ACI Guide

Front cover
A Guide to Using
ACI Worldwide’s
BASE24-es on z/OS
Set up and use BASE24-es on z/OS
High availability scenarios
Configure for highest

availability
Alex Louwe Gordon Fleming

Kooijmans Claus Koefoed
Edward Addison Ron Schmidt
Alexei Alaev Bob Spanke
Dan Archer Mingming Tao
Dore Teichman
ibm.com/redbooks
International Technical Support Organization
A Guide to Using ACI Worldwide’s BASE24-es on

z/OS
August 2006
SG24-7268-00
Note: Before using this information and the product it supports, read the information in
“Notices” on page vii.
First Edition (August 2006)
This edition applies to Release 1, Version 06.2 of ACI Worldwide’s BASE24-es product on z/OS.
© Copyright International Business Machines Corporation 2006. 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 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
The team that wrote this redbook. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x
Become a published author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii
Comments welcome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii
Chapter 1. Running a payments system . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Industry growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.1 The requirement for reliability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.2 System availability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.3 Basic banking transaction categories . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 How ATM and POS processing is accomplished . . . . . . . . . . . . . . . . . . . . 4
1.2.1 Transaction acquisition and authorization . . . . . . . . . . . . . . . . . . . . . . 7
1.3 How System z addresses ATM/EFT/POS processing requirements . . . . . 7
1.3.1 Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3.2 Parallel Sysplex clustering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3.3 Manageability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.3.4 Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3.5 Expandability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3.6 Dynamic workload balancing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Chapter 2. Introducing BASE24-es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.1 BASE24-es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.1.1 Consumer transaction support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.1.2 Transaction switching and routing. . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.1.3 Flexible authorization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.1.4 Integrated consumer data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.1.5 Traditional and emerging delivery channels . . . . . . . . . . . . . . . . . . . 16
2.1.6 Reliable security infrastructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.1.7 Intuitive graphical user interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1.8 Scriptable extracts and reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1.9 National switch and international card scheme interfaces . . . . . . . . 17
2.2 Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3 Operability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.4 Scripting component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.5 The user interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.6 Rich functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
© Copyright IBM Corp. 2006. All rights reserved. iii

2.7 Platform independence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.8 Flexible architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.9 The ACI Payments Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Chapter 3. System z and z/OS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

3.1 How z/OS addresses non-functional requirements . . . . . . . . . . . . . . . . . . 24
3.1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.1.2 Traditional z/OS strengths. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.1.3 Centralized computing model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Chapter 4. BASE24-es technical architecture on z/OS . . . . . . . . . . . . . . . 41

4.1 BASE24-es logical architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.1.1 Memory-based message queuing . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.1.2 Indexed structured file system or relational database . . . . . . . . . . . . 43
4.1.3 Transaction manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.1.4 Data communications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.1.5 Other requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.2 BASE24-es physical architecture on System z . . . . . . . . . . . . . . . . . . . . . 44
4.2.1 Long-running tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.2.2 Non-recoverable TDQs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.2.3 Context-free servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.2.4 Message routing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.2.5 Single-region deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.2.6 Two-tier deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.2.7 Three-tier deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.2.8 Workload management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.2.9 TOR architecture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.2.10 AOR architecture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.2.11 Financial transaction flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.2.12 BASE24-es logical architecture on z/OS. . . . . . . . . . . . . . . . . . . . . 90
4.2.13 User interface region architecture . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.2.14 Scalability considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Chapter 5. Designing the system layout . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

5.1 Pain versus gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.1.1 Planned versus unplanned outages . . . . . . . . . . . . . . . . . . . . . . . . 100
5.1.2 Single site versus dual site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.2 High availability considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.2.1 Terminal Owning Regions (TORs) . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.2.2 Application Owning Regions (AORs) . . . . . . . . . . . . . . . . . . . . . . . 105
5.2.3 File access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
5.2.4 External system connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
5.2.5 Back-end authorization system connectivity . . . . . . . . . . . . . . . . . . 114
5.3 BASE24-es configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
iv A Guide to Using ACI Worldwide’s BASE24-es on z/OS

5.3.1 Single LPAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
5.3.2 Multiple LPARs/Single CPCs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
5.3.3 Multiple LPARs/Multiple CPCs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
5.3.4 Multiple LPARs/Multiple CPCs/Dual site . . . . . . . . . . . . . . . . . . . . . 120
Chapter 6. Installing BASE24-es on z/OS . . . . . . . . . . . . . . . . . . . . . . . . . 123

6.1 The ITSO environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
6.1.1 Hardware we used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
6.1.2 Software we used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
6.2 BASE24-es requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
6.2.1 Hardware requirements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
6.2.2 Software requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
6.2.3 Desktop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
6.3 Overview of pre-installation steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
6.3.1 General considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
6.3.2 Installing and implementing the workstation requirements . . . . . . . 128
6.3.3 Additional requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
6.3.4 New time library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
6.4 Installation considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
6.4.1 Expanding the BASE24-es installation to a CICSplex . . . . . . . . . . 132
6.5 Setting up integration with the back-end system . . . . . . . . . . . . . . . . . . . 133
Chapter 7. Setting up and running the initial workload . . . . . . . . . . . . . . 137

7.1 Setting up the workload. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
7.1.1 Configuring the BASE24-es infrastructure . . . . . . . . . . . . . . . . . . . 138
7.1.2 Business configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
7.2 Our ITSO system layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
7.3 Setting up monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
7.3.1 Routing region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
7.3.2 Target region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
7.4 Installation verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
7.4.1 Simulate single transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
7.4.2 Increasing the workload . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Chapter 8. Tuning BASE24-es on z/OS . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

8.1 BASE24-es File Partitioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
8.2 Basic considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
8.3 Single files as potential bottlenecks . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
8.4 CICS logging impacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
8.5 RMF Monitor III indicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
Chapter 9. Achieving high availability on z/OS . . . . . . . . . . . . . . . . . . . . 159

9.1 Test scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
9.2 CICS and CP/SM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Contents v
9.2.1 CMAS failure after workload has been activated . . . . . . . . . . . . . . 160
9.2.2 TCP/IP client - TOR failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
9.2.3 TCP/IP server - TOR failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
9.2.4 AOR failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
9.2.5 TCP/IP server - TOR maintenance . . . . . . . . . . . . . . . . . . . . . . . . . 167
9.2.6 AOR maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
9.2.7 Dynamically adding a TOR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
9.2.8 Dynamically adding an AOR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
9.3 Data environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
9.3.1 SMSVSAM failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
9.3.2 Lock structure full - IGWLOCK00 . . . . . . . . . . . . . . . . . . . . . . . . . . 176
9.3.3 Read-only data table at initial time . . . . . . . . . . . . . . . . . . . . . . . . . 184
9.3.4 Read-only flat file at initial time . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
9.4 BASE24-es application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
9.4.1 Journal file full . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
9.4.2 IP client failure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
9.4.3 IP Server failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
9.4.4 Integrated Server failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
9.4.5 Planned application upgrade. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
9.4.6 Remote program link to back-end authorization failure. . . . . . . . . . 202
9.4.7 Usage file full. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
9.5 Hardware. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
9.5.1 Central Processor (CP) failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
9.5.2 CPC failure/LPAR failure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
9.5.3 Coupling Facility (CF) failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
9.6 z/OS and relevant subsystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
9.6.1 z/OS maintenance - best practices . . . . . . . . . . . . . . . . . . . . . . . . . 215
9.6.2 CF maintenance - best practices . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Chapter 10. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219

10.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
10.2 The payment environment requirements. . . . . . . . . . . . . . . . . . . . . . . . 220
10.2.1 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
Related publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227

IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Other publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Online resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
How to get IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Help from IBM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
vi A Guide to Using ACI Worldwide’s BASE24-es on z/OS

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 Web sites are provided for convenience only and do not in any
manner serve as an endorsement of those Web sites. The materials at those Web sites are not part of the
materials for this IBM product and use of those Web sites 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 illustrates 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. 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 IBM's application
programming interfaces.
© Copyright IBM Corp. 2006. All rights reserved. vii

Trademarks
The following terms are trademarks of the International Business Machines Corporation in the United States,
other countries, or both:
Eserver Parallel Sysplex™ RMF™

Eserver GDPS® Sysplex Timer®
eServer™ HiperSockets™ System z™
AIX® IBM® System z9™
Chipkill™ IMS™ Tivoli®
CICS® Language Environment® Tivoli Enterprise™
CICS/ESA® Maestro™ TotalStorage®
CICSPlex® MVS™ VTAM®
DB2® OS/390® WebSphere®
DFSORT™ Parallel Sysplex® z9™
DS6000™ pSeries® z/OS®
DS8000™ Redbooks™ zSeries®
FlashCopy® Redbooks (logo) z/VM®
Geographically Dispersed RACF® z/VSE™®
The following terms are trademarks of other companies:
ACI, ACI Worldwide, and BASE24-es are trademarks or registered trademarks of ACI Worldwide Inc. or its
affiliates in the United States, other countries or both.
Java™, JSP™, and all Java-based trademarks are trademarks of Sun™ Microsystems™, Inc. in the United
States, other countries, or both.
Windows®, and the Windows logo are trademarks of Microsoft® Corporation in the United States, other
countries, or both.
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.
UNIX® is a registered trademark of The Open Group 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.
viii A Guide to Using ACI Worldwide’s BASE24-es on z/OS

Preface
In this IBM® Redbook we explain how to use the ACI BASE24-es product on
z/OS®. BASE24-es is a payment engine utilized by the financial payments
industry. The combination of BASE24-es and System z™ is a competitive and
attractive end-to-end retail payments solution for the finance sector.
Failure in a financial payments environment is a high-visibility customer service

issue, and outages at any level have debilitating effects on customer loyalty. The
entire payments cycle must be conducted in near real-time. In such an
environment, high availability is a mission-critical requirement. In this guide, we
demonstrate how you can achieve a high availability configuration for BASE24-es
on z/OS.
We begin by outlining the requirements of a payments system, and then

introduce the structure and functionality offered by the BASE24-es product. Next
we describe the strengths and abilities of System z and z/OS, and explain the
technical and physical architecture of BASE24-es on z/OS.
We guide you in designing a system layout and in installing, tailoring, and

configuring your workload on z/OS. Finally, we detail the numerous failure
scenarios that we tested in order to verify the robustness of the solution. These
test scenarios were carefully selected in areas such as CICS/CICSplex, data
environment, BASE24-es application, and hardware. Communication was
handled by ATM or POS device simulators and a Visa network simulator. (Testing
for high availability of communication, however, was beyond the scope of this
redbook.)
This redbook is focused on business verification and can be used by system

administrators who work with payment transactions systems. It will help you to
achieve the highest possible availability on z/OS by running the supplied test
scenarios.
BASE24-es is an evolving product. The information provided in this redbook is

specific to BASE24-es release 06.2, and is subject to change in subsequent
releases of the product.
© Copyright IBM Corp. 2006. All rights reserved. ix

The team that wrote this redbook
This redbook was produced by a team of specialists from around the world
working at the International Technical Support Organization, Poughkeepsie
Center.
Alex Louwe Kooijmans is a Project Leader with the International Technical

Support Organization (ITSO) in Poughkeepsie, NY. He specializes in
WebSphere®, Java™ and SOA on System z with a focus on integration,
security, high availability and application development. Previously, he worked as
a Client IT Architect in the Financial Services sector with IBM in The
Netherlands, advising financial services companies on IT issues such as
software and hardware strategy and on demand. Alex has also worked at the
Technical Marketing Competence Center for zSeries® and Linux in Boeblingen,
Germany, providing support to customers implementing Java and WebSphere
on zSeries. From 1997 to 2002, he worked on a previous assignment with the
ITSO, managing various IBM Redbook projects and delivering workshops
around the world.
Edward Addison is a Software Engineer working in Raleigh, NC as the

Technical Lead for CICS® Level 2 support. Prior to working at CICS Level 2, he
was a member of the VSAM Level 2 support group in San Jose, CA. Edward has
been with IBM for 18 years, supporting customer VSAM and CICS. He holds a
BS degree in Information Systems from the University of Phoenix.
Alexei Alaev is a Senior Engineer with ACI Worldwide, located in the USA.
Alexei has 28 years of experience in the IT industry. He holds degrees in
Computer Science from Moscow State University in Russia. His areas of
expertise include IBM mainframe application and system programming, including
the sysplex environment. For the past 8 years, Alexei has been working on the
development and implementation of the multi-platform BASE24-es application.
Dan Archer is a Senior Product Manager with ACI Worldwide Inc. located in
Omaha, NE. He has 11 years of online transaction processing experience with
ACI, and has provided technical support of BASE24-es since its inception.
Gordon Fleming is a Master Engineer with ACI Worldwide Inc., located in the
USA. He has 19 years of experience in online transaction processing, primarily in
BASE24 and BASE24-es development. He holds degrees from the University of
Chicago and the University of Nebraska at Omaha. His areas of expertise
include multi-platform infrastructure design and implementation.
Claus Koefoed is a Program Manager/zSeries from CompeteCenter in IBM

Denmark. He has 30 years of experience in the I/T industry with IBM. His area of
professional focus is on IBM mainframe competition, EFT solutions. Claus is
x A Guide to Using ACI Worldwide’s BASE24-es on z/OS

editor of a newsletter/Sinet on mainframe subjects for the IBM community. He
holds a Msc. Econ. from the University of Copenhagen.
Ron Schmidt is a Systems Engineer with ACI Worldwide, Inc., located in the
US. He has 15 years of experience in online transaction processing, primarily in
TRANS24-eft development/deployment and BASE24-es performance. He holds
a degree in Computer Science and Business Administration from Kearney State
College. His areas of expertise include development/deployment of online
transaction processing systems and the performance analysis of zSeries
applications.
Bob Spanke is a Master Engineer in ACI Worldwide's Technical Steering group.

