
restore SYSBAS for this IBM i node.
 If this IBM i backup node was powered down before, restart the IBM i backup node
connected to the new secondary storage system with restarting clustering.
 Re-establish and start the remote copy paths and IASP volume relationships between the
primary and new secondary storage system. This requires a full-resynchronization
between primary and secondary volumes.
 Change the ASP copy descriptions for the secondary storage system. Even if the logical
configuration did not change, at least the serial number information needs to be updated
for the new storage system.
 Start the remote copy ASP sessions again.

If a DS8000 or SVC/V7000 remote copy secondary storage system is not replaced but needs
to be powered-cycled for a disruptive maintenance action, the primary storage system keeps
track of the changes such that only a partial resynchronization of the out-of-sync tracks will be
required after the secondary storage system becomes operational again. Usually this

Chapter 15. Best practices 427


resynchronization would have to be started manually for both DS8000 and SVC/V7000. An
exception is DS8000 Global Mirror, which automatically resumes suspended Global Copy
relationships and restarts the Global Mirror session.

15.12 Problem data collection


Prior to taking the steps to gather data, ensure that the HA environment is always up to date
with PTFs, which can be found on the Recommended Fixes website. There are PTFs added
frequently that address various HA-related issues. Also, when missing the suggested PTFs,
data capture can also sometimes have difficulties. See 8.1.3, “Power Systems requirements”
on page 135, for the PTF requirements.

Note: If you are using the manual method of collecting data, you must collect it on all
nodes of your cluster.

15.12.1 IBM i clustering


In this section we discuss IBM i clustering.

Node not starting


A node in the cluster cannot be started. Using STRCLUNOD results in an error message or
hangs indefinitely. Take these steps:
1. Check the joblog of the process in which STRCLUNOD was run.
If there is CPFBB98, it gives reason codes for the problems with a node start.
2. Check the QCSTCTL, QCSTCRGM, or the actual CRG joblog.
3. Check the QHASVR job (PowerHA for i 7.1 and later). Check the joblog if there are any
problems.
4. Make sure that user QCLUSTER has no jobs in the QBATCH jobqueue waiting to be
started. Usually, the QBATCH job queue has only one maximum job. Check this on
all nodes.

Cluster resources group or administrative domain


If there is problem with the cluster resource group or the administrative domain, check that
the following information is true:
 The jobs are named the same as the cluster resource groups. Its jobs should not be
running in any of the subsystems.
 The job are named the same as the administrative domain.
 Check the QCSTCTL and QCSTCRGM jobs. These are the system jobs not running in
any of the subsystems.

Problem data collection to be sent to IBM


In this section we discuss problem data collection to be sent to IBM.

Tip: IBM Support might instruct you to collect additional data not included in these
instructions, or they might require less data then listed here. Contact IBM Support with
your problem and you will get the instructions for your specific case.

428 PowerHA SystemMirror for IBM i Cookbook


The problem data collection that needs to be sent to IBM can be gathered by using QMGGTOOLS
(see the 15.12.3, “The Must Gather Data Collector” on page 438) or manually. Using
QMGTOOLS is the preferred way to collect the diagnostic data for the HA solution problems. In
the following sections we describe both of these ways to collect the data.

Collecting the data with QMGTOOLS


To collect the data with QMGTOOLS, you need to collect general data on one of your nodes,
following the instructions in “Collecting the general cluster data for all nodes” on page 439.
Then collect on each of the nodes QSYSOPR and QHST messages for the time of failure.

Collecting the data manually


To prepare data to be sent to IBM for problem analysis, take the following steps on all of the
nodes in your cluster:
1. Run this command:
DSPCLUINF CLUSTER(cluster_name) DETAIL(*FULL) OUTPUT(*PRINT)
2. Run this command:
DSPCRGINF CLUSTER(cluster_name) CRG(*LIST) OUTPUT(*PRINT)
3. Run this command:
DSPCRGINF CLUSTER(cluster_name) CRG(cluster_resource_group_name) OUTPUT(*PRINT)
4. Joblogs of QCSTCTL, QCSTCRGM, and the joblogs for the Cluster Resource Group
(CRG) and administrative domain: These are named same as CRG and Administrative
Domain will be required.
5. Run this command:
DMPCLUTRC CLUSTER(cluster_name) CRG(*ALL) NODE(*ALL) LEVEL(*ERROR)
6. Run this command for each copy description:
DSPASPCPYD ASPCPY(asp_copy_description) OUTPUT(*PRINT)
7. Run this command for each session:
DSPASPSSN SSN(session_name) OUTPUT(*PRINT)
8. Collect the QSYSOPR and QHST messages for the time of failure.

Chapter 15. Best practices 429


9. Start SST (strsst), log in, and collect the following data:
a. LIC Logs (strsst → 1. Start a service tool → 5. Licensed Internal Code log → 2. Dump
entries to printer from the Licensed Internal Code log).
On page shown in Figure 15-11, set these options:
• Dump option 1 and option 3. As a result you get two spool files.
• Set the starting and ending date and time to at least two hours before the
problem occurred.

Dump Entries to Printer from Licensed Internal Code Log

Type choices, press Enter.

Dump option . . . . . . . . . . . 1 1=Header


2=Header and note entry
3=Header and entire entry
Entry ID:
Starting . . . . . . . . . . . 00000000 00000000-FFFFFFFF
Ending . . . . . . . . . . . . FFFFFFFF 00000000-FFFFFFFF

Entry type:
Major code . . . . . . . . . . 0000 0000-FFFF
Minor code . . . . . . . . . . 0000 0000-FFFF

Starting:
Date . . . . . . . . . . . . . 09/20/11 MM/DD/YY
Time . . . . . . . . . . . . . 00:00:00 HH:MM:SS
Ending:
Date . . . . . . . . . . . . . 09/20/11 MM/DD/YY
Time . . . . . . . . . . . . . 00:00:00 HH:MM:SS

F3=Exit F12=Cancel

Figure 15-11 Dump LIC Log

430 PowerHA SystemMirror for IBM i Cookbook


b. Get the product activity log (PAL) entries for a couple of days before the problem
occurred (for example, a 7-day period) (strsst → login → 1. Start a service tool → 1.
Product activity Log → 1. Analyze log).
When you see the page shown in Figure 15-12, specify the following settings:
• Log type 1 (All logs)
• The From and To dates to include last days (for example, 7) including the time of
the problem occurrence
Press Enter.

Select Subsystem Data

Type choices, press Enter.