He is located in Omaha, NE. Bob has 19 years of online transaction process
experience with ACI. His primary area of focus is BASE24-es application
architecture. Bob holds Bachelor of Science degrees in Computer Science and
Information Systems from Kansas State University.
Mingming Tao is a z/OS Systems Programming leader with the China

Merchants Bank Data Center, ShenZhen. He has 9 years of experience with
mainframe systems. He holds a degree in Computer Science from FuDan
University. His areas of expertise include SNA/IP integration, CICS dump
analysis, and performance tuning. He has taught extensively on the subjects of
systems programming and database administration.
Dore Teichman is an IBM Certified IT Architect and Open Group Certified

Master IT Architect with IBM. He has more than 25 years of experience in online
transaction processing, primarily with Base-24 and Tandem, where he served as
a Regional Designated Specialist for Performance. He holds a degree in
Computer Information Systems and Engineering from California Polytechnic
University. His areas of expertise are Performance Engineering, deployment and
optimization of high volume online transaction processing systems, and Server
Consolidation. Prior to joining IBM, Dore served as a Regionally Designated
Specialist for performance with Tandem Computers, Inc., and is the original
author of Guardian Performance Analyzer.
Thanks to the following people for their contributions to this project:
Paola Bari, Robert Haimowitz

International Technical Support Organization, Poughkeepsie Center
Stephen Anania
IBM Poughkeepsie
Ron Beauchamp, Rick Doty, Ajit Godbole, Jeff Hansen, Karen Jarnecic, Kurt
Lawrence, Charlie Linberg, Catherine McCarthy, Calvin Robertson
ACI Worldwide
Preface xi
Become a published author
Join us for a two- to six-week residency program! Help write an IBM Redbook
dealing with specific products or solutions, while getting hands-on experience
with leading-edge technologies. You'll team with IBM technical professionals,
Business Partners and/or customers.
Your efforts will help increase product acceptance and customer satisfaction. As
a bonus, you'll develop a network of contacts in IBM development labs, and
increase your productivity and marketability.
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 Redbooks™ to be as helpful as possible. Send us your comments

about this or other Redbooks in one of the following ways:
򐂰 Use the online Contact us review redbook 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
xii A Guide to Using ACI Worldwide’s BASE24-es on z/OS

1
Chapter 1. Running a payments system

Banks and financial institutions are highly motivated by competitive forces to
broadly deploy automated teller machines (ATMs) and engage in electronic funds
transfer (EFT) activities. “The size of the ATM network has a significant impact on
the demand for bank deposits”.1 Customers consider the number and location of
ATMs in a bank’s network when choosing a bank or financial institution, and have
demonstrated that they are willing to pay a surcharge for convenience. While
increasing customer convenience, ATMs also save financial institutions money
and increase profits by reducing resource costs through automation.
In this chapter we provide an overview of bank payment systems and

requirements, and then describe how System z offers the availability,
manageability, security, and workload balancing to help you meet those
requirements.
1
Knittel, C., and V. Stango. 2004. "Compatibility and Pricing with Indirect Network Effects: Evidence
from ATMs." NBER working paper 10774
© Copyright IBM Corp. 2006. All rights reserved. 1

1.1 Industry growth
Bank card processing (ATM, point-of-sale (POS) and credit), more than any other
activity, wields an immediate and lasting impact on the reputation and image of a
financial institution. Banks and credit card companies encourage customers to
use their cards, inviting them to essentially live “cashless”. As a result, the ATM
and POS marketplace has undergone a rapid, expansive change in recent years.
“Installation of ATMs and the proliferation of Retail POS has been particularly
rapid in recent years. ATM growth was 9.3 percent per year from 1983 to 1995
but accelerated to an annual pace of 15.5 percent from 1996 to 2002. Much of
the acceleration is due to placing ATMs in locations other than bank offices.
These off-premise ATMs accounted for only 26 percent of total U.S. ATMs in
1994, but now account for 60 percent. On the debit card side of the industry,
growth has been extremely rapid in point-of-sale (POS) debit card transactions.
With an annual growth rate of 32 percent from 1995 to 2002, POS debit is the
fastest growing type of payment in the United States. Today it accounts for nearly
12 percent of all retail non-cash payments, a fivefold increase in just five years.
Growth has been sharp in both online (PIN-based) and offline (signature-based)
debit. From 1995 to 2002, annual growth of online debit was 29 percent, while
offline debit grew at 36 percent.”2
1.1.1 The requirement for reliability

Few issues impact the reputation of a financial institution, retailer, or customer
more than ATM/credit card processing reliability. Trying to explain that a card
being declined is the fault of the bank is a difficult task, and failures at the
point-of-sale embarrass customers and damage the reputation of the bank itself.
As the transition to a cashless society continues, customers will quickly revert to
using cash, and—more damagingly—take their business elsewhere, over a
single embarrassing event.
Moreover, the brand name of the ATM has to be considered. Maestro™, Star,
and others value their name brand and prefer to avoid being the target of poor
performance complaints by banking customers.
1.1.2 System availability

Typically, the issuing financial institution or bank has about 10 seconds for
domestic transactions to respond before the network begins to stand in. In some
2
Knittel, C., and V. Stango. 2004. "Compatibility and Pricing with Indirect Network Effects: Evidence
from ATMs." NBER working paper 10774.
2 A Guide to Using ACI Worldwide’s BASE24-es on z/OS

instances, such as an overseas banks and gateway transactions, more time is
allocated.
In the event that a financial institution is unable to respond to a transaction within

a specified time period (usually about 10 seconds), the ATM or POS network may
have the right to “stand in” for the bank, and following specified rules, then
approve or decline the transaction. This could potentially expose the bank to
overdraft of the customer’s account, and the possibility of approving a fraudulent
transaction. The specific rules and charges for stand in are usually the result of a
negotiated agreement between the financial institution and the shared network.
A financial institution or card issuer with unsatisfactory or poor reliability or poor

response-time performance may have to accept the risk of a shared network
“standing in” for the bank, an activity that is neither free, nor without risk.
System reliability and data integrity

A financial transaction may pass through several switches before it is completed
on the round-trip journey from acquisition, to authorization, and back to its
starting point, ATM or POS station. When everything goes according to plan, the
transaction completes and is prepared for final settlement which may take place
real-time, or in some form of batch processing scenario. When failures and errors
occur, the host issuer authorizing system must keep track of all transactions,
never losing a single one.
There are several methods available for meeting the response time, availability,
and data integrity requirements imposed on financial processing systems. These
techniques go beyond simple fault-tolerance; the system must be continuously
available. Fault tolerance is not a requirement for high availability; it is simply one
of the tools that contributes to the high availability capability.
1.1.3 Basic banking transaction categories

ATM or POS processing can occur across multiple paths, either directly (via a
bank’s owned ATM typically called “On Us”), or through one of many shared
networks where the card holder is using a terminal that is owned and operated by
a entity.
An example of the cooperative networks is STAR, owned by First Data

Corporation. With more than 1.7 million locations across the United States,
handling over 5,700 financial institutions with over 134 million cards, STAR is just
one of the many networks available to banks and other financial institutions that
can be used to project their financial services to their customer base.
Chapter 1. Running a payments system 3

Common shared networks do more than simply provide credit card and ATM
processing. They may also provide other services to financial institutions such
as:
򐂰 Debit and ATM network
򐂰 Surcharge-free programs
򐂰 Deposits, deposit sharing program
򐂰 Gateway connections
򐂰 ATM driving
򐂰 PIN-secured debit, signature debit and stored value card processing
򐂰 Card authorization, activation and production
򐂰 Merchant acquirer and agent bank programs
򐂰 Bill payment services
򐂰 Risk management and fraud prevention services
Other industry competitors in this marketplace worldwide are Plus, Cirrus,

Maestro, Pulse, Multibanco, Interlink, StarNet, Alto, Midas, Euronet and others.
In total, there are more than 60 major networks worldwide available to both
consumers and financial institutions.
Shared ATM networks provide added availability to financial institution

customers, although they can also introduce a certain degree of risk. Most
networks require stand-in authorization, which enables a network to authorize
transactions when a card issuer or processor cannot do so. However, this in turn
may lead to an increased risk of fraud.
To cite a useful example of the levels of uptime that are typically maintained, a
major bank ATM processing system in The Netherlands has had zero downtime
over the past four years, including maintenance and upgrades. This type of
reliability is not unusual for the banking industry; even systems in developing
countries maintain availability targets higher than 96%.
1.2 How ATM and POS processing is accomplished

In many ways, ATM and POS processing and POS card processing are similar
and follow similar paths though the network and associated systems. Four
categories of transaction can take place, as described in Table 1-1.

Table 1-1 Transaction switching schemes
Category Description
On Us The transaction arrives via a bank-owned ATM or POS

device and is never routed outside the bank.
Example: A customer of the “SampleA” Credit Union uses
an ATM card to withdraw money from the ATM located in
the local credit union office.
Network On Us The transaction originates from a sharing network such

as STAR or Cirrus, in which both the bank and the ATM or
POS-owning bank are members of the same network.
Example: A local customer travels somewhere away from
home and is a member of “SampleA” Credit Union. The
customer needs to use an ATM or POS at the “SampleB”
Credit Union. Both “SampleA” and “SampleB” are
members of the STAR network.
Reciprocal Transaction The card holder initiates a transaction at an ATM or POS

owned by a bank that is a member of a different regional
network.
Example: A New York resident attempts to withdraw
money from an ATM or POS in Omaha, Nebraska. An
agreement between the network in Nebraska and the
network in New York allows the transaction to be switched
from one regional network to another.
National Bridge The card holder uses an ATM or POS at a bank not their
Transactions own, and the two banks belong to different regional
networks that do not have any agreement. Both banks
must belong to the same national network. The
transaction is handed from the ATM or POS regional
network to the national network, and finally to the
authorizing bank’s regional network. In this case, there
are three switches involved.
Transactions are routed through the originating ATM or POS, either the bank’s
own network or through some combination of regional or national network.
Figure 1-1 demonstrates, at high level, the possible paths that a transaction
might follow to completion.

"Network On Us"
"On Us"
Bank/Financial Institution Auth System
Bank-Owned Bank-Owned
Network Regional Network
"On Us" Network "On Us"
National Bridge or Reciprocal
Databases Databases
or Record or Record
National
Network
Financial
Institution Bank/Financial Institution Auth Systems
Network
"On Us" Regional
Network
Databases
or Record
Figure 1-1 Transaction flows
Both the regional and national networks are switching systems that, to a fair
extent, resemble the systems used within the financial institution. The switching
systems drive transactions from initiation to destination. We use the word
“switched” to describe the hand-off of a transaction from host-to network-to host.
򐂰 An ATM or POS transaction is accepted by the host of the card-issuing
institution, and either processed locally “On Us”, or switched to one of the
regional networks “Network On Us”.
򐂰 The regional networks either switch the transaction to another institution host
for processing “Network On Us”, or switch the transaction to a national
network for hand-off to another regional network “National Bridge, or
Reciprocal”.
򐂰 Responses are switched from the owning host of an ATM or POS “On Us”, to
the ATM or POS, or switched to a regional network “Network On Us”.

򐂰 The regional network either switches a Host Response to an ATM or POS
transaction to the host of a member bank “Network On Us”, or switches the
response to a national network “National Bridge, or Reciprocal”.
򐂰 The national network switches the response to another regional network for
later switching to the card-issuer host.
Although there are other variations on this theme, transactions are switched from
host to host via network switches until the original transaction request is
completed by a transaction response, or it times out.
1.2.1 Transaction acquisition and authorization

ATM and POS transactions have to traverse three basic steps through complete
processing: Acquisition, Authentication, and Authorization. (Back-end settlement
and reconciliation activities are not a part of the actual real-time transaction.)
Note that BASE24-es does not have different paths for Authentication and
Authorization. Both functions are performed as part of a single task.
1.3 How System z addresses ATM/EFT/POS processing

requirements
In the following sections we list the processing requirements that can be met
when running on a System z platform. For further details, refer to Chapter 3,
“System z and z/OS” on page 23.
1.3.1 Availability
System z provides 24-hour a day, 7-day a week availability, including scheduled
maintenance. Continuous availability goes beyond just hardware fault tolerance;
it is achieved by a combination of hardware, application code, and good system
management practices.
The System z servers are the result of a long evolution beginning in 1964. The
sysplex configuration is a high availability solution that not only provides a
platform for continuous availability, but also for system maintenance and capacity
additions. The System z hardware platform is capable of providing up to five 9s of
availability.
On a server basis, System z systems are equipped with features that provide for
very high availability exclusive of clustering:
򐂰 Redundant I/O interconnect

򐂰 Concurrent Capacity Backup Downgrade (CBU Undo)
򐂰 Concurrent memory upgrade
򐂰 Enhanced driver maintenance
򐂰 Capacity backup upgrade
򐂰 On/Off capacity
1.3.2 Parallel Sysplex clustering

Within a Parallel Sysplex® cluster, you can construct a parallel processing
environment with no single points of failure. Because all systems in the Parallel
Sysplex can have concurrent access to all critical applications and data, the loss
of a system due to either hardware or software failure does not necessitate loss
of application availability. Peer instances of a failing subsystem executing on
remaining healthy system nodes can take over recovery responsibility for
resources held by the failing instance.
Alternatively, the failing subsystem can be automatically restarted on still-healthy

systems using automatic restart capabilities to perform recovery for work in
progress at the time of the failure. While the failing subsystem instance is
unavailable, new work requests can be redirected to other data-sharing
instances of the subsystem on other cluster nodes to provide continuous
application availability across the failure and subsequent recovery. This provides
the ability to mask planned as well as unplanned outages to the end user.
Parallel Sysplex clustering is designed to bring the power of parallel processing

to business-critical System z9™ applications. A Parallel Sysplex cluster consists
of up to 32 z/OS images coupled to one or more Coupling Facilities (CFs or
ICFs) using high-speed specialized links for communication.
The Coupling Facilities, at the heart of the Parallel Sysplex cluster, enable high
speed, read/write data sharing and resource sharing among all the z/OS images
in a cluster. All images are also connected to a Sysplex Timer® to ensure all
events are properly sequenced in time.
Figure 1-2 on page 9 presents an overview of Parallel Sysplex.

Coupling
Technology
Sysplex
Timer
System z System z
ESCON/FICON
Shared
data
Figure 1-2 Parallel Sysplex overview
When configured properly, a Parallel Sysplex cluster has no single point of failure
and can provide customers with near continuous application availability over
planned and unplanned outages. Events that otherwise would seriously impact
application availability (such as failures in hardware elements or critical operating
system components) have no, or reduced, impact in a Parallel Sysplex
environment.
1.3.3 Manageability
A wide array of tools, including the IBM Tivoli® product and other operational
facilities, contribute to continuous availability. IBM Autonomic Computing
facilities and tools provide for completely fault tolerant, manageable systems that
can be upgraded and maintained “on the fly” without downtime.
Autonomic computing technologies that provide Self-Optimizing,

Self-Configuring and Self-Healing characteristics go beyond simple hardware
fault tolerance. Additionally, the System z hardware environment provides:
򐂰 Fault Detection
򐂰 Automatic switching to backups where available
򐂰 (Chipkill™ memory, ECC cache, CP, Service Processor, system bus,
Multipath I/O, and so on)
򐂰 Plug and Play and Hot swap I/O
򐂰 Capacity Upgrade on Demand

The combination of tools, facilities, hardware design, and software architecture
provides 100% uptime.
1.3.4 Security
On March 14, 2003, IBM eServer™ zSeries 900 was the first server to be
awarded EAL5 security certification. The System z architecture is designed to
prevent the flow of information among logical partitions on a system, thus helping
to ensure that confidential or sensitive data remains within the boundaries of a
single partition.
On February 15, 2005, IBM and Novell announced that SUSE Linux Enterprise
Server 9 successfully completed a Common Criteria (CC) evaluation to achieve a
new level of security certification (CAPP/EAL4+). IBM and Novell also achieved
US DoD Common Operating Environment (COE) compliance, a Defense
Information Systems Agency requirement for military computing products.
On March 2, 2006, z/OS V1.7 with the RACF® optional feature achieved EAL4+
for Controlled Access Protection Profile (CAPP) and Labeled Security Protection
Profile (LSPP). This prestigious certification assures customers that z/OS V1.7
has gone through an extensive and rigorous testing process and conforms to
standards sanctioned by the International Standards Organization.
1.3.5 Expandability
The Parallel Sysplex environment can scale near linearly from 2 to 32 systems.
This can be a mix of any servers that support the Parallel Sysplex environment.
The aggregated capacity of this configuration meets every processing
requirement known today.
The Capacity Upgrade on Demand (CUoD) capability allows you to

nondisruptively add one or more Central Processors (CPs), Internal Coupling
Facilities (ICFs), zSeries Application Assist Processor (zAAP), and Integrated
Facility for Linux (IFLs) to increase server resources when they are needed,
without incurring downtime.
Capacity Upgrade on Demand can quickly add processors up to the maximum

number of available inactive engines. Also, additional books (up to a maximum of
four in total) can be installed concurrently, providing additional processing units
and memory capacity to a z9-109 server.
In addition, the new Enhanced Book Availability function also enables a memory
upgrade to an installed z9-109 book in a multibook server. This can provide
customers with the capacity for much needed dynamic growth in an
unpredictable ATM/EFT world.

The CUoD functions are:
򐂰 Nondisruptive CP, ICF, IFL, and zAAP upgrades
򐂰 Dynamic upgrade of all I/O cards in the I/O Cage
򐂰 Dynamic upgrade of memory
1.3.6 Dynamic workload balancing

To end users and business applications, the entire Parallel Sysplex cluster can
be seen as a single logical resource. Just as work can be dynamically distributed
across the individual processors within a single SMP server, so too can work be
directed to any node in a Parallel Sysplex cluster having available capacity. This
avoids the need to partition data or applications among individual nodes in the
cluster, or to replicate databases across multiple servers.

2
Chapter 2. Introducing BASE24-es

At the heart of any payments network is a payment engine. The payment engine
must acquire, switch, route and authorize transactions from a variety of sources.
ACI Worldwide is a leader in mission-critical e-payment software, offering
solutions to manage transactions from the point of access to settlement.
BASE24-es® is part of the ACI Payments Framework, a broad suite of products
developed by ACI.
In this chapter we describe the BASE24-es product and its functions and
features.

2.1 BASE24-es
The ACI BASE24-es product represents the latest in payment engine technology,
offering robust features and functions to the financial payments industry.
2.1.1 Consumer transaction support

BASE24-es provides comprehensive support for consumer e-payment
transactions initiated by a variety of transaction instruments that include credit,
debit and chip cards.
As the payment industry evolves and new instruments emerge (for example,
customer ID, mobile telephone numbers), the flexible nature of its architecture
enables BASE24-es to easily adapt to provide continued value.
Today, BASE24-es supports a comprehensive cardholder and administrative

transaction set that can be accessed from any appropriate delivery channel.
Administrative transactions are also supported for settlement and reconciliation
purposes.
BASE24-es can be configured to maintain card and associated account

information. Authorization logic can be scripted to perform a variety of tasks
including check current status of the card, compare cardholder use against limit
profiles, determine whether the transaction is allowed based on a number of
configurable options, and more. In addition, BASE24-es can provide alternate
routing or stand-in authorization if a configured primary external authorizer
becomes unavailable.
2.1.2 Transaction switching and routing

BASE24-es provides a highly flexible routing structure for transactions. This
flexibility not only routes transactions to the appropriate network, card
association, processor or internal system for authorization, but it also helps users
gain lower interchange charges by factoring in the total path when determining
the authorization destination.
Transactions are routed based on a combination of the following:

򐂰 Source Profile
򐂰 Destination Profile
򐂰 Transaction Type
򐂰 Account Type 1 (FROM account type)
򐂰 Account Type 2 (TO account type)
򐂰 Method of consumer authentication (PIN present, chip card, and so on)

This flexibility in transaction routing accommodates different account types that
may reside on different systems, as well as different platforms. Users can
customize their transaction processing at various stages in the transaction life
cycle, including:
򐂰 Pre-screening before transactions are sent to an external authorizer
򐂰 Defining the processing steps for real time internal authorization
򐂰 Defining the processing steps for stand-in authorization
򐂰 Specifying the destination of advice messages following authorization
򐂰 Specifying how the database should be impacted during the
post-authorization process
2.1.3 Flexible authorization

BASE24-es supports consumer authentication and authorization processing via
a powerful scripting engine. The scripting engine enables organizations to control
and define application logic without having to modify product source code.
Using an interpreted scripting language with JavaScript™-like syntax,

BASE24-es allows users to create a variety of authorization scripts to tailor the
authorization logic to meet specific business requirements or service
agreements. Organizations can decide what data is used in the authorization
process and when the data is used, regardless of whether the data is part of the
transaction or from an alternative source.
ACI provides a set of sample scripts that cover the basic positive, negative or
usage-based authorization processes. This flexibility shortens the time needed to
develop new products and services or accommodate changes requested by the
business department of an organization. Separating the business logic (in
scripts) from the product source code also facilitates script compatibility with
future releases.
Users can choose what level of authorization should be performed on

BASE24-es. This can range from full authorization using a card, limit, account
and balance information, to handling pre-screening before using a host for
authorization, to just providing a stand-in capability using negative card data. All
limits are user-defined to provide full flexibility in controlling use of the cards and
accounts.
2.1.4 Integrated consumer data

The component architecture of BASE24-es is designed for flexibility to leverage
consumer data resident in external systems and databases. Users can develop
Chapter 2. Introducing BASE24-es 15

components that expose consumer data to the scripting engine to allow more
intuitive authorization functionality. This consumer data can be held in customer
relationship management (CRM) systems, fraud management systems,
customer information files, core banking systems and other applications.
By exposing more consumer information to the authorization process,

organizations can improve consumer relationships by approving transactions
based on more comprehensive consumer information. They can also intelligently
manage risk by denying transactions based on certain risk factors.
2.1.5 Traditional and emerging delivery channels

In addition to support for traditional delivery channels, including ATM and
point-of-sale (POS), BASE24-es can process payments initiated through Internet
shopping networks, personal digital assistants (PDAs), mobile telephones, Web
ATMs, and home banking and branch systems. BASE24-es offers a powerful,
flexible foundation for delivering common services across multiple consumer
access channels, computer systems and databases.
Through the use of XML- and ISO-based interface standards and other
industry-specific formats, transaction services can be exposed to any channel.
Thus, BASE24-es can provide a single point of access across an enterprise for
the service of consumer payments, eliminating the costs of maintaining multiple
service points.
2.1.6 Reliable security infrastructure

Organizations’ transaction security requirements can vary greatly depending on
the environment. An organization typically requires an integrated system of
software, industry-standard hardware, and procedures to properly implement
financial transaction security.
BASE24-es is designed to be flexible in its transaction security support and to

provide a range of hardware options. The application addresses the diverse
needs of large-scale transaction processing systems where the originator of a
transaction may operate under an entirely different transaction security scheme
than the authorizer. Regardless of the origin or destination of payments,
BASE24-es meets the current industry requirements for security, including Triple
DES and EMV support.
BASE24-es operates in a network environment where sensitive data, such as the

personal identification number (PIN), is secured via encryption, and the system
provides cryptographic functions such as PIN encryption, PIN verification,
message authentication, chip authentication and card verification using
interfaces to external hardware security modules (HSMs).

2.1.7 Intuitive graphical user interface
The ACI user interface presents a task-oriented view of the application for
multiple users ranging from business to technical to administrative, and it
incorporates graphical elements such as hyperlinks, buttons and pull-down
menus. Users can also choose to display text labels in their local language to
accommodate adaptation into their environment. Integrated help at the window
level and the field level minimizes the need for extensive user training, while
streamlining business processes and providing greater flexibility.
Written in Java and C++ and using XML message formats, the ACI user interface
provides a flexible operating environment that is used by multiple ACI
applications. A security administrator configures access permissions through the
ACI user interface where a user security and user audit environment are shared
by all ACI applications.
2.1.8 Scriptable extracts and reports

The BASE24-es scripting engine gives users added flexibility and control over
the reporting and extracting of financial transaction data. In an area where
customization is virtually required to meet integration needs, the scripting feature
allows real time definition of essential and ad hoc reports, as well as user-defined
file layouts to serve as input to existing batch processes or reporting tools.
2.1.9 National switch and international card scheme interfaces

BASE24-es incorporates off-the-shelf support for a range of international card
scheme interfaces, including Visa, MasterCard and American Express, as well as
many national switch interfaces. These interfaces are built using a framework
methodology covering ISO 8583 (1987), ISO 8583 (1993) and XML standards.
This methodology makes use of “inheritance” to facilitate reusability of
components and allows new interfaces to be built quickly by either ACI or the
customer organization.
2.2 Architecture
ACI uses an object-oriented design and development style to implement its
Enterprise Services architecture of BASE24-es. This architecture helps reduce
impacts associated with extending the core product code. The use of
object-oriented programming languages, such as C++ and Java, enhance the
extensibility of BASE24-es solutions and minimize time-to-market for new
products and services. By extending integration flexibility, BASE24-es allows
access to more customer information.

BASE24-es software components use this architecture to create flexible
business services that allow users to quickly develop and deploy new products
and deliver enhanced customer service. The components are organized
according to the function they perform to support processing for the required
business services. Each business component performs a specific type of
processing (that is, authorization or routing) or controls a specific part of the file
system (for example, account or customer).
The architecture of BASE24-es is designed for multiple platform configurations.

The platform is defined as the hardware, operating system, middleware, and file
system. However, platform-specific processing is isolated into specific
components to allow the rest of the application to be common across all
platforms.
BASE24-es software components are organized according to the function they

perform, as described here:
򐂰 Adaptors manage the information exchanged between the end user and the
business components. Adaptors can be designed to communicate with any
acquiring or issuing entity, including Internet portals, hardware devices,
service providers or interchanges (Visa, MasterCard, and so on), and
in-house systems.
򐂰 Business components perform the processing required for the business
services offered. Each business component performs a specific type of
processing (that is, authorization, routing) or controls a specific part of the file
system (for example, prefix, perusal).
򐂰 Foundation components provide information and processing required by
business components regardless of the software module. These are the basic
building blocks for any application. For example, the time component obtains
the current system time in the format needed by the application. When any
business component needs the system time, it obtains it from this foundation
component.
򐂰 Platform-specific components insulate the application from changes in the
operating system, middleware, and data structure. For example, processing
performed by the time component differs across platforms.