Log . . . . . . . . . . 1 1=All logs


2=Processor
3=Magnetic media
4=Local work station
5=Communications
6=Power
7=Cryptography
8=Licensed program
9=Licensed Internal Code

From:
Date . . . . . . . . 09/22/11 MM/DD/YY
Time . . . . . . . . 14:55:28 HH:MM:SS

To:
Date . . . . . . . . 09/29/11 MM/DD/YY
Time . . . . . . . . 14:55:28 HH:MM:SS

F3=Exit F5=Refresh F12=Cancel

Figure 15-12 PAL options

Chapter 15. Best practices 431


On the next page (Figure 15-13) chose these options:
• Report type 3 (Print options)
• Optional entries to include, Statistic Y (for Yes)
Press Enter.

Select Analysis Report Options

Type choices, press Enter.

Report type . . . . . . . . . 3 1=Display analysis, 2=Display summary,


3=Print options
Optional entries to include:
Informational . . . . . . . Y Y=Yes, N=No
Statistic . . . . . . . . . Y Y=Yes, N=No

Reference code selection:


Option . . . . . . . . . . 1 1=Include, 2=Omit
Reference codes
*ALL *ALL...

Device selection:
Option . . . . . . . . . . 1 1=Types, 2=Resource names
Device types or Resource names
*ALL *ALL...

F3=Exit F5=Refresh F9=Sort by ... F12=Cancel

Figure 15-13 PAL print options

432 PowerHA SystemMirror for IBM i Cookbook


On next page (Figure 15-14), choose these options:
• Report type 4 (full report)
• Include hexadecimal data Y (Yes)
Press Enter.

Select Options for Printed Report

Type choices, press Enter.

Report type . . . . . . . 4 1=Print analysis report


2=Print summary report
3=Print partial report
4=Print full report

For report type 4:


Include hexadecimal
data . . . . . . . . Y Y=Yes
N=No

F3=Exit F5=Refresh F12=Cancel

Figure 15-14 PAL print report options

c. The general instructions for how to run the tool can be found in “Instructions for running
the advanced analysis macro” on page 434.

Data to collect for cluster resource group problem


When a problem is related with the CRG, in addition to general, data you might be required to
collect the advanced analysis macros (see “Instructions for running the advanced analysis
macro” on page 434) for these:
 IOSW with no options
 HRIFR with -ipl0 option
 IOHRIDEBUG with option -walklhrichildren -oneliner
 IOHRIDEBUB with option -walklhrichildren -detail
 IOHRITOOLHDWRPT with option -hidden
 IOHRITOOLHDWRPT with option -partofcrg
 IOHRITOOLHDWRPT with option -assocrscofcrg

Chapter 15. Best practices 433


Data to collect when problem is related to administrative domain
In addition to the general data collection, collect the following data:
 If you have a problem with a specific monitored resource run PRTMRE DETAIL(*FULL) for
this resource.
The following steps should be done if you are not using QMGTOOLS for general
data collection.
 Run the following commands and collect the spool files produced by them:
DMPSYSOBJ OBJ(QFPATRES) CONTEXT(QSYS)
DMPSYSOBJ OBJ(QFPATATR) CONTEXT(QSYS)
DMPSYSOBJ OBJ(QFPATMAP) CONTEXT(QSYS)
DMPSYSOBJ OBJ(QFPATCAE) CONTEXT(QSYS)
 The joblog of the QPRFSYNCH job and any other spoolfiles that were generated by
this job.

Data to collect if you are using geographic mirroring


In addition to the general data collection, run the following advanced analysis macros (see
“Instructions for running the advanced analysis macro” on page 434):
 GEOSTAT with option -ALL
 DSMINFO without any options
 ASMINFO with option -ASP nnn -r 50000 (Where nnn is the 3-digit IASP number. For
example, for IASP no. 33 you need to specify -ASP 033.)

Data to collect when using DS8000 TotalStorage


When using DS8000 for your IASP and PowerHA solutions, in addition to general data,
collect DSCLI logs from /QIBM/UserData/HASM/hads/dscli/log.

In case you are not using QMGTOOLS, you need to collect XSM logs that can be found in
/QIBM/UserData/HASM/hads.

Instructions for running the advanced analysis macro


Depending on the problem area within the high-availability environment, IBM Support might
ask you to collect advanced analysis macros. The following steps can be used to display or
print advanced analysis macros:
1. Start service tools with strsst.
2. Log in using the DST user and password (passwords are case sensitive).
3. Select option 1. Start a service tool.
4. Select option 4. Display/Alter/Dump.
5. Select option 2. Dump to printer (or 1. Display/Alter storage if you just want to see the
results on the page without creating spoolfile).
6. Select option 2. Licensed Internal Code (LIC) data.
7. Select option 14. Advanced analysis.

434 PowerHA SystemMirror for IBM i Cookbook


You will get a page similar to Figure 15-15.

Select Advanced Analysis Command

Output device . . . . . . : Printer

Type options, press Enter.


1=Select

Option Command

FLIGHTLOG
ADDRESSINFO
ALTSTACK
BATTERYINFO
CLUSTERINFO
CONDITIONINFO
COUNTERINFO
DISABLEFLASHSYNC
DSTINFO
EXCEPTCHAIN
FINDFRAMES
FINDPTF
More...
F3=Exit F12=Cancel

Figure 15-15 AA panel

Now you can search a macro that you want to run:


1. Specify 1 in the Option column and press Enter.
2. On the next page specify the options that you need (if any), and press Enter.
3. Chose the printing options (if any), and press Enter again.

If you do not find the macro name that you want to run on the list, you can run it by specifying
1 in the Options column in the first row (the empty one) and specifying the macro name in the
Command column.

Chapter 15. Best practices 435


Figure 15-16, Figure 15-17 on page 437, and Figure 15-18 on page 437 show an example of
running the AAExample macro with the -ALL option.

Select Advanced Analysis Command

Output device . . . . . . : Printer

Type options, press Enter.


1=Select

Option Command
1 AAExample
FLIGHTLOG
ADDRESSINFO
ALTSTACK
BATTERYINFO
CLUSTERINFO
CONDITIONINFO
COUNTERINFO
DISABLEFLASHSYNC
DSTINFO
EXCEPTCHAIN
FINDFRAMES
FINDPTF
More...
F3=Exit F12=Cancel

Figure 15-16 AAExample macro example

436 PowerHA SystemMirror for IBM i Cookbook


Specify Advanced Analysis Options

Output device . . . . . . : Printer

Type options, press Enter.

Command . . . . : AAExample

Options . . . . . -all

F3=Exit F4=Prompt F12=Cancel

Figure 15-17 AAExample options

Specify Dump Title

Output device . . . . . . : Printer

Type choices, press Enter.

Dump title . . . . . . . . . . .

Perform seizes . . . . . . . . . 1 1=Yes, 2=No

Partial print page numbers:


From page . . . . . . . . . . . 1 1-2147483647
Through page . . . . . . . . . 9999 1-2147483647

F3=Exit F12=Cancel

Figure 15-18 AA_Example printing options

Chapter 15. Best practices 437


15.12.2 PowerHA GUI
In this section we discuss the PowerHA GUI.

Problems with accessing the GUI for PowerHA


The PowerHA for System i provides a graphical user interface within IBM Systems Director
Navigator for i that is a part of the HTTP Server ADMIN instance. We recommend that you
have the current level of group PTFs for HTTP Server and Java.

If you have problems accessing the IBM Systems Director Navigator for i, make sure that you
have started the HTTP Server ADMIN domain. Use STRTCPSVR SERVER(*HTTP)
HTTPSVR(*ADMIN). Check the ADMIN and ADMIN1-4 jobs in QHTTPSVR subsystem.

Additional information about the problem can be found in the following directories:
 /QIBM/UserData/HASM/logs
 /QIBM/UserData/OS/OSGi/LWISysInst/admin2/lwi/logs

Collect the Java dump of the ADMIN2 job:


Wrkjob admin2
Option 45
Option 32

Get the most recent file named javacore.xxx from the


/QIBM/UserData/OS/OSGi/LWISysInst/admin2/lwi/runtime/core directory.

Problems with operations in PowerHA GUI


When problems occur in PowerHA, you can see detailed messages about the problem in the
web GUI.

Additional information about PowerHA GUI operations can be found in the


/QIBM/UserData/HASM/logs/ directory.

Problem data collection for PowerHA GUI


To collect the diagnostic data to be sent to IBM Support, see “Collecting the data for the
PowerHA GUI with QMGTOOLS” on page 445.

15.12.3 The Must Gather Data Collector


The Must Gather Data Collector is a tool provided by IBM Support for automation of the
diagnostic data collection. The tool functionality is continuously developed and new functions
are added, so we recommend that you have the latest version of the tool1.

Installation of the Must Gather Data Collector


A *SAVF containing the QMGTOOLS library can be found at this website:
https://www-912.ibm.com/i_dir/idoctor.nsf/downloadsMGT.html

The PTFs listed on that website are related to the iDoctor functions. If you are not planing to
use QMPGTOOLS for iDoctor functionality, you do not need to install them. Check the PTFs
for high availability mentioned in 8.1.3, “Power Systems requirements” on page 135.

1
We based our example in this book on the tool version that is current at the time of writing. The tool might have
been updated since then.

438 PowerHA SystemMirror for IBM i Cookbook


You need to download QMGTOOLS for your IBM i version, transfer SAVF to the system, and
restore the QMGTOOLS library2. To restore the QMGTOOLS library, use this command:
RSTLIB SAVLIB(QMGTOOLS) DEV(*SAVF) SAVF(SAVF_LIB/QMGTOOLvrm)

Where SAVF_LIB is the library where you put the downloaded SAVF, and QMGTOOLvrm is
a SAVF name for your IBM i release.

To collect the data with QMGTOOLS it is enough to install the tool on one of the nodes only.

When you restore the QMGTOOLS library you can then use ADDLIBLE QMGTOOLS and
issue GO MG to access the tools main menu and then select option 1 to access the HA data
collection, or you can go straight to this menu with GO HASMNU.

Collecting the general cluster data for all nodes


To collect the diagnostic data for all of the cluster nodes, take the following steps on the node
on which you installed the QMGTOOLS:
1. Go to the HASMNU: GO QMGTOOLS/HASMNU.
2. Select option 1. Collect and retrieve cluster data from multiple nodes.
3. Specify the name of your library, and it will be created (Figure 15-19). Press Enter.

(DMPCLU)

Type choices, press Enter.

Library to store data . . . . . LIB4HADTA Library Name


Dump cluster trace only . . . . N Y, N

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys

Figure 15-19 Options for data collection

2 There is one menu option used by IBM Support (GO MG, opt.1, opt.4) that requires the 5799PTL to be installed.

Chapter 15. Best practices 439


4. Specify the user IDs and passwords for the nodes (Figure 15-20). If you are using the
same user profile and password on all nodes, you can choose option F6 and specify the
same user and password for all nodes (Figure 15-21 on page 441). Press F1 to continue.

Note: To successfully access all nodes, collect the data and get it on single node. User
profiles should be able to FTP successfully to and from every node. In case you cannot
use FTP among your nodes, you need to collect the data on each node separately, as
described in “Collecting the general data on a single node” on page 443.

Nodes UserID Password Confirm Password


----- ------ -------- ----------------

DEMOPROD userid
DEMOHA userid
DEMOFC userid

F1=Continue F3=Exit F6=Options


Figure 15-20 Specify UID and PWD for the nodes

440 PowerHA SystemMirror for IBM i Cookbook


Advanced Options

Use same user/pass for all nodes . . . . . . . . . . . .: Y


User ID : userid Password :
Confirm :

F1=Continue F3=Exit
Figure 15-21 Specify same user and password for all nodes (optional)

Chapter 15. Best practices 441


5. Wait for the collection to complete (Figure 15-22), the press F1.

Nodes Status
----- ------

DEMOPROD Done Retrieving data from remote


DEMOHA Done QDMPCLU/POWERHA/257093 done
DEMOFC Done Retrieving data from remote

FTP done, press F1 to continue...

Figure 15-22 Data collection completed

442 PowerHA SystemMirror for IBM i Cookbook


6. Send the data to IBM. SAVF is placed in the library given in step 4, and the name of SAVF
is in the message at the bottom of the page (Figure 15-23).

HASMNU QHASTOOLS menu

Select one of the following:

1. Collect and retrieve cluster data from multiple nodes


2. Dump cluster data on local node only
3.
4. Cluster Debug Tool (Internal IBM only)
5. Alternative Debug Tool
6.
7.
8. Collect SBG (Solution Based GUI) data
9.
10.
11.
12.