2.3 Operability
BASE24-es supports high volume, high availability, 24/7 operability through a
scalable, high available software architecture that runs on a variety of platforms,
as explained here:
򐂰 Flexible journal configuration and settlement cutover
This allows for 24/7 cutover processing and uninterrupted processing across
separate time frames.
򐂰 Implement new business logic without downtime
Code and file system changes that affect configuration do not require a
system restart.
򐂰 Hardware resilience
The system and data access layers take advantage of each platform’s
failover processing capabilities, all with the same set of application code.
򐂰 Consistent processing cost
The asynchronous messaging model of BASE24-es provides a consistent per
transaction processing cost regardless of the transaction volume, which
allows the application to grow as needed with a predictable hardware
requirement.
2.4 Scripting component

The BASE24-es scripting facility gives organizations a powerful JavaScript-like
syntax to allow modification of application logic without having to modify the
source code. The application uses these scripts to create journal perusal
queries, define journal extract and reporting requirements, and as part of the
authorization process.
Scripts are maintained and compiled through the user interface. Users can
display a script repository that shows all scripts available for use by BASE24-es.
A script editor allows the user to add, edit, delete and compile scripts through the
user interface. During the compile process, scripts are checked for syntax errors
and saved on the BASE24-es system. Rather than compiling into machine code,
these scripts are ordered into a list of serial instructions that the script engine
may interpret in real time during transaction processing.
Compiled scripts are loaded into memory where they can be retrieved for
execution during transaction processing. If a change must be made to script
logic, then the script can be updated, recompiled and placed back into use
without ever taking the affected programs out of service.

Scripts have access to data from multiple sources; the primary source is from
transaction data elements. Application files for authorization are also available to
the script facility, and these include card file, limits file, usage accumulation file,
account with balance file and pre-authorization file. For greater decision-making
flexibility, access to additional customer information can be obtained through
custom written components that “expose” the proprietary structure of a
customer’s file to the script. These custom files may be core banking system
files, ACI or other third-party card management systems, or a variety of other
sources.
A single script can contain all of the tasks that BASE24-es must perform to
authorize a transaction. The tasks can also be split into multiple scripts that are
organized in a hierarchical structure.
Because of the component-based design of the application and the scripting

language, scripts implemented by the user may not need to be modified when a
new release of the application becomes available, allowing users to easily
upgrade to new releases of the product. If a user plans to take advantage of new
functionality from within the script, then changes will be required.
2.5 The user interface

The BASE24-es user interface employs Java, C++ and XML technologies,
providing the user interfaces needed to manage all components of the
application.
With the system’s built-in user security feature, users are assigned roles that
grant them permission to specific functions and tasks associated with various
windows. Users are authenticated during the logon process, thereby minimizing
the risk associated with unauthorized users gaining access to functions they are
not permitted to perform.
The user audit function is responsible for maintaining a secure audit database
where all file maintenance transactions and modifications to the user security
database are recorded. Before and after images of the affected record will be
logged wherever appropriate.
The user interface design incorporates the flexibility for users to alter the layout
and wording on the desktop to meet individual organization needs. All text and
positional information is maintained in configuration files, so adapting the user
interface without altering the product code is particularly easy. This structure also
incorporates multi-language capability.

2.6 Rich functionality
BASE24-es provides full functionality to support payment transactions across
multiple channels. The software is parameter-driven, allowing users to configure
a system that will meet their unique business requirements. ACI’s product
investment strategy accommodates periodic new releases of software providing
support for both regulatory changes as well as new trends in electronic delivery.
2.7 Platform independence

BASE24-es supports a broad range of computing environments, allowing
customers to operate ACI’s best-of-breed software on their choice of
industry-standard platforms. BASE24-es operates on a variety of HP, IBM and
Sun™ servers. On each platform, our ACI software takes advantage of the best
in systems software for reliability, availability and high-performance throughput.
2.8 Flexible architecture

Since the design of BASE24-es includes support for scripting, there is little need
for customer technical staff to have knowledge of the core languages of the
application. A working knowledge of JavaScript programming methods will
prepare an experienced programmer for maintenance and creation of the
business logic necessary to meet the institution’s authorization processing
needs.
Because the BASE24-es application is component-based, ACI customers have

the freedom to develop components in-house, which extends product
functionality. To accomplish this task, some basic skills concepts are required as
outlined below:
򐂰 UML (Unified Modeling Language)
򐂰 C++
򐂰 Java
2.9 The ACI Payments Framework

ACI offers mission-critical e-payment software solutions to manage transactions
from the point of access to settlement. BASE24-es is part of the ACI Payments
Framework, a broad suite of products developed by ACI. The ACI Payments
Framework includes software to enable transaction processing through evolving

Internet and wireless channels as well as traditional ATM and POS channels.
ACI solutions process transactions in real time and automate the back-office
functions associated with settlement, dispute processing, fraud detection and
account service.

3
Chapter 3. System z and z/OS

In this chapter we describe the features and the quality of service provided by the
IBM System z platform and the z/OS operating system.

3.1 How z/OS addresses non-functional requirements
In this section we describe the traditional strengths of the mainframe. We focus
on the qualities that the mainframe provides, thereby addressing the relevant
non-functional requirements of most mission-critical IT systems.
3.1.1 Background
For more than 40 years, IBM mainframes have successfully run the core IT
systems of many mid-size, large, and very large enterprises. Over this period,
IBM has consistently invested in the evolution of the mainframe’s unparalleled
technology. Mainframes have incorporated new technologies and computing
models, ranging from the centralized model to computing to the Web model, and
from assembler and COBOL languages to Java.
IBM mainframes have become the computing industry benchmark. They are
used by thousands of enterprises worldwide, with trillions of dollars invested in
applications and skills. The platform hosts a large portion of the total business
transactions and data in the world, providing 24x7 availability and other unique
qualities.
IBM mainframes run five different operating systems, consisting of z/OS, z/VM®,
z/VSE™, z/TPF and Linux on System Z. In the following section we focus on the
z/OS operating system, the flagship mainframe operating system.
3.1.2 Traditional z/OS strengths

The generally recognized strengths of z/OS fit into a number of broad categories:
򐂰 Centralized computing model
򐂰 Security
򐂰 Manageability
򐂰 Virtualization and workload management
򐂰 Reliability
򐂰 Scalability
򐂰 Availability
򐂰 Communications and I/O
򐂰 Transaction processing
򐂰 Batch processing
To provide these, the z/OS operating system includes a comprehensive set of

capabilities and tools out of the box. z/OS provides its unique Quality of Service
(QoS) in combination with System z hardware and Parallel Sysplex capabilities.
In the following sections we discuss each of these areas in more detail.

3.1.3 Centralized computing model
We define a computing model as a structure that organizes the way
communication and sharing occurs. A computing model describes the topology
of interconnected computer resources, as well as how users access these
resources.
In the history of commercial computing, we have seen three major computing

models: centralized, client/server, and network computing, as described here:
򐂰 The centralized model is characterized by a single (or a few) large computing
nodes.
In this model, a variety of applications are hosted on the nodes, ranging from
tens to several hundred or even thousands of applications. The applications
all share the same computing resources such as memory, CPU, and disk and
tape storage. All users are directly connected to this system.
򐂰 The client/server model spreads computing resources over a large number of
distributed computing nodes that do not share the same resources.
In this model, each computing node typically runs a single (or just a few)
applications, and the applications communicate over the network. The
objective of this model was to offload processing from the central computers.
򐂰 The network computing model (also known as the Web model) allows users
to connect to applications using a Web browser, which is independent of a
particular operating system or hardware.
All data, business logic, and applications are brought back to servers, and
only the information required for display is sent to the user device.
Small applications (or applets) can be downloaded from the server and run in
the local memory of the terminal to improve functionality or performance. The
Web model allows users to connect to applications from anywhere over the
Internet.
Figure 3-1 on page 26 offers a simplified view of these computing models.
Chapter 3. System z and z/OS 25

COMPUTING MODELS
Network Computing
Client/Server
Centralized
Figure 3-1 Computing models
The centralized computing model was less used in the late 1980s and early
1990s due to a perceived lack of flexibility and openness, and high cost of
acquisition. By the end of the 1990s, however, this computing model regained
favor following improvements in mainframe hardware and software. Also,
businesses had by then come to understand the serious manageability
complications and security exposures of the client/server model, and realized
that centralized computing offered some clear advantages.
The implementation of the centralized computing model in the mainframe

hardware and the z/OS operating system differentiates itself by the provision of
integrated management and control capabilities to manage the complex
workloads running on the platform.
Another key differentiator offered by centralized computing is data proximity.

Having your applications in close proximity to the data they use significantly
reduces communication overhead, increases security, and minimizes points of
failure.
Resource sharing is a further key differentiator among computing models. For

example, in a centralized model every resource (storage, memory, processors,

I/O channels, and so on) is shared between the different applications. This leads
to overall better utilization of those resources and less idle time. In a distributed
model with dedicated resources, efficiency is significantly lower and idle time
higher.
Partitioning and Parallel Sysplex

Centralizing all components on a single system does not imply that mainframe
computing is limited to a single machine or to a single operating system. On the
contrary, mainframes have the capability of being partitioned on up to 60 logical
partitions (LPARs), and each LPAR has the ability to run one of the five
z operating systems. There are also management tools and resource sharing at
the machine level, and at the operating system level.
The z/OS Operating System allows you to create a cluster of up to 32 systems;

this is known as Parallel Sysplex. In Parallel Sysplex, a “system” can be a full
machine or just an LPAR. Most computing platforms today offer clustering
capabilities. Parallel Sysplex, however, is a completely different kind of clustering
solution because it can share every resource among the elements in the cluster,
and also dynamically reconfigure, add, or remove resources.
Notably, z/OS also allows all members in a cluster to share all data, even up to
the record level. In contrast, other cluster implementations allow you to partition
data among the elements of the cluster, and each system can access just the
data attached to it.
Parallel Sysplex also provides significant network optimizations for

communication across its cluster members. After a client request reaches the
Sysplex Distributor there is no more external network traffic required; all traffic
flows over the System z hardware. As a consequence, network latency is kept to
a minimum and the typical network issues you normally see in a physical
decentralized infrastructure are inherently absent.
Even when the Parallel Sysplex is physically spread over several different
machines, the communication between them flows over high speed fiber optic
connections managed by the Cross Coupling Facility (XCF), a specific protocol
for those connections with a magnitude of GigaBytes of transfer rate. Within a
physical machine, communication between z/OS images is accomplished
through memory-to-memory, and there is no network protocol faster than that.
Security
z/OS provides deep security integration, from the application level down to the
operating system level and the hardware. In this section we highlight a number of
major functions in z/OS.

Centralized security
Approximately 30 years ago IBM developed the Resource Access Control Facility
(RACF) to provide centralized security functions such as user identification and
authentication, resource access control, and auditing for both the operating
system and applications running on the system. As a next step, to improve the
usability of the system's security interfaces and thus encourage more
applications to use consistent system-level security rather than implement their
own separate security mechanisms, IBM implemented the System Authorization
Facility (SAF) within the MVS™ operating system. SAF combined the various
security function invocations into a single extensible security mechanism.
This new security structure provided several benefits to encourage use by

application programs and components of the operating system itself. Most
recently, IBM has again modified SAF to provide additional security interfaces for
the UNIX environment on z/OS.
With the introduction of the System Authorization Facility suite of services, a

centralized point was created within the operating system infrastructure through
which both operating system components and applications could invoke the
services of the z/OS resident security manager.
Auditing and logging

A very important feature of a centralized authentication and access control
mechanism is the ability to record and analyze security information. The audit
data is essential for ensuring that the client’s installation security policy is being
followed. The RACF option of the z/OS Security Server provides multiple ways to
specify what security-relevant events are recorded in the audit stream, and how
that information is reduced and analyzed. RACF provides a wide variety of
capabilities and tools to the auditor for data analysis.
Accountability
Accountability in z/OS is achieved by a combination of user authentication and
the ability to propagate a user’s credentials throughout the application.
The System Management Facility (SMF) interface is designed for collecting

performance and accounting information. z/OS will collect all accounting and
performance data and present this back to the installation through RMF™ and
SMF records. These records can be analyzed and aggregated in performance
and capacity reports, but can also be used as a base for security analysis and
accounting policies.
Cryptography
The System z hardware has two distinctly different cryptographic hardware
engines that are supported under z/OS; the CMOS (Complementary Metal Oxide

Semiconductor) Cryptographic Coprocessor and the newer PCI (Peripheral
Component Interface) Cryptographic Coprocessor (PCICC). The PCICC
provides the capability for rapid response to client requirements that was
sometimes difficult to achieve with the CMOS Cryptographic Coprocessor alone.
For further details about this topic, refer to 4.2.10, “AOR architecture” on
page 56.
Network security
Networking and communications security on z/OS is provided by the
Communications Server element of z/OS. The Communications Server provides
networking and communication services for accessing z/OS applications over
both SNA and TCP/IP networks.
Firewall Technologies within the Communications Server provide the

network-level protections. The Communications Server provides a further,
optional layer of protection by supporting multiple TCP/IP “stacks” which can
allow clients to configure TCP/IP communications in a way that segregates the
external (public) traffic from the internal traffic used in the private network. Thus,
the z/OS Firewall Technologies can provide extra protection for applications
using the stack that handles the external communications traffic.
Furthermore, the Firewall Technologies include packet filtering to limit the kind of
network requests that can reach a machine, proxy and SOCKS servers to control
TCP/IP connectivity, Domain Name Services (DNS), encryption of IP traffic (IP
tunnels, or Virtual Private Networks (VPN)) to allow private communication over
the public network. Other security features provided by z/OS and the
Communications Server include LDAP, PKI, Kerberos, and openSSH.
Manageability
The z/OS operating system has been designed to manage multiple workloads in
a single system image. This design has, from the beginning, put the
requirements on the operating system to provide the means to manage these
complex environments. As a consequence, over the years z/OS has become
equipped with an extensive set of management tools for managing and
controlling thousands of complex simultaneously running applications,
comprising batch and online components, databases, transaction managers and
so on.
The management tools included in z/OS have become a very mature set of
system management and automation software. Clients over the years have
based their sophisticated procedures and automation scenarios on this software.

These tools include:
򐂰 SAF (discussed in “Security” on page 27) is the security interface built into the
operating system, and RACF is provided for security administration and
management.
򐂰 SMF and RMF are used for resource management, performance
management, capacity planning capabilities and logging of system and
middleware events.
򐂰 SMP/E is used for product installation and maintenance.
򐂰 DFSMS provides data management software that automatically manages
data from creation to expiration. DFSMS provides allocation control for
availability and performance, backup/recovery and disaster recovery services,
space management, tape management, and reporting and simulation for
performance and configuration tuning.
򐂰 SDSF provides the abilities to monitor, manage, and control the z/OS sysplex.
򐂰 System Automation provides the capabilities to increase application
availability through policy-based self-healing, and to automate z/OS
Input/Output, processor and systems operations.
򐂰 Tivoli Workload Scheduler (TWS) provides functions to automate, plan, and
control the processing of an enterprise's entire production workload.
Virtualization and workload management

Why is System z the most trusted platform for core enterprise systems and
applications? One of the reasons for this is virtualization. Unlike most other
computer systems, the IBM Mainframe is designed from the silicon chip and up
for virtualization. By comparison, RISC and Intel® chips are architected to meet
computational benchmarks. What they gain in the processing of instructions,
they lose in task switching, error handling, data movement and input-output
operations. IBM mainframes create a virtual environment, not only for existing
and new CICS, IMS™, DB2® and Batch workloads, but also for workloads such
as Java (J2EE™) applications, Linux, UNIX and GRID applications. This means
that an enterprise can run its entire enterprise server workload within a virtual
mainframe environment.
Mainframes typically meet their service requirements at 80% to 100% utilization

levels, and the varying nature of application workloads means that a shared
infrastructure is more efficient than a set of dedicated servers. Peak mainframe
workloads are dealt with by reducing the allocation of resources to lower-priority
background tasks and through dynamic reallocation of resources.
The virtualization capabilities of IBM mainframes are diverse (see Figure 3-2 on
page 31).

Virtual Resources
Linux
CICS IMS
IFL
WebSphere z/OS
zAAP DB2 zVM
zVSE
HiperSockets – Low latency virtual network
Physical Resources
Memory
Memory
Memory
I/O Memory CPUs Networks
Memory
Figure 3-2 Mainframe virtualization capabilities
z/OS Parallel Sysplex capabilities allow configurations of large numbers of

partitions to be run as a single system image sharing resources. z/OS Workload
Manager (WLM) controls the distribution of workload over the partitions and
prioritization of work within a partition. Through Intelligent Resource Director
(IRD), CPU resources can be dynamically reassigned. WLM and IRD working
together provide unique on-demand capacity services. WLM can work with
Sysplex Distributor to load balance workload across the systems in the sysplex,
and make the network entry point into the system highly available.
All of these capabilities make the mainframe an ideal platform for running mixed
workloads of hundreds of different business-critical applications, thereby
achieving very high resource utilization. The mainframe manages these diverse
workloads while maintaining a consistent performance.
Reliability
System z and z/OS provide an advanced combination of availability and
resilience. The combination of redundancy and virtualization delivers

extraordinary levels of application availability and recoverability. IBM mainframe
architecture helps to protect end users from hardware and software failures. If an
application fails, its workload can be picked up by a backup of the same
application, running on the same physical hardware. On a more granular level,
z/OS uses a feature called Automatic Restart Manager (ARM) for automatic
recovery and restart of services and transactions based on predefined polices.
IBM mainframe software is designed for transactional integrity with system-wide

two-phase commit. That means that a client’s application data is recoverable,
even when hardware and software failures occur. z/OS provides a special set of
services for this activity, called Resource Recovery Services (RRS), which are
commonly used by middleware solutions, but which can also be used in custom
applications, as the APIs are public. When resources in the commit scope are all
on z/OS, RRS can provide two-phase commit functionality without the need for
distributed transaction technologies such as XA.
Mainframe workloads can be shared by systems running up to 100 km apart,

allowing for smooth disaster recovery procedures. And when the time comes to
add system capacity, or replace a failed hardware module, those hardware
changes can be done without interrupting customer service. Some mainframes
run uninterrupted for years, adapting to changing needs, workloads, and
applications while continuously in production.
System z hardware provides a solution with no single point of failure. There is

redundancy within the power supply and cooling, an inside battery for backup,
and built-in spare processors and memory. Another hardware advantage is the
ability to perform a hardware upgrade or hardware maintenance without taking
the machine down.
Scalability
System z hardware can be scaled up vertically to a maximum of 54 CPUs at this
time. The z/OS Parallel Sysplex configuration provides additional horizontal
scalability. A sysplex can be a cluster of up to 64 z/OS images in different
partitions on different System z machines (possibly geographically dispersed),
with full data sharing and high availability and recovery.
The machines participating in a Parallel Sysplex can be physically dispersed up

to 100 km distance each to the other, which gives the capability of having a
physically decentralized infrastructure being logically centralized.
In z/OS, the horizontal scalability is not limited to specific workloads. Any

workload can take advantage of the Parallel Sysplex scaling options.
Scalability is fundamental in an on demand world, and the IT infrastructure

should be able to scale when business demands more resources. This does not

necessarily mean a new server or system image must be integrated to the
complex in some way or another, as is required on distributed platforms.
Mainframes offer more capabilities to scale on demand. Other platforms scale
out, while mainframes scale in, as explained here:
򐂰 Scaling out means adding resources to a logical system by adding a physical
system (or operating system image), thereby increasing the configuration
complexity and reducing manageability.
򐂰 Scaling in means (automatically) aggregating new resources to the existing
infrastructure white-hot increasing its complexity, without demanding
redesign of data bases, application patterns, network traffic and so on and
without the need to bring the system down.
This is another mainframe differentiator, by design. For instance, you can add
new processors, new storage, or reconfigure partitions, without any impact on
applications, data, and existing infrastructure.
Another key performance and scalability benefit of mainframes is provided by the

HiperSockets™ virtual TCP/IP network, which eliminates network delays
between applications and subsystems running within a mainframe.
Availability
System z has many components to address availability, as well as z/OS.
Parallel Sysplex, the clustering solution for the z/OS environment, provides both
scalability and availability. With Parallel Sysplex, a failure of one image in the
cluster does not affect any other image. And any specific transaction running on
the failed image can be fully dispatched and recovered in any other image, thus
making use of the full data sharing architecture.
For higher availability and disaster recovery purposes, a Parallel Sysplex can be
configured in a Geographically Dispersed Parallel Sysplex™ (GDPS®) mode.
There are two GDPS modes:
򐂰 GDPS/PPRC
GDPS/PPRC is a configuration in which a Parallel Sysplex is distributed
across two sites, connected together up to 100 km, with data synchronized
and shared continuously.
One site (part of the sysplex) acts as a primary, and the second site acts as a
secondary, in stand-by mode. GDPS controls and automates a full swap to
the backup site in case of failures.
򐂰 GDPS/XRC
GDPS/XRC is a configuration in which the distance between sites can be
more than 100 km, theoretically without limitation. In GDPS/XRC, the sysplex

does not span both sites. Instead, a full system image is swapped to the
alternate site in an emergency situation.
Both modes use HiperSwap, which allows you to activate replicated data in the
disaster recovery site without application outage.
Figure 3-3 illustrates a simplified view of z/OS continuous availability.
Single System Parallel Sysplex GDPS
12
11 1
10 2
9 3
8 4
12 1
7
6
5 11 1 12 11
10 2 2 10
9 3 3 9
8 4 4 8
7 6 5 5 7
6
1 to 32 systems Site 1 Site 2
Built-in Redundancy Addresses Addresses Site Failure

Capacity Upgrade on Planned/Unplanned Maintenance
Demand HW and SW Outages Sync/Async Data
Capacity Backup Flexible, Nondisruptive Mirroring
Hot Pluggable I/O Growth Eliminates tape/disk
Capacity beyond SPOF
largest CPC No/Some data loss
Scales better than Application-Independent
SMPs
Dynamic
Workload/Resource
Management
Figure 3-3 z/OS continuous availability
The System z and z/OS availability features are:

򐂰 Unique mainframe clustering technology for maximum up-time (Parallel
Sysplex)
򐂰 The ability to deploy participating nodes in the sysplex cluster remotely
(Geographically Dispersed Parallel Sysplex)
򐂰 The ability to recover virtual Linux servers running remotely (xDR)
򐂰 Replicate data real-time at remote locations (PPRC)
Communications and I/O

Over the past decade, focus has shifted from SNA networks and applications to
TCP/IP and Internet technologies. The TCP/IP protocol suite has become a
de facto standard for transmitting data over networks. But SNA is still used

extensively in banks and other financial transaction networks. In some cases,
SNA application traffic now runs over IP-based networks using Data Link
Switching (DLSw). In other cases, applications have been changed, and
processes reengineered, using TCP/IP rather than SNA.
For some organizations, the network traffic that traverses IBM SNA
communication controllers has declined to the point where it is in the business
interest to find functional alternatives for the remaining uses of these controllers
by consolidating and possibly eliminating the controllers from the networking
environments.
Estimates are that the current use of TCP/IP in the EFT, ATM, and POS
environments is more than 70% and rapidly increasing, but there are still a
significant percentage of ATMs on SNA protocol and some (less than 5%) on
X.25 in the POS segment.
BASE24-es supports VTAM®-attached terminals. The BASE24-es CICS VTAM

communication handler currently supports SNA LU 0 and 2. Other BASE24-es
VTAM protocols are scheduled on individual client business requirements.
Support for SNA on z/OS is provided by the Communication Controller for Linux
(CCL). The 3745/3746 SNA communications controller has been withdrawn from
the market.
With CCL you can continue using the traditional SNA subarea (INN), NCP
boundary (BNN), SNA Network Interconnect (SNI), and X.25 connectivity that
are currently supported by the IBM 3745.
Transaction processing
Traditionally, the mainframe has offered extremely robust transaction processing,
providing thousands of concurrent online users with the ability to retrieve
information and make updates. The IBM transaction management solutions on
the mainframe have since long been CICS and IMS on the z/OS platform, and
TPF as a specialized operating system on the mainframe for airline and financial
transactions.
WebSphere Application Server for z/OS was recently added to the family of
transaction managers on z/OS. Businesses have invested heavily in the
development of transactional applications, and despite efforts to mimic
transaction processing on non-mainframe platforms, these solutions have never
provided the same scalability and reliability as mainframe transaction processing.
As discussed, the z/OS operating system provides transactional services that are
not limited to applications running in one of the transaction managers. Resource

Recovery Services (RRS) assures integrity by providing transaction and
two-phase commit services to any application that requires those.
Batch processing
Another outstanding capability of the mainframe architecture is the running of
batch workloads. In the z/OS environment, dedicated subsystems (JES2 or
JES3), in combination with special job scheduling software such as Tivoli
Workload Scheduler (TWS), manage batch processing. No other architecture
can support batch processing as well as z/OS.
The value of batch processing is often underrated. With the requirement for
integrating new applications with existing back-end systems, batch processing
becomes ever more important. The IBM mainframe is well-positioned to handle
batch processing. Offering the ability to run batch processes written in the Java
language making use of zAAP processors, and the ability to access back-end
data either on z/OS or any other server, z/OS is a very cost-effective platform for
running business-critical batch applications.
Additional capabilities of z/OS

z/OS provides a number of facilities that provide unique implementation
capabilities for the products running on the z/OS platform. These capabilities
significantly increase the Qualities of Service to a level above what can be
reached on other platforms, while preserving functionality at the same level, at a
minimum.
Openness
Virtualization has provided “openness” to the mainframe, enabling it to run not
only the traditional operating systems of z/OS, z/VM, z/TPF and z/VSE, but also
to support Linux. Once called “proprietary” technology, the mainframe has
adopted open standards technology as well and is as open as any other
UNIX-based platform or the Windows platform. There are no limitations on the
mainframe in using standards such TCP/IP, HTTP, SOAP and so on.
Today’s mainframe not only supports open standards, but enables and integrates
traditional technologies with the new ones. Core business applications, usually
running in IMS or CICS, can be integrated with new-style applications using XML
and SOAP, thereby bridging the technology landscape and maximizing
investments in skill and resources.
Cost-effectiveness
Cost is one of the most sensitive and complex aspects of any solution analysis. It
is sensitive because everyone responsible has a different opinion of what
constitutes a reasonable cost. It is complex, because determining the total cost
of a solution requires a critical analysis of many factors.

Major consulting companies have published many studies related to cost
comparison among platforms, and delving further into this topic is beyond the
scope of this publication. However, in the following sections we offer some points
to consider.
TCA and TCO

Total Cost of Acquisition (TCA) refers to the buying cost of a certain product.
Total Cost of Ownership (TCO) refers to the whole cost involved in a process,
and it extends much further than just the buying cost. TCO takes into account all
aspects related to a product’s installation, maintenance, personnel and training
costs, the required infrastructure, and so on.
Companies often do not distinguish between TCO and TCA, because lack of
knowledge or lack of data makes full analysis impossible. Various consulting
companies specialized in performing this kind of analysis have published very
positive results for the System z platform, considering five years analysis.
Economies of scale
Due to its centralized computing model and almost endless scalability, the
System z platform provides economies of scale, meaning that you can add new
physical resources (CPU, I/O devices, servers) without the need to proportionally
expand the existing infrastructure and the number of staff managing it.
People are often the most expensive resource in a business. In the distributed
world, for every additional set of servers or pool of storage, the amount of human
resource needed will increase. On the System z platform, however, the number
of people supporting it is practically flat and only increases when very large
chunks of additional infrastructure are added.
As mentioned in “Scalability” on page 32, another key scalability differentiator

between the System z platform and distributed platforms is the difference
between scale in and scale out. Scalability is a major factor in achieving service
level agreements (SLAs). And with the increasing popularity of the Internet and
e-business, companies find it difficult to predict when and how often users will
access corporate Web sites and demand IT resources. As a result, system and
application scalability have became critically important to success in the world of
e-business. With System z, stability is accomplished on the same machine (scale
in). In contrast, on distributed platforms you need to add new machines (scale
out).
An “out-of-the-box” solution
z/OS is not just an operating system. It is a package that encompasses the
operating system itself and more than 30 built-in solutions. The full package
comes with everything that is required to build a robust IT software infrastructure.