Selection or command
===>

F3=Exit F4=Prompt F9=Retrieve F12=Cancel


F13=Information Assistant F16=System main menu
Data saved into save file CLUDOCS001 in library LIB4HAMNU
Figure 15-23 Information about the SAVF with collected data

Collecting the general data on a single node


In case you cannot collect data using the QMGTOOLS on a single node, you need to install
QMGTOOLS on each node and run the collection on each node. To do this use the following
steps on each node.
1. Add the QMGTOOLS library to your library list:
ADDLIBLE LIB(QMGTOOLS)
2. Go to HASMNU: GO MG and select option 1 (or GO HASMNU).

Chapter 15. Best practices 443


3. Select option 2 (Dump cluster data on local node only) and input the correct parameters
for data collection (a library will be created) (Figure 15-24).

(DMPCLUINF)

Type choices, press Enter.

Cluster Name . . . . . . . . . . > PWRHA_CLU Character value


Local node name . . . . . . . . > DEMOHA Character value
Save into save file? . . . . . . Y Y, N
Library to store data . . . . . LIB4HALOC Character value

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys

Figure 15-24 Parameters for local data collection

444 PowerHA SystemMirror for IBM i Cookbook


4. Wait for the collection to complete. You will get the page shows in Figure 15-25. Press
Enter and send the SAVF that was created to IBM.

Completed...save file CLUDOCS001 created.

Figure 15-25 Panel with SAVF name to be sent to IBM

Collecting the data for the PowerHA GUI with QMGTOOLS


To collect the diagnostic data for the problems that occurred when you were using the
PowerHA GUI, you need to have QMGTOOLS installed on the system on which you are
using the IBM Systems Director Navigator for System i.

Chapter 15. Best practices 445


To collect the data, take these steps:
1. Go to HASMNU: GO QMGTOOLS/HASMNU.
2. Select option 8. Collect SBG (Solution Based GUI) data.
3. Wait for the collection to complete (Figure 15-26) and send the data in the message
to IBM.

HASMNU QHASTOOLS menu

Select one of the following:

1. Collect and retrieve cluster data from local/remote nodes


2. Dump cluster data on local node only
3.
4. Cluster Debug Tool
5. Alternative Debug Tool
6.
7.
8. Collect SBG (Solution Based GUI) data
9.
10.
11.
12.

Selection or command
===>

F3=Exit F4=Prompt F9=Retrieve F12=Cancel


F13=Information Assistant F16=System main menu
Completed, file is in IFS directory /tmp/hasmlogs1003111627.zip
Figure 15-26 Send /tmp/hasmlogs1003111627.zip to support

Reviewing the data collected by QMGTOOLS


As a result of your data collection, there is a library created and physical files put into it. There
are two files for each node. You can review the content of the files with the available system
tool, such as PDM, or with DSPPFM or EDTF.

The file content might vary due to the development of QMPGTOOLS and your configuration.

446 PowerHA SystemMirror for IBM i Cookbook


Using the Alternative Debug Tool option
In the HASMNU, there is an option that you can use to diagnose certain kinds of problems in
your cluster. The types of problems that it can diagnose vary depending on the QMGTOOLS
release, so we recommend that you have the latest version of the tool. To use it, enter GO
HASMNU and select option 5. Alternative Debug Tool. You will be presented a page similar to
the one shown in Figure 15-27. Input the name of the library containing the data that you
collected previously.

Cluster Debug (CLUDBGALT)

Type choices, press Enter.

Library containing dumps . . . . LIB4HAMNU Library Name

Bottom
F3=Exit F4=Prompt F5=Refresh F12=Cancel F13=How to use this display
F24=More keys

Figure 15-27 Alternative debug tool options panel

As a result of this analysis, there will be a spoolfile created in your interactive job. In the file
you can find the problems that the tool has found and additional information about what to
collect to debug the problem further, if necessary.

Development of the Must Gather Data Collector


As it was stated previously, QMGTOOLS is being continuously developed and new functions
are being added. Therefore, we advise that you have the most current version of the tool to
use the new features of this tool.

If you have any suggestions for improvements that might help us improve this tool, you can
contact Benjamin Rabe ([email protected]) or the authors of this publication to have your
suggestions forwarded to the appropriate development team.

Chapter 15. Best practices 447


15.12.4 PowerVM Virtual I/O Server
If you have a problem with storage hosted by Virtual I/O Server, check the following items:
1. The VIOS LPAR operability and configuration. Go to HMC and check the VIOS LPAR
properties for the status and the allocated resources.
2. You can log in to the VIOS LPAR and check the mapping of the devices with lsmap -all.
3. You can use diagmenu to diagnose problems in your VIOS partition.
More information about Virtual I/O Server problem solving and management can be found
in the Power Systems Information Center at:
http://publib.boulder.ibm.com/infocenter/powersys/v3r1m5/index.jsp
4. Collect the snap data. When you sign on to the VIOS LPAR use snap with no parameters.
When it ends there will be file.

15.12.5 DS8000 Copy Services


When using DS8000 for your IASP storage and PowerHA for System i, you can log onto the
TotalStorage device and check the status of the replication using the DSCLI commands:
1. The Metro Mirror status of a volume can be checked with lspprc (Figure 15-28).

dscli> lspprc -fmt default 6000-6002 6100-6102


Date/Time: October 3, 2011 4:44:47 PM CEST IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY031
ID State Reason Type SourceLSS Timeout (secs) Critical Mode First Pass Status
==================================================================================================
6000:6000 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6001:6001 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6002:6002 Full Duplex - Metro Mirror 60 60 Disabled Invalid
6100:6100 Full Duplex - Metro Mirror 61 60 Disabled Invalid
6101:6101 Full Duplex - Metro Mirror 61 60 Disabled Invalid
6102:6102 Full Duplex - Metro Mirror 61 60 Disabled Invalid

Figure 15-28 lspprc for volumes used in Metro Mirror

448 PowerHA SystemMirror for IBM i Cookbook


2. The Global Mirror status for the volumes can be checked with lsgmir (Figure 15-29).

dscli> lsgmir 0
Date/Time: October 6, 2011 4:55:24 PM CEST IBM DSCLI Version: 7.6.10.530 DS:
IBM.1750-13ABGAA
SessionID MasterID ID State %Success CGtime
=======================================================================
0x02 IBM.1750-13ABGAA 02 Running 100 10/06/2011 15:32:12 CEST
dscli> showgmir 0
Date/Time: October 6, 2011 4:55:26 PM CEST IBM DSCLI Version: 7.6.10.530 DS:
IBM.1750-13ABGAA
ID IBM.1750-13ABGAA/02
Master Count 1
Master Session ID 0x02
Copy State Running
Fatal Reason Not Fatal
CG Interval Time (seconds) 0
Coord. Time (milliseconds) 50
Max CG Drain Time (seconds) 30
Current Time 10/06/2011 15:32:15 CEST
CG Time 10/06/2011 15:32:14 CEST
Successful CG Percentage 100
FlashCopy Sequence Number 0x4E8DADDE
Master ID IBM.1750-13ABGAA
Subordinate Count 0
Master/Subordinate Assoc -
Figure 15-29 lsgmir for Global Mirror