Among others components, it comes with security services, communication
services, performance and accounting services, compilers, storage management
services, an HTTP server, clustering capabilities, and so on. Other operating
systems do not provide the equivalent set of services out of the box; instead,
clients need to purchase them separately.
IBM flexible charge model

IBM has introduced various new ways of charging for mainframe software, which
has significantly reduced the platform total cost of ownership (TCO). The most
recently added models include specialty processors for Java and DB2 tasks
called zAAP and zIIP, respectively.
Data proximity
As previously mentioned, data proximity delivers additional benefits. When
integration logic is deployed on z/OS close to the resources that are being
integrated, composition and integration with multiple z/OS resource managers
will give optimal performance because data access can be realized
cross-memory, requiring no network traffic and duration of held locks can be
minimized. This also significantly increases availability because the configuration
will comprise less points of failure. Also, recovery will be much faster in rollback
situations. The ability to run the transactions using native z/OS services rather
than a distributed 2-phase commit protocol offers far better performance.
Unique z/OS values

Table 3-1 maps the strengths of the z/OS platform to current products and
solutions.
Table 3-1 Enabling z/OS technologies

Deployment requirement z/OS technology
Accounting, logging and auditing SAF/RACF

SMF
RMF
z/OS logging
Security SAF/RACF
Cryptographic hardware
Network level security
MLS
Cost-effectiveness TCO
Manageability/manpower
Scalability
Integrated management solutions
Charge model
zIIP, zAAP, IFL specialty processors

Deployment requirement z/OS technology
Integrity Transaction processing

RRS, 2-PC
Parallel Sysplex, GDPS
Recoverability, disaster recovery
Predictable and consistent performance WLM

IRD
Parallel Sysplex
Hipersockets
Standards-based integration platform All J2EE products available on z/OS,

including WebSphere Application Server,
ESB, Process Server, Portal Server
Centrally managed platform z/OS

System z
Parallel Sysplex
Mature management products
Reliable operating environment System z hardware

z/OS stability
Parallel Sysplex, GDPS

4
Chapter 4. BASE24-es technical

architecture on z/OS
The combination of BASE24-es and System z is a competitive and attractive
end-to-end retail payments solution for the finance sector. BASE24-es on the
System z demonstrates the value of the ACI/IBM partnership, leveraging
best-of-breed capabilities from each company.
In this chapter, we cover the following topics:

򐂰 BASE24-es physical and logical architecture
򐂰 Workload management mechanisms
򐂰 Scalability considerations

4.1 BASE24-es logical architecture
BASE24-es was designed to run on many hardware/database/middleware
configurations. The bulk of the code is platform-independent. System Interface
Services provide a platform-independent API and a platform-dependent
implementation that allows the code to perform its business functions on a
defined list of platforms that provide the prerequisite services.
Figure 4-1 shows the BASE24-es logical architecture:
External Endpoints
Automated Point-of-Sale Host Public Card

Teller Devices Acquiring or Networks
Machines Authorization
Protocol &
Middleware Data Communication Stack
BASE24-es Communication Handlers
Messaging Middleware
Memory Based Queues
Application
BASE24-es BASE24-es BASE24-es
Integrated Server Integrated Server Integrated Server

System Interface System Interface System Interface
Services Services Services
Database
Transaction Manager
Database
Figure 4-1 BASE24-es logical architecture
4.1.1 Memory-based message queuing

BASE24-es requires a message queuing subsystem. While it is capable of
working with recoverable message queues, it does not require them. The
possibility of message loss in the data communications network requires the use
of application-level end-to-end protocols that provide recovery mechanisms for
lost messages. Those end-to-end protocols also make it possible to use more

efficient memory-based message queuing internally within BASE24-es; in the
unlikely event that messages on a memory-based queue are lost, end-to-end
protocols assure financial integrity.
IBM System z platforms offer at least two subsystems that provide high
performance memory-based message queuing that meets the requirements of
BASE24-es.
򐂰 WebSphere MQ is an industry-leading stand-alone message queuing product
that meets the requirements of BASE24-es.
򐂰 CICS Transient Data Queues meet the requirements of BASE24-es.
4.1.2 Indexed structured file system or relational database

BASE24-es requires indexed access to structured data, along with record locking
and the other features commonly associated with a structured access method.
VSAM with Record Level Sharing (RLS) meets the requirements of BASE24-es
for an efficient, scalable and reliable file system providing shared record-level
access to multiple concurrent instances of the BASE24-es Integrated Server. In
addition to enhanced performance, scalability, and reliability, the CICS API to
VSAM/RLS allows access to additional functionality relative to traditional
CICS/FOR file sharing.
DB2 support is planned for a future release. Contact ACI for details.
4.1.3 Transaction manager

BASE24-es requires a transaction manager that can provide transactional
integrity and rollback capability.
On IBM System z, CICS provides an industry-leading transaction manager.
4.1.4 Data communications

At a minimum, BASE24-es requires a sockets-based TCP/IP communications
stack.
IBM System z provides a fully functional sockets implementation; furthermore,

the Virtual Telecommunications Access Method (VTAM) also provides robust
access to other communications methods such as 3270 Bisync and SNA to
CICS-based applications.
Chapter 4. BASE24-es technical architecture on z/OS 43

4.1.5 Other requirements
In addition to the functional requirements defined here, BASE24-es has inherent
non-functional requirements for an environment that meets requisite standards of
scalability, manageability, and reliability. IBM System z hardware with the z/OS
Operating System, sysplex, and CICSplex System Manager, provides a level of
manageability, reliability, and scalability that is sufficient to meet BASE24-es
stringent standards for providing a highly available Online Transaction
Processing (OLTP) system. For further information, refer to 3.1, “How z/OS
addresses non-functional requirements” on page 24.
4.2 BASE24-es physical architecture on System z

Because of its industry-leading position as an OLTP processing environment,
and the access to other required services that it can provide, CICS was selected
as the primary runtime-environment for BASE24-es on z/OS.
CICS sockets provide efficient TCP/IP communications, while the CICS VTAM
API provides access to existing communications protocols. CICS file access in
conjunction with SMSVSAM provides efficient and scalable record-level shared
file access, while CICS itself is a world-class transaction manager. Although
WebSphere MQ is a supported alternative, CICS itself provides entirely
adequate message queuing and delivery capabilities. CICS Multiple Region
Operations (MRO) and sysplex provide scalability and help provide availability.
4.2.1 Long-running tasks

Many BASE24-es CICS components, including many of the most heavily-used
components, are implemented as long-running CICS tasks.
Most of the external endpoints with which BASE24-es communicates via TCP/IP
use long-lived sockets, and require a long-running task to maintain the socket. In
addition, the BASE24-es architecture is based on large C++ executables with
substantial initialization cost. It is far more efficient to process many units of work
in a single task instance. Accordingly, BASE24-es makes use of long-lived IP
handler and Integrated Server tasks.
The Integrated Server is a single executable program that consists of C++

components responsible for the various elements of transaction processing,
including interfaces to acquiring endpoints, issuing endpoints, cryptographic
facilities, routing and authorization. The integrated facilities minimize the
overhead of invoking many separate small CICS programs.

As of BASE24-es Version 06.2, interfaces to all supported acquiring endpoints
except ATMs are linked into the Integrated Server program. And as of
BASE24-es Version 06.4, ATM interfaces will be included as well.
Multiple instances of the long-lived Integrated Server task run in a single CICS
region to achieve the parallelism normally achieved with many short-lived CICS
tasks running in a single region.
4.2.2 Non-recoverable TDQs

As noted in 4.1.1, “Memory-based message queuing” on page 42,the end-to-end
application-level protocols used by all remote endpoints make it generally
unnecessary for BASE24-es to make use of persistent queuing, enabling it to
take advantage of the far better performance characteristic of memory-based
queuing. In general the use of recoverable TDQs in the CICS implementation of
BASE24-es results in unacceptable performance degradation, and all TDQs are
assumed to be non-recoverable unless otherwise noted.
4.2.3 Context-free servers

BASE24-es Integrated Servers (IS) are inherently context-free, and thus have
zero CICS affinities. Any message directed to a server class can be processed
by any instance of that server class in any CICS region or any logical partition
(LPAR) as long as all instances of the server class have access to a common file
system or database.
BASE24-es TCP/IP communications handlers hold context in the form of the

socket established with the remote endpoint. The BASE24-es Message Delivery
subsystem (part of System Interface Services) ensures that messages are
delivered to the correct instance of the communications handler, in whatever
region or LPAR it is located.
4.2.4 Message routing

BASE24-es applications in most cases will send messages asynchronously; that
is, they send a message and then proceed with other work and process the
response to that message if and when it is received, rather than specifically
waiting for a response.
In some cases, BASE24-es applications may send certain messages

synchronously; for example, when a request is sent to a low-latency and
generally reliable server, the applications may eliminate the overhead of
asynchronous processing and wait for a response before proceeding.

BASE24-es routing is configured by symbolic name. System Interface Services
configuration files map the symbolic name exposed to the platform-independent
application code to platform-dependent constructs such as TDQs, CICS
transactions, CICS programs, and ACI proprietary constructs built on CICS
primitives for enhanced synchronous and asynchronous message delivery.
4.2.5 Single-region deployment

At its very simplest, BASE24-es can theoretically be deployed in a single CICS
region. Files and VTAM terminals (if any) are all defined in the same region. A
single-region deployment is undesirable for reasons of scalability, availability,
performance, and system maintenance. Figure 4-2 illustrates the flow.
CICS Region
BASE24-es IP BASE24-es
Handler IPTDQ IntegratedServer
IPTDQ
BASE24-es BASE24-es
VTAM IntegratedServer
Reader/Writer
Figure 4-2 BASE24-es one-tier deployment
A communications handler (IP or VTAM) receives a message from an external

endpoint and places the message on an Integrated Server TDQ read by multiple
instances of the IS server class. One instance of the IS server class reads the
message from the TDQ and processes it. Based on configuration data and
context contained in the message, the IS task routes the response to the
component that originated the request.
If the originator was an IP handler, the message is placed on a dedicated TDQ. If

the originator was a VTAM terminal-based transaction, a VTAM terminal-based
transaction is started on the originating terminal.

4.2.6 Two-tier deployment
A two-tier deployment is generally preferable from the standpoint of scalability
and performance, and provides some incremental advantages over a single
region in the areas of availability and maintenance. Figure 4-3 illustrates the
process.
TOR
BASE24-es IP IP
Handler TDQ
AOR AOR
XDYR XDYR
IS IS
Trans - Trans -
TDQ action TDQ action
IS Task IS Task
IS Task IS Task
Figure 4-3 BASE24-es two-tier deployment
In a two-tier deployment the BASE24-es Communications Handlers (IP and

VTAM) are located in one or more separate routing regions, along with
associated file and queue definitions. The routing regions will also typically
contain any VTAM terminal definitions. By convention, even though these routing
regions may not contain VTAM terminal definitions, and do contain code such as
the BASE24-es Communications Handlers, they are referred to as Terminal
Owning Regions (TORs).
The target region typically contains definitions for the BASE24-es Integrated
Server and associated queues and file definitions, along with associated
ancillary tasks. By convention these target regions are referred to as Application
Owning Regions (AORs).
In this deployment, the BASE24-es communications handler in the TOR takes

advantage of CICS Multi-region Operation (MRO) facilities to distribute work

across multiple AORs. Since TDQ writes are not dynamically routed, BASE24-es
uses a dynamic routing transaction XDYR. The communications handler starts
XDYR, and the start is dynamically routed to an available AOR. Once running in
a AOR, XDYR writes to a TDQ in that AOR, where the message is subsequently
read and processed by an available member of the Integrated Server class.
4.2.7 Three-tier deployment

Note that a traditional CICS three-tier deployment, where the lowest tier is a
CICS File-Owning Region (FOR), is not an option for BASE24-es. The role of the
traditional FOR is better served by the facilities of VSAM Record Level Sharing.
BASE24-es relies on certain API features available only for RLS files, and will not
function in an FOR environment.
4.2.8 Workload management

Asynchronous messaging from the TOR to the AOR combines BASE24-es
long-lived Integrated Server tasks to make both CICSPlex® Workload
Management (WLM) goal-mode and CICSplex WLM queue mode inappropriate
for routing BASE24-es transactions.
BASE24-es provides both a Dynamic Transaction Routing Program for use in a

CICSplex environment and a routing user exit for use in a non-CICSplex MRO
environment, both of which serve to evenly distribute work across the available
BASE24-es AORs.
When an AOR initializes, BASE24-es runs a handshake transaction in the AOR,

which starts the BASE24-es Integrated Servers in the AOR and then notifies all
configured TORs that the AOR is available to accept workload. The handshake
command should also be run when shutting down the region to notify the TORs
that the AOR will no longer be available to accept new work.
Figure 4-4 on page 49 shows BASE24-es WLM in a CICSplex environment.