3. The FlashCopy status for the copied volumes can be checked with lsflash (Figure 15-30).

dscli> lsflash -fmt default a010-a013


Date/Time: October 3, 2011 4:12:40 PM CEST IBM DSCLI Version: 7.6.10.530 DS: IBM.2107-75AY032
ID SrcLSS SequenceNum Timeout ActiveCopy Recording Persistent Revertible SourceWriteEnabled TargetWriteEnabled BackgroundCopy
====================================================================================================================================
A010:A020 A0 0 60 Disabled Enabled Enabled Disabled Enabled Enabled Enabled
A011:A021 A0 0 60 Disabled Enabled Enabled Disabled Enabled Enabled Enabled
A012:A022 A0 0 60 Disabled Enabled Enabled Disabled Enabled Enabled Enabled
A013:A023 A0 0 60 Disabled Enabled Enabled Disabled Enabled Enabled Enabled

Figure 15-30 lsflash for volumes in FlashCopy

If you need additional help with the command you can use the -help option with any of them.
The list of available commands is shown in DSCLI when you issue help.

For more information about DS8000 see this website:


http://publib.boulder.ibm.com/infocenter/dsichelp/ds8000ic/index.jsp

15.12.6 SVC/V7000 Copy Services


To collect problem data when using the SVC, collect the information for Virtual I/O Server
(15.12.4, “PowerVM Virtual I/O Server” on page 448). Also collect this information:
1. DSPSVCSSN SSN(session_name) OUTPUT(*PRINT).
2. Collect the XSM logs from /QIBM/UserData/HASM/hads.

Chapter 15. Best practices 449


450 PowerHA SystemMirror for IBM i Cookbook
A

Appendix A. IBM i data resilience options


Whether you need continuous availability for your business applications or are looking to
reduce the amount of time that it takes to perform daily backups, IBM i high-availability
technologies provide the infrastructure and tools to help achieve your goals.

Modern IBM i high-availability solutions are built on IBM i cluster resource services, or more
simply clusters. A cluster is a collection or group of multiple systems that work together as a
single system from an application perspective. Clusters provide the underlying infrastructure
that allows resilient resources, such as data, devices, and applications, to be automatically or
manually switched between systems or nodes (partitions on a single frame or on multiple
frames). It provides failure detection and response, so that in the event of an outage, cluster
resource service responds accordingly, keeping your data safe and your business
operational. PowerHA SystemMirror for i offers complete end-to-end integrated solutions for
high availability and disaster recovery, with special focus on the application availability
through planned or unplanned outage events.

This appendix discusses two other options for providing data resiliency:
 Full-system storage-based Copy Services solutions
 Logical replication solutions

We also provide a comparison of the resiliency solutions available for IBM i:


 PowerHA SystemMirror for i
 Full-system storage-based Copy Services
 Logical replication

© Copyright IBM Corp. 2012. All rights reserved. 451


IBM i full-system storage-based Copy Services solutions
The introduction of IBM i boot from SAN further expanded IBM i availability options by exploiting
solutions such as FlashCopy, provided through IBM System Storage Copy Services functions.
With boot from SAN introduced with i5/OS V5R3M0, you no longer needed to use remote load
source mirroring to mirror your internal load source to a SAN-attached load source. Instead, the
load source can be placed directly inside a SAN-attached storage subsystem and with IBM i 6.1
provide multi-path attachment to the external load source for redundancy.

Boot from SAN makes it easier to bring up a system environment that has been copied using
Copy Services functions such as FlashCopy or remote Copy Services. During the restart of a
cloned environment, you no longer have to perform the Recover Remote Load Source Disk
Unit through Dedicated Service Tools (DSTs), thus reducing the time and overall steps
required to bring up a point-in-time system image after FlashCopy or remote Copy Services
functions have been completed.

Cloning IBM i
Cloning has been a concept for the IBM i platform since the introduction of boot from SAN
with i5/OS V5R3M5. Previously, to create a new system image, you had to perform a full
installation of the SLIC and IBM i. When cloning IBM i, you create an exact copy of the
existing IBM i system or partition. The copy can be attached to another System i/Power
Systems models, a separate LPAR, or, if the production system is powered off, the existing
partition or system. After the copy is created, you can use it for offline backup, system testing,
or migration.

Boot from SAN enables you to take advantage of some of the advanced features that are
available with IBM system storage. One of these functions is FlashCopy. It allows you to
perform a point-in-time instantaneous copy of the data held on a LUN or group of LUNs.
Therefore, when you have a system that only has SAN LUNs with no internal drives, you can
create a clone of your system.

Important: When we refer to a clone, we are referring to a copy of a system that only uses
SAN LUNs. Therefore, boot from SAN is a prerequisite for this.

452 PowerHA SystemMirror for IBM i Cookbook


Full system FlashCopy
FlashCopy not only allows you to take a system image for cloning but is also an ideal solution
for increasing the availability of your IBM i production system by reducing the time for your
system backups (Figure A-1).

Production IBM i Backup IBM i

SLIC, OS, LPP &


User Data
Snapshots for backups,
Flash OS and application upgrade
Copy
role-backs, BI and queries.

DS8000
Figure A-1 Full system FlashCopy

To obtain a full system backup of IBM i with FlashCopy, either a system shutdown or, as of
IBM i 6.1, a quiesce function is supplied that flushes modified data from memory to disk.
FlashCopy only copies the data on the disk. The IBM i quiesce for Copy Services function
(CHGASPACT) introduced with 6.1 allows you to suspend all database I/O activity for
*SYSBAS and IASP devices before taking a FlashCopy system image, eliminating the
requirement to power down your system.

Full system replication by Metro Mirror or Global Mirror


Metro Mirror offers synchronous replication between two DS models or between a DS and
ESS model 800. In Figure 3-9, two IBM i servers are separated by distance to achieve a
disaster recovery solution at the second site. This is a fairly simple arrangement to implement
and manage. Synchronous replication is desirable because it ensures the integrity of the I/O
traffic between the two storage complexes and provides you with a recovery point objective
(RPO) of zero (that is, no transaction gets lost). The data on the second DS system is not
available to the second IBM i system while Metro Mirror replication is active (that is, it must be
powered off).