TOR
DTR6002
BASE24-es EYU9XLOP CICS CPSM BASE24-es
TCP/IP Dynamic Tran Workload CPSM Region
Handler Routing Manager Selection
(EYU9WRAM)
Start XDYR Target AOR
CICS Relay DRT6002 BASE24-es

Program MRO Region DTRCTL TSQ DTR6001
Selection Handshake
Register AOR
Routed Start of DTRCTL Register

XDYR Dynamic AOR
Transaction
Routing
Control
Database
AOR
IS TOR Scope
DRT6005 XDYR IS TDQ

Message Delivery DTR6001
Transaction IS Handshake
Figure 4-4 BASE24-es WLM in a CICSPlex environment
Figure 4-5 on page 50 shows BASE24-es WLM in a non-CICSPlex MRO

environment.

TOR
DTR6002
BASE24-es EYU9XLOP CICS CPSM BASE24-es
TCP/IP Dynamic Tran Workload CPSM Region
Handler Routing Manager Selection
(EYU9WRAM)
Start XDYR Target AOR
CICS Relay DRT6002 BASE24-es

Program MRO Region DTRCTL TSQ DTR6001
Selection Handshake
Register AOR
Routed Start of DTRCTL Register

XDYR Dynamic AOR
Transaction
Routing
Control
Database
AOR
IS TOR Scope
DRT6005 XDYR IS TDQ

Message Delivery DTR6001
Transaction IS Handshake
Figure 4-5 BASE24-es workload management - non-CPSM
4.2.9 TOR architecture

A BASE24-es TOR is a Terminal Owning Region by convention only, as noted in
4.2.6, “Two-tier deployment” on page 47. It will commonly not contain any VTAM
terminal definitions, and will always contain user code. It will also contain file
definitions for the BASE24-es DDMF and SDMF routing files, and normally
contain file definitions for the TCPIPCFG and SOCKRECS TCP/IP
Communications Handler configuration files.
All communications handlers that are in the TOR route messages to the AOR
using symbolic routing (see 4.2.4, “Message routing” on page 45), and can route
messages using most of the available routing methods. Typically routing will
result in a distributed start of the XDYR BASE24-es Message Delivery program
as previously described (see 4.2.8, “Workload management” on page 48).
The BASE24-es communications handler code running in the TORs typically

consumes relatively little CPU resource.

TCP/IP client considerations
When discussing TCP/IP communications, it is necessary to make a distinction
between the TCP/IP client and the application-level client. For example, most
public card networks prefer that the member institution act as the TCP/IP client,
even though the institution may be solely a card issuer processing requests sent
from the network.
The BASE24-es TCP/IP client for CICS can be configured to establish

connections with multiple remote IP addresses and ports. It can also be
configured to bind to a specified local port. It associates each remote endpoint
with a specific BASE24-es symbolic name (see 4.2.4, “Message routing” on
page 45) for purposes of internal routing.
Each remote endpoint is associated with a specified symbolic name, which is in

turn mapped to a specific CICS destination type and identifier by System
Interface Services (SIS) configuration files. All messages from a remote endpoint
are delivered to the specified local destination.
See 4.2.6, “Two-tier deployment” on page 47 for an illustration of communications

between the TCP/IP client and the BASE24-es authorization components.
Figure 4-6 on page 52 illustrates the components of the TOR associated with the
TCP/IP client.

TOR
SOCKRECS
BASE24-es
IP Client
TCPIPCFG
SDMF
BASE24-es
Task
Monitor DDMF
Start XDYR
Figure 4-6 BASE24-es TOR – IP client
The components illustrated are as follows:

򐂰 SOCKRECS has one record for each configured remote endpoint.
򐂰 TCPIPCFG has information about the IP Client handler process itself,
specifying information such as the symbolic name of the IP handler, and
whether it is client or server.
򐂰 DDMF has routing information for each symbolic destination in the system.
򐂰 SDMF has configuration information for each BASE24-es task that makes use
of SIS Message Delivery services.
򐂰 The BASE24-es Task Monitor runs at intervals to monitor the health if the IP
Client and restart it as required.
TCP/IP server considerations

For some specific purposes (see 4.2.13, “User interface region architecture” on
page 90), BASE24-es makes use of the IBM-provided CSKL listener. However,
for OLTP it provides its own listener, the BASE24-es TCP/IP server. Unlike the
IBM listener, the BASE24-es server retains ownership of the sockets on which it
accepts connections. It is architecturally similar to the BASE24-es TCP/IP client
in that respect.

The BASE24-es TCP/IP server for CICS binds to and accepts connections on a
specific TCP port. To accept connections on multiple ports, it is necessary to
configure multiple copies of the server.
The server can be configured to associate specific remote endpoints with

specific BASE24-es symbolic names (typically in an ATM acquirer environment),
or to associate remote endpoints with its own symbolic name (typically in a POS
or dial-up ATM environment) based on configuration in the TCPIPCFG and
SOCKRECS files.
See 4.2.6, “Two-tier deployment” on page 47 for an illustration of communications

between the TCP/IP client and the BASE24-es authorization components.
Figure 4-7 illustrates the components of the TOR associated with the TCP/IP
server.
TOR
SOCKRECS
BASE24-es
IP Client
TCPIPCFG
SDMF
BASE24-es
Task
Monitor DDMF
Start XDYR
Figure 4-7 BASE24-es TOR - TCP/IP server
򐂰 SOCKRECS has one record for each configured remote endpoint that the IP
Server will associate with a unique symbolic name. Remote endpoints not
configured in SOCKRECS will take on the symbolic name of the TCP/IP
Server itself.

򐂰 TCPIPCFG has information about the IP Server process itself, specifying
information such as the symbolic name of the IP handler, and whether it is
client or server.
򐂰 DDMF has routing information for each symbolic destination in the system.
򐂰 SDMF has configuration information for each BASE24-es task that makes use
of SIS Message Delivery services.
򐂰 The BASE24-es Task Monitor runs at intervals to monitor the health of the IP
Server and restart it as required.
VTAM considerations
BASE24-es makes use of CICS/VTAM facilities to support non-IP based
protocols on z/OS. VTAM reader/writer programs are developed as the need
arises. Current offerings include SNA LU.0 and SNA LU.2.
Figure 4-8 illustrates a BASE24-es TOR – VTAM Reader/Writer configuration.
VTAM
Terminal
1 6
TOR AOR
2 XDYR
VTAM
Reader/Writer IS
TDQ
3
4
5 Integrated
Server
Figure 4-8 BASE24-es TOR – VTAM Reader/Writer
1. An incoming message causes the VTAM reader/writer transaction to be

started on the originating terminal. The reader/writer calls RECEIVE to get
the message.
2. The reader/writer adds a BASE24-es internal header to the message,
specifying among other things the symbolic source associated with its
terminal. It starts the AOR Message Delivery transaction XDYR, and
terminates. The start is routed to an available AOR.

3. XDYR places the incoming message on the Integrated Server TDQ in the
region where it was started.
4. The BASE24-es Integrated Server reads the message and processes it.
5. The Integrated server reads the DDMF routing configuration file to get routing
information for the symbolic source of the request, and starts the VTAM
reader/writer on the associated terminal.
6. The VTAM reader/writer calls SEND to deliver the message to the originating
terminal, and terminates.
WebSphere MQ considerations
In some cases, it may make sense for an external endpoint to send requests to
BASE24-es using WebSphere MQ. This form of message delivery has no
components in the TOR, but is handled directly in the AOR (see “Operator
notification considerations” on page 55).
Operator notification considerations

For detailed information about this topic, refer to the discussion under AOR
Operator Notification considerations (“Operator notification considerations” on
page 58).
TOR file definitions

The BASE24-es TOR contains definitions for System Interface Service (SIS)
Message Delivery and Event files, TOR-specific communications configuration
files, and BASE24-es workload management files. Table 4-1 lists the files that
may be defined in a BASE24-es TOR.
Table 4-1 TOR file definitions

Dynamic Destination Map File (DDMF) This SIS routing configuration file maps BASE24-es symbolic
names to CICS facilities. It contains one record per routable
endpoint in the BASE24-es system that supports
asynchronous requests. It is required in the TOR.
Static Destination Map File (SDMF) This SIS routing configuration file contains static information
about an endpoint. It contains one record per CICS transaction
in the BASE24-es system, plus one record per SOCKRECS file
entry.
Synchronous Destination Map File This SIS routing configuration file maps BASE24-es symbolic
(SYDMF) names to CICS facilities. It contains one record per routable
endpoint in the BASE24-es system that supports synchronous
requests. It is not generally required in the TOR.
Terminal Configuration File This BASE24-es communications configuration file contains

(TERMCFG) one record per VTAM reader/writer CICS transaction.

TCP/IP Configuration File (TCPIPCFG) This BASE24-es communications configuration file contains
one record per BASE24-es TCI/IP communications handler
task.
Socket Records file (SOCKRECS) This BASE24-es communications configuration file maps
remote endpoints to BASE24-es symbolic names. It contains
one record per remote endpoint defined to the system.
Endpoints not defined to the system are assigned a generic
symbolic name.
Event Log (EVTLOGR) This SIS event file is a Relative Record Data Set that is a
circular buffer of Operator Notification events.
DTR Control file (DTRCTL) This file is used by the BASE24-es Workload Management
User Exit as a source of information for the dynamically created
Applications Table and Routing Tables. The records in this file
are used to create tables in the DTRCTLTS CICS TS queue.
4.2.10 AOR architecture

The BASE24-es AOR contains file definitions for the VSAM/RLS files in the
authorization database, as well as definitions for the SDMF, DDMF, and SYDMF
routing files used by the BASE24-es System Interface Services Message
Delivery Service. It also contains definitions for the BASE24-es Integrated Server
and ancillary programs and transactions such as the BASE24-es Task Monitor
and AOR Message Delivery programs.
Long-running task considerations

BASE24-es uses many long-running CICS tasks as described in 4.2.1,
“Long-running tasks” on page 44 including the Integrated Server in the AOR.
Figure 4-9 on page 57 depicts the components associated with the long-running
Integrated Server tasks in the AOR.

AO1R
Timer ISnn (BASE24-es

XDYR Integrated Server)
(BASE24-es AOR
Message Delivery
Program)
ISnn TDQ
ISnn (BASE24-es
Integrated Server)
Timer
AOR1 ISnn
MDS Memory
Table
BASE24-es
Task Monitor
Figure 4-9 BASE24-es AOR - Long-running tasks
The BASE24-es Integrated Server tasks are started by the BASE24-es Task
Monitor program during CICS region initialization. They in turn create a
configurable number of instances of one or more BASE24-es Integrated Server
classes. (It may be desirable to create more than one Integrated Server class to
handle specific workloads. A separate TDQ is configured for each server class.)
When the Integrated Servers initialize, they enter themselves in a shared

memory table associated with the server class in the AOR. The Task Monitor is
then scheduled to rerun at configurable intervals to monitor the state of the
servers it has started, and restart them as required.
When the Integrated Server long-running tasks find no messages on their TDQ,
they post a timer and wait. When the AOR Message Deliver Program runs, it
writes the message to the TDQ associated with the correct server class. It then
examines the shared memory table and checks the state of class member tasks.
If it finds a task that is waiting on a timer, it cancels the timer so the task can
wake and process a message from the TDQ.
It is possible that another task has already processed the delivered message. If
so, the newly-awakened task simply re-posts its timer and waits; the associated
cost is low.

Since the shared memory table is specifically designed to be associated with the
tasks running in a single AOR, it does not introduce any undesirable CICS
affinities.
Each group of long-running tasks associated with a single queue is commonly

referred to as a BASE24-es server class.
Operator notification considerations

From time to time, BASE24-es may encounter conditions that it wishes to bring to
the attention of a human or programmatic operator.
Figure 4-10 illustrates how those operator notifications are generated.
BASE24-es region
BASE24-es
Application BASE24-es
EVTLOGR
TDQ
CICS
EINF Spooler
BASE24-es Event (CEEOUT)
Distributor EWRN
Transaction
EEER
ECRT
BASE24-es R300 Event

Perusal Transaction
CICS
BASE24-es EVTLOGR (VSAM RLS Operator
Shared Relative Record Data Set)
Figure 4-10 BASE24-es Operator Notifications
When an operator notification is generated, the generating task writes it to the

Event Log TDQ. The Event Log TDQ then triggers the BASE24-es Event
Distributor transaction.

The Event Distributor transaction first writes the event to a circular Relative
Record data set, which is intended to give a system-wide picture of associated
events. You can peruse this file with the BASE24-es R300 Event Log Perusal
transaction.
Then, based on the severity of the notification, the Event Distributor starts one of
four Alternate Distributor transactions: EINF (Informational Notification), EWRN
(Warning Notification), EERR (Error Notification) or ECRT (Critical Notification).
By default, these four transactions are associated with the same program,
SIADIST, which simply writes to the local CICS spooler. However, the real
flexibility of the notification architecture lies in the fact that the program
associated with these transactions is intended to be user-replaceable. The user
might, for example, associate EERR and ECRT with a program that also writes to
the z/OS console.
Tivoli Monitoring considerations

Currently there is no API available to CICS applications to let them raise events
to the Tivoli Enterprise™ Console. CPSM has an API into Tivoli, and CPSM can
be used to send messages or alerts to NetView®, and if NetView is set up, it can
route to the Tivoli Event Console. Though it is not part of BASE24-es standard
product, an Alternate Distributor (described in “Operator notification
considerations” on page 55) could be developed to make use of these facilities.
WebSphere MQ considerations
A BASE24-es Integrated Server class or other BASE24-es Message Delivery
Service-based task can be configured to read a WebSphere MQ queue rather
than a CICS TDQ.
Figure 4-11 on page 60 shows an Integrated Server class configured to read

messages with no BASE24-es internal header, rather than the more typical
configuration in which the Integrated Server expects messages with a
BASE24-es internal header from a CICS TDQ. In this configuration, the
BASE24-es System Interface Services Message Delivery Service assigns a
symbolic name to the remote endpoint, much as a BASE24-es Communications
Handler in the TOR normally would, and creates an internal BASE24-es header.
The Integrated Server processes the request and routes the response to the
originating endpoint by sending the response to symbolic name assigned by the
Message Delivery Service.
The Message Delivery Service determines that symbolic name is associated with
a WebSphere MQ, and that messages to that destination should be placed on
the queue without a BASE24-es internal header.