The main consideration with this solution is distance. The solution is limited by the distance
between the two sites. Synchronous replication needs sufficient bandwidth to prevent latency
in the I/O between the two sites. I/O latency can cause application performance problems.

Appendix A. IBM i data resilience options 453


Testing is necessary to ensure that this solution is viable depending, on a particular
application’s design and business throughput.

With Global Mirror all the data on the production system is asynchronously transmitted to the
remote DS models. Asynchronous replication via Global Copy alone does not guarantee the
order of the writes, and the remote production copy will lose consistency quickly. To
guarantee data consistency, Global Mirror creates consistency groups at regular intervals, by
default, as fast as the environment and the available bandwidth allows. FlashCopy is used at
the remote site to save these consistency groups to ensure that a consistent set of data is
available at the remote site, which is only a few seconds behind the production site. That is,
when using Global Mirror a RPO of only a few seconds can be achieved normally without any
performance impact to the production site.

Figure A-2 shows an overview of a full system remote copy solution by using IBM
system storage.

Production IBM i Backup IBM i

Mirror copy for planned and


unplanned outage.
Copy all data (ex., SLIC, OS,
LPP, user data) including
temporary storage data.

SAN

Synchronous copy Metro Mirror

SLIC, OS, LPP &


User Data

Asynchronous copy Global Mirror

DS8000 DS8000
Figure A-2 Full system remote Copy Services

This is an attractive solution because of the extreme distances that can be achieved with
Global Mirror. However, it requires a proper sizing of the replication link bandwidth to ensure
that the RPO targets can be achieved. Testing should be performed to ensure that the
resulting image is usable.

When you recover in the event of a failure, the IPL of your recovery system will always be an
abnormal IPL of IBM i on the remote site.

Note: Using IBM i journaling along with Metro Mirror or Global Mirror replication solutions
is highly recommended to ensure transaction consistency and faster recovery.

454 PowerHA SystemMirror for IBM i Cookbook


Logical replication solutions
Logical replication is currently the most widely deployed multisystem data resiliency topology
for high availability (HA) in the IBM i world. It is deployed by using IBM iCluster or a High
Availability Business Partner (HABP) solution package. Replication is executed (via software
methods) on objects. Changes to the objects (for example, file, member, data area, or
program) are replicated to a backup copy. The replication process is near real time. Typically,
if the object, such as a file, is journaled, replication is handled at a record level. For such
objects as user spaces that are not journaled, replication is handled at the object level. In this
case, the entire object is replicated after each set of changes to the object is complete.

Production IBM i Backup IBM i


Replicate changed data via
IBM i remote journal.
DB and Non-DB objects can
replicate but it depends on
the software.

Journal IBM i Remote journal (DB and other) Journal

Logical
Apply
Replication
Management and Control Process
Software

Figure A-3 Logical replication

Most logical replication solutions allow for additional features beyond object replication. For
example, you can achieve additional auditing capabilities, observe the replication status in
real time, automatically add newly created objects to those being replicated, and replicate a
subset of objects in a given library or directory.

To build an efficient and reliable multi-system HA solution using logical replication,


synchronous remote journaling as a transport mechanism is preferable. With remote
journaling, IBM i continuously moves the newly arriving database data in the journal receiver
to the backup server journal receiver. At this point, a software solution is employed to “replay”
these journal updates, placing them into the object or replacing the object on the backup
server. After this environment is established, there are two separate yet identical objects, one
on the primary server and one on the backup server.

With this solution in place, you can activate your production environment on the backup
server via a role-swap operation.

One characteristic of this solution category is that the backup database file is “live”. That is, it
can be accessed in real time for backup operations or for other read-only application types,
such as building reports.

Appendix A. IBM i data resilience options 455


Another benefit of this type of solution is that different releases of IBM i can be utilized on the
primary and backup nodes. This means that the solution can be used to assist in migrations
to new operating system levels. As an example, you can be replicating from IBM i 6.1 on the
production system to IBM i 7.1 on the backup system, and when you want to migrate the
production system to IBM i 7.1 you can perform a roleswap to the backup, perform your OS
upgrade on production, and when the migration has been validated and tested, roleswap your
users back to production.

The challenge with this solution category is the complexity that can be involved with setting up
and maintaining the environment. One of the fundamental challenges lies in not strictly
policing undisciplined modification of the live copies of objects residing on the backup server.
Failure to properly enforce such a discipline can lead to instances in which users and
programmers make changes against the live copy so that it no longer matches the production
copy. Should this happen, the primary and backup versions of your files are no longer
identical. There are new tools provided by the software replication providers that perform
periodic data validation to help detect this situation.

Another challenge associated with this approach is that objects that are not journaled must go
through a checkpoint, be saved, and then be sent separately to the backup server. Therefore,
the granularity of the real-time nature of the process might be limited to the granularity of the
largest object being replicated for a given operation.

For example, a program updates a record residing within a journaled file. As part of the same
operation, it also updates an object, such as a user space, that is not journaled. The backup
copy becomes completely consistent when the user space is entirely replicated to the backup
system. Practically speaking, if the primary system fails, and the user space object is not yet
fully replicated, a manual recovery process is required to reconcile the state of the
non-journaled user space to match the last valid operation whose data was completely
replicated. This is one of the reasons why the application state and the state of the replicated
data at the target box are inherently asynchronous.

Another possible challenge associated with this approach lies in the latency of the replication
process. This refers to the amount of lag time between the time at which changes are made
on the source system and the time at which those changes become available on the backup
system. Synchronous remote journal can mitigate this for the database. Regardless of the
transmission mechanism used, you must adequately project your transmission volume and
size your communication lines and speeds properly to help ensure that your environment can
manage replication volumes when they reach their peak. In a high-volume environment,
replay backlog and latency might be an issue on the target side even if your transmission
facilities are properly sized. Another issue can arise when the apply process on the target
system cannot keep up with the incoming data. The target system cannot be used until this
lag data is fully applied.

456 PowerHA SystemMirror for IBM i Cookbook


Comparison characteristics
In this section, we selected major characteristics to consider. However, you might have other
characteristics that are equally or more important to your environment. To compare the
various availability techniques that use some form of data resiliency, we use the
characteristics in the technology comparison shown in Table A-1 and Table A-2.

Table A-1 Comparison with data resiliency options (technology)


PowerHA Full-system Logical replication
SystemMIrror for i storage-based Copy
Services

Base technology IBM i Storage Storage Replication software


management

Cluster resource Yes No No


management

Multi-site cluster Yes No No


resource management

Multi-site data Yes Yes Yes


replication service

Synchronous to Yes Yes No


application state

Environment resiliency Cluster Admin Domain No Depends on software

Integrated heartbeat Yes No No

Integrated flash copy Yes Yes N/A

Automated/user Yes No Depends on software


control fail over and applications

Internal disk support Yes No Yes

Switched disk-based Yes No No


resiliency

Host-based replication Geographic Mirroring No Remote journal

Storage-based Metro Mirror Metro Mirror No


replication Global Mirror Global Mirror

Table A-2 Comparison with data resiliency options (IT operating environment)
PowerHA Full-system Logical replication
SystemMIrror for i storage-based Copy
Services

IT operation skill base IBM i/Storage Storage IBM i/Replication


software

Technical skill support IBM IBM IBM/HABP


base

Data synchronization Disk mirroring Disk mirroring Journal replay

Required bandwidth Data change in IASP Data change in full Journal changes
system

Appendix A. IBM i data resilience options 457


PowerHA Full-system Logical replication
SystemMIrror for i storage-based Copy
Services

Save window FlashCopy FlashCopy Save while active

Primary/secondary Yes Yes No


storage synchronously
mirrored

Use of journal Yes, best practice Yes, best practice Source of replicated
data

Application Not required Not required Required, after


environment change changing or adding
management for applications, setting up
replication services replication

Object type support IASP data All Most

IPL is required when No Yes No


switchover/failover

Recovery point Last data written to Last data written to Depends on


objective IASP Storage transaction boundary

Recovery time IASP vary on time Abnormal IPL time Time of applying lag
objective plus switchover
overhead plus sync
check

458 PowerHA SystemMirror for IBM i Cookbook


Related publications

The publications listed in this section are considered particularly suitable for a more detailed
discussion of the topics covered in this book.

IBM Redbooks
The following IBM Redbooks publications provide additional information about the topic in this
document. Note that some publications referenced in this list might be available in softcopy
only.
 Striving for Optimal Journal Performance on DB2 Universal Database for iSeries,
SG24-6286
 Journaling - User ASPs Versus the System ASP, TIPS0602
 Journaling at object creation on DB2 for iSeries, TIPS0604
 Journaling: Why Is My Logical File Journaled?, TIPS0677
 Journaling - How Can It Contribute to Disk Usage Skew?, TIPS0603
 Journaling · Journal Receiver Diet Tip 1: Eliminating Open and Close Journal Entries,
TIPS0607
 Journaling - *RMVINTENT: The preferred fork in the road for heavy journal traffic,
TIPS0605
 Journaling · Journal Receiver Diet tip 2: Consider using skinny headers, TIPS0654

You can search for, view, download or order these documents and other Redbooks,
Redpapers, Web Docs, draft and additional materials, at the following website:
ibm.com/redbooks

Online resources
These websites are also relevant as further information sources:

These websites are also relevant as further information sources:


 IBM i Information Center
http://publib.boulder.ibm.com/infocenter/systems/scope/i5os/index.jsp
 IBM PowerHA
http://www-03.ibm.com/systems/power/software/availability/i5os.html
 IBM iCluster for i
http://www-03.ibm.com/systems/power/software/availability/i5os.html#icluster

© Copyright IBM Corp. 2012. All rights reserved. 459


Help from IBM
IBM Support and downloads
ibm.com/support

IBM Global Services


ibm.com/services

460 PowerHA SystemMirror for IBM i Cookbook


Index
Change Network Attributes (CHGNETA) command 30
Symbols CHGASPSSN 62
*QMQRY 32 CHGCLUNODE 43
*SYSBAS 16 CHGNETA command 30
CIM 43
Numerics CIM event 43
5799-HAS PRPQ 18 CKD (count key data) 85
class object 26
Client Access/400 30
A Cluster 36
Add Admin Domain Node Entry 55 cluster administrative domain 55
ADDCADMRE 56 Cluster Node 38
ADDCADNODE 55 Cluster Resource Group 41–42
ADDCLUMON 44 Cluster Resource Group object types 41
Administrative Domain 27 commitment control 49
Administrative domain 55 Common Information Model (CIM) server 43
Advanced Function Printing Data Stream (AFPDS) 28 Configuration message queue 28
AFPDS (Advanced Function Printing Data Stream) 28 connections 18
Alert Filters parameter 30 consistency groups 122, 124
alert table 26 Consistent stopped 121
Allow user domain objects in libraries 28 Consistent synchronized 121
ALRFTR (Alert Filters) 30 Controlling subsystem 28
Application CRG 41 copy on write 129
application migration 33 copy rate 126
application requester (AR) 16 Copy Services 118
Application Requester Driver (ARD) program 16 copying 127
application server 16 CPFBB22 43
AR (application requester) 16 CPFBB4F 43
ARD (Application Requester Driver) 16 create an IASP 18
Array site 84 CRG 42
ASP CRTDEVASP 18
group component 16
ASP Copy Descriptions 57
ASP Sessions 60 D
asynchonous delivery mode 68 Data CRG 41
asynchronous delivery 76 Data port services 69
Asynchronous geographic mirroring 68 DataLink 26
asynchronous mirroring mode 74 DB2 Web Query 33
Asynchronous transmission delivery 76 DDM 32
Attention program 28 DDM (distributed data management) 30
auxiliary storage pools 14 DDMACC (Distributed Data Management Access) 30
Dedicated Service Tools (DST) 19
default sort sequence algorithm 29
B Device CRG 42
background copy 126 Device Domain 39
Backup node 39 Devices adapters 82
backup system 50 Disks enclosures 82
Basic disk pools 14 distributed data management (DDM) 30
basic user ASP 15, 30 Distributed Data Management Access (DDMACC) 30
Battery BackUp (BBU) 83 Double-byte code font 28
DRDA 30, 32
DS8000 storage 82
C DS8700 82
Capacity BackUp (CBU) 133 DS8800 82
CFGDEVASP 25, 134 DSPASPSSN 70
CFGGEOMIR 134

© Copyright IBM Corp. 2012. All rights reserved. 461


E L
Extent pools 86 library name space 16
library-capable IASPs 15
link tolerance 123
F logical subsystem 93
Failback 104 LOOPBACK 32
Failover 103 LPP 134
FB (fixed block) 85 LUN level switching 54, 105
Fibre Channel adapter 92
FlashCopy 54, 102, 111
establish 108 M
reading from the source 109 MDisks 116
reading from the target 109 Metro Mirror 52
terminating the FlashCopy relationship 110 Metro Mirror pair 94
writing to the source 109 Micro-Partitioning™ 136
writing to the target 110 mirror copy 49, 69
FlashCopy pair 54, 108 Mirroring Mode 70
FlashCopy relationship states 127 Mirroring Mode. 50
FlashCopy SE 54, 111 monitored resource 56
flexible service processor (FSP) 43 monitored resource entries (MREs) 56
Full synchronization 71 MOUNT 32
Full volume copy 110 MREs 56

G N
Geographic mirroring 50 N_Port ID Virtualization (NPIV) 136
Global Copy 102 Name Space 15
Global Mirror 53, 99 name space 16
Global Mirror session 102 network attributes 30
Nocopy option 110

H
Hardware Management Console 83 O
Hardware Management Console (HMC) 77, 136 object creation 18
Heartbeat monitoring 37 ODBC 31
host based replication 51 OptiConnect connections 30
host connection 92 Outqueues 33
HSL loop 48

P
I Partial synchronization 71
IBM PowerVM 136 Password validation program 29
IBM Storwize V7000 54, 114 Peer CRG 42
idle or copied 127 POWER Hypervisor™ (PHYP) 43
Idling 121 POWER6 48
Inactive job message queue 29 prepared 128
Inconsistent copying 121 preparing 127
Inconsistent stopped 121 prestarted job 31
incremental FlashCopy 126 Primary disk pool 14
independent auxiliary storage pools 14 Primary node 39
indirection layer 125 Problem log filter 29
Integrated Virtualization Manager (IVM) 136 production copy 49, 69
IO enclosures 82 Program Request Pricing Quotation (PRPQ) 134
IP address 16

Q
J QALWUSRDMN 28
JDBC 31 QATNPGM 28
job descriptions 27 QBOOKPATH 28
Job queues 27 QCFGMSGQ 28
journal receivers 27 QCTLSBSD 28
Journals 27 QDFTJOBD 33

462 PowerHA SystemMirror for IBM i Cookbook


QGPL 33 suspended 127
QINACTITV 29 Switchable IASPs 78
QINACTMSGQ 29 Switched disk 47
QPFRADJ 146 Switched disks 77
QPRBFTR 29 switching RDBs 18
QPWDVLDPGM 29 SWSC 18
QSRTSEQ 29 SWSC (system-wide statement cache) 18
QSTRUPPGM 29 Synchronization priority 72
QSYS 28 Synchronous geographic mirroring 74
QSYS/QSYSOPR 30 synchronous mirroring mode 74
QSYS2 18 SYSBAS 136
QSYSLIBL 33 system ASP 15
QSYSOPR 29 System disk pool 14
QSYSSBSD 28 System part of the library list 29
QTEMP 15 Systems Network Architecture (SNA) 28
QUPSMSGQ 30 system-wide statement cache (SWSC) 18
QUSRLIBL 30, 33
QUSRSYS 33
T
target system 49
R Thin-provisioned FlashCopy 128
Rank 85 Tracking space 73
RDB directory 16 Transmission Delivery 50, 70
Recovery Domain 42
Redbooks website 459
Contact us xvi U
Relational Database (RDB) directory 16 uninterruptible power supply (UPS) 30
Reliable message function 37 UPS (uninterruptible power supply) 30
Remove Admin Domain Node Entry 55 user domain user 28
Remove Cluster Admin Domain Node 55 User part of the library list 30
Replicate node 39
RMVCADMRE 56 V
RMVCADNODE 55 V7000 53
VDisks 116
S VIOS 43
SAN Volume Controller 54, 114 Virtual I/O Server 136
SAN Volume Controller (SVC) 114 Virtual I/O Server (VIOS) 43
Secondary disk pool 14 Volumes 87
Service Activity Manager 29 volumes group 91
service tools user ID 19
SETASPGRP 28, 32 W
SI44148 134 Work with Cluster 55
SNA (Systems Network Architecture) 28 WRKASPCPYD 64
Sort sequence 29 WRKCLU 55
source system 49 WRKSHRPOOL 145
Space Efficient 54
space-efficient FlashCopy 111
Spare disks 84
spool files 33
SQL catalog 16
SQL CONNECT 18
SQL interface 16
Startup program 29
stopped 127
stopping 127
Storage based replication 52
Storwize V7000 114
STRQQRY 33
STRTCPSVR 44
Suspend timeout 72

Index 463
464 PowerHA SystemMirror for IBM i Cookbook
PowerHA SystemMirror for IBM i
Cookbook
PowerHA SystemMirror for IBM i Cookbook
PowerHA SystemMirror for IBM i Cookbook
PowerHA SystemMirror for IBM i Cookbook
PowerHA SystemMirror for IBM i
Cookbook
PowerHA SystemMirror for IBM i
Cookbook
Back cover ®

PowerHA SystemMirror
for IBM i Cookbook
®

Take advantage of IBM PowerHA SystemMirror for i is the IBM high-availability disk-based
clustering solution for the IBM i 7.1 operating system. When combined INTERNATIONAL
PowerHA to configure
with IBM i clustering technology, PowerHA for i delivers a complete TECHNICAL
and manage high
high-availability and disaster-recovery solution for your business SUPPORT
availability applications running in the IBM System i environment. PowerHA for i ORGANIZATION
enables you to support high-availability capabilities with either native
Find guidance on disk storage or IBM DS8000 or DS6000 storage servers or IBM
planning and Storwize V7000 and SAN Volume Controllers.
implementing The latest release of IBM PowerHA SystemMirror for i delivers a
PowerHA brand-new web-based PowerHA graphical user interface that BUILDING TECHNICAL
effectively combines the solution-based and task-based activities for INFORMATION BASED ON
your HA environment, all in a single user interface. PRACTICAL EXPERIENCE
Benefit from the latest
PowerHA solution This IBM Redbooks® publication provides a broad understanding of IBM Redbooks are developed
enhancements PowerHA for i. This book is intended for all IBM i professionals who are by the IBM International
planning on implementing a PowerHA solution on IBM i. Technical Support
Organization. Experts from
IBM, Customers and Partners
from around the world create
timely technical information
based on realistic scenarios.
Specific recommendations
are provided to help you
implement IT solutions more
effectively in your
environment.

For more information:


ibm.com/redbooks

SG24-7994-00 ISBN 0738436364

You might also like