AOR
IS (MQ, No
IS Internal Header)
MQ
External
Endpoint MQ
Figure 4-11 WebSphere MQ external endpoints
The attributes of a destination may be configured independently; that is, a

symbolic destination may be configured as a WebSphere MQ that takes
messages with BASE24-es internal headers, or a CICS TDQ that takes
messages without BASE24-es internal headers. However, a server class must be
configured with both attributes. That is, a single server class can accept
messages from:
򐂰 A CICS TDQ that contains only messages with BASE24-es internal headers
򐂰 A CICS TDQ that contains only messages without BASE24-es internal
headers
򐂰 A WebSphere MQ that contains only messages with BASE24-es internal
headers
򐂰 A WebSphere MQ that contains only messages without BASE24-es internal
headers
Supporting all the options above would require configuring four Integrated Server
classes; that is, four CICS transactions, each with one of the preceding
configurations, each with a specified number of instances of long-lived tasks, and
each associated with the BASE24-es Integrated Server program.
See “Data communications to external authorization system” on page 65 for a

more complex example involving both MQ and TDQ-enabled Integrated Server
classes.

Cryptographic considerations
BASE24-es requires access to a secure cryptographic facility for cardholder PIN
translation and verification, Message Authentication Code generation and
verification, and other applications. There are several options available in the
z/OS implementation of BASE24-es.
TCP/IP connected security module

BASE24-es provides a special synchronous implementation of its TCP/IP Client
communications handler for use with TCP/IP-attached Hardware Security
Modules (HSMs). It is usually desirable to have at least two HSMs available on
the system. BASE24-es will distribute the load across available configured
security modules.
Figure 4-12 provides a simplified view illustrating connectivity to a single

synchronous TCP/IP client in a single TOR.
TOR
HSM
Synchronous
IP Client
IP Client
TDQ
AOR
Integrated
TSQ Server
Record
1
Record
2
Figure 4-12 BASE24-es TCP/IP-attached HSM
In the figure, the BASE24-es Integrated Server, if necessary, creates a CICS

Temporary Storage Queue (TSQ). When it generates a request to the HSM, it
writes information about the request into the first record of the CICS TSQ, then

places the request on the CICS TDQ associated with the Synchronous IP Client,
and sets a timer.
The Synchronous IP Client reads its TDQ and sends the request to the HSM over
a long-lived socket connection. It then reads the response from the socket,
validates that the response is still related to the request specified in the first
record of the TSQ, places the response in the second record of the TSQ, and
cancels the timer set by the Integrated Server.
The Integrated Server wakes up and, finding a response in the second record of
the TSQ, processes the response from the TCP/IP-attached HSM.
VTAM attached security module

BASE24-es also supports synchronous communication with VTAM-attached
security modules; see Figure 4-13.
TOR
VTAM
Writer/Reader
HSM
AOR
TSQ Integrated
Server
Record
1
Record
2
Figure 4-13 BASE24-es VTAM-attached HSM
As shown in the figure, the BASE24-es Integrated Server, if necessary, creates a

CICS Temporary Storage Queue (TSQ). When it generates a request to the
HSM, it writes information about the request into the first record of the CICS

TSQ, then starts the configured VTAM Writer/Reader task on the specified
terminal, and sets a timer.
The VTAM Writer/Reader is passed the HSM request when it is started. It sends
the request to the HSM and waits for a response. When a response is received,
the Writer/Reader validates that the response is still related to the request
specified in the first record of the TSQ, places the response in the second record
of the TSQ, cancels the timer set by the Integrated Server, and terminates.
The Integrated Server wakes up and, finding a response in the second record of
the TSQ, processes the response from the TCP/IP-attached HSM.
Vendor-specific CICS API

Hardware Security Module vendors may provide a CICS API to facilitate
communications with their HSMs. Contact ACI for a list of supported vendors.
IBM Integrated Cryptographic Service Facility

Although BASE24-es support for the IBM Integrated Cryptographic Service
Facility is in plan, as of BASE24-es Version 06.2 it is not yet supported.
Integration with external authorization systems

While BASE24-es provides extensive authorization capabilities of its own, it is
often called upon to route messages to external authorization subsystems. On
most platforms, BASE24-es is limited to various means of data communications
to access a public card network or back-end host authorization subsystem.
However, the System z implementation of BASE24-es has other options,
especially when interacting with an existing host-based authorization system.
In general, the method to access the external authorization system is determined

by the requirements of the authorization system itself.
Local link to external authorization system

One option that is reliable and performs well is a CICS LINK to a component of
the external authorization system; Figure 4-14 on page 64 illustrates this
approach.
This may be appropriate if the component to which BASE24-es will link is itself
small, with small CPU utilization and resource requirements; perhaps it is a
simple front-end for a more extensive system located in another AOR, and is
capable of handling the associated availability and workload distribution
concerns. In this case the communications between the BASE24-es AOR and
the external authorization system AOR is determined by the external
authorization system itself.

Because the external authorization system is generally either a remote card
network authorizer or a z/OS host application, this option is only viable for the
z/OS implementation of BASE24-es, and only when the external authorizer is a
z/OS-based application.
When sizing the system it is important to remember that since a single thread of
execution in BASE24-es is blocked waiting for a response from the host, it is
necessary to configure more threads of execution, that is, more Integrated
Server long-lived tasks.
At a minimum, (peak_arrival_rate)*(is_latency +
external_authorization_system_latency) instances of the Integrated Server
long-lived task should be configured. (This number is a general guideline only;
more specific sizing information should be obtained before finalizing any plans.)
Configuring additional IS tasks is usually an acceptable trade-off since the cost of

adding additional instances of the IS task is low in terms of CPU utilization;
low-to-moderate in terms of memory utilization; and the processing cost of a
synchronous communication to the external authorization system is generally
significantly lower than the processing cost of an asynchronous communication.
TOR
BASE24-es AOR External Authorization

AOR [s]
BASE24-es Integrated
Server
CICSLINK
Linkable EAS component
Figure 4-14 Link to external authorization system
Distributed link to external authorization system

An alternative to the synchronous communications described in “Local link to
external authorization system” on page 63 is a Distributed Program Link (DPL) to
the external authorization system; Figure 4-15 on page 65 illustrates this
approach.

This approach is generally to be preferred if the component being linked to by
BASE24-es is large with large CPU utilization and resource requirements, or if
using CPSM facilities to achieve availability and workload management goals is a
consideration.
The considerations regarding the number of instances of the IS task described in

“Local link to external authorization system” on page 63 apply here as well.
Again, this is a general guideline and more specific sizing information should be
obtained.
TOR
BASE24-es AOR External Authorization System

AOR [s]
BASE24-es
Integrated Linkable EAS component
Server CICS
DISTRIBUTED
PROGRAM
LINK
Figure 4-15 Distributed link to external authorization system
Data communications to external authorization system

The use of data communications protocols to communicate with an external
authorization system is required when the external authorizer is a public card
network; Figure 4-16 on page 66 illustrates this approach.
While it is possible, and on rare occasions necessary, to use data

communications to communicate with a back-end host authorization system that
is collocated with the BASE24-es system, it is generally better to use one of the
previously defined CICS-based methods.
The BASE24-es cost associated with asynchronous communications to the

back-end system is generally higher, because of the need to save context for the
context-free server class (see 4.2.3, “Context-free servers” on page 45) in the
database and the additional overhead associated with the data communications
itself.

However, asynchronous communications may be required due to requirements
of the external authorization system, or because excessive latency in the
external authorization system makes it desirable for BASE24-es to perform this
operation asynchronously.
Acquiring Issuiing
external external
endpoint endpoint
TOR 1 12 6 7
IP Handler 1 IP Handler 2
11 5
IP 1Q IP 2Q
2
AOR
10 4
IS Q
3, 9
Integrated Server
Figure 4-16 Data communications to external authorization system
1. The acquiring endpoint sends an IP request.

2. The IP handler starts an XDYR transaction to put the request on an IS TDQ in
some AOR.
3. The IS reads the message from the queue and determines that it must be
routed to an external authorizer.
4. The IS saves the transaction context in the database and places the request
on the TDQ associated with the correct external endpoint. The TDQ may be
associated with the same or a different IP handler than the one that originated
the request, depending on where the socket associated with the external
endpoint is located.

5. The IP handler reads the request from the TDQ and locates the socket
associated with the external endpoint.
6. The IP handler sends the request to the external authorizer.
7. The external authorizer sends a response.
8. The IP handler starts a XDYR transaction to place the response on the TDQ
associated with messages from the external endpoint.
9. The IS reads the response from the TDQ.
10.The IS retrieves the context from the database, updates the database, and
routes the response to the TDQ associated with the originating endpoint.
11.The IP handler reads the message from the TDQ and locates the socket
associated with the originating external endpoint.
12.The IP handler routes the response to the originating external endpoint.
WebSphere MQ to external authorization system

BASE24-es can be configured to use WebSphere MQ to communicate with
WebSphere MQ-enabled External Authorization Systems or acquiring endpoints.
Figure 4-17 illustrates one possible configuration with asynchronous messaging
to the External Authorization System.
(This example was chosen for illustrative purposes. Asynchronous messaging is

generally to be preferred when data communications latency and reliability
issues are introduced into the interface between BASE24-es and the External
Authorization System; this may or may not be the case with WebSphere MQ.)
TOR AOR
IS01
IS01 (TDQ/Internal EAS
Header Server
TDQ Class) MQ
IS02
(MQ/No IS02
Header Server MQ
Class)
Figure 4-17 WebSphere MQ to external authorization system

1. In the flow illustrated in Figure 4-17 on page 67, the TOR places a message in
the IS01 TDQ, which is associated with an Integrated Server class that is
configured to expect messages with the BASE24-es internal header from the
IS01 TDQ.
2. The IS01 task that processes the message determines that it should be
authorized by an External Authorization System that is accessed via a
symbolic destination associated with the EAS MQ, and that messages
intended for that symbolic destination are accessed by placing a message
without a BASE24-es internal header on that queue.
3. The External Authorization System processes the request, and places the
response on the WebSphere MQ IS02 queue.
4. The IS02 Integrated Server class is configured to read messages from the
IS02 WebSphere MQ queue, and to assume that those messages do not
have a BASE24-es internal header. The BASE24-es System Interface
Services Message Delivery Service adds a header and the Integrated Server
task processes the response, impacting the financial database and
forwarding the response to the originating external endpoint.
AOR file definitions

A BASE24-es AOR contains definitions for files used by the platform-specific
infrastructure: System Interface Service (SIS) Message Delivery and Event files,
TOR-specific communications configuration files, and BASE24-es workload
management files. Table 4-1 on page 55 lists and explains the AOR file
definitions.
Table 4-2 AOR file definitions

Dynamic Destination Map File (DDMF) This SIS routing configuration file maps BASE24-es symbolic
names to CICS facilities. It contains one record per routable
endpoint in the BASE24-es system that supports
asynchronous requests.
Static Destination Map File (SDMF) This SIS routing configuration file contains static information
about an endpoint. It contains one record per CICS transaction
in the BASE24-es system, plus one record per SOCKRECS file
entry.
Synchronous Destination Map File This SIS routing configuration file maps BASE24-es symbolic
(SYDMF) names to CICS facilities. It contains one record per routable
endpoint in the BASE24-es system that supports synchronous
requests.
Event Log (EVTLOGR) This SIS event file is a Relative Record Data Set that is a
circular buffer of Operator Notification events.

DTR Control file (DTRCTL) This file is used by the BASE24-es Workload Management
User Exit as a source of information for the dynamically created
Applications Table and Routing Tables. The records in this file
are used to create tables in the DTRCTLTS CICS TS queue.
Card File (CARDD) This file contains one record per card known to the associated
financial institution.
Context File (CTXD) This file stores context for a client component. If client context
exceeds the limits of a single record, the context is stored in
multiple records. A context record has an expiration time stamp
after which a clean-up program will delete the record if the
client has not already done so. The assigns for the context
tables are defined in the Context Configuration table to allow a
number of Context tables to be shared or kept separate by
different clients.
A BASE24-es AOR also contains definitions for nearly two hundred files that are
optional or required for various options and configurations of the
platform-independent BASE24-es application logic. Most of these files are small
and static, and (if required) are read into CICS main storage at CICS region
initialization. They are never again read, and never updated by the Online
Transaction Processing (OLTP) logic. They can be updated only by operator
command and do not introduce any CICS region affinities.
The platform-independent files are listed in Table 4-3. Files which have both a
base name and an OLTP name (for example, Acquirer_Issuer_Relation and
Acquirer_Issuer_Relation_OLTP) are of this type; the file exists only as input to
build the table in main storage.
Table 4-3 Platform-independent file definitions

Acquirer_Issuer_Relation The Acquirer Issuer Relation table contains one record for
each issuer route profile, acquirer route profile, and
transaction code (that is, the first two characters of a
processing code) combination in the system.
Through configuration of AIR records, a relationship
between an acquirer and an issuer is defined by specifying
an allowed transaction code. The key to the AIRD is Issuer
Route Profile, Acquirer Route Profile, and Transaction Code.
The Issuer Route Profile and Acquirer Route Profile may be
wildcarded with asterisks.
This table defines a set of transactions that is valid for every
Issuer/Acquirer Route Profile combination. The route code is
used as a key to the Route Table.

Acquirer_Issuer_Relation_OLTP The Acquirer Issuer Relation OLTP table contains one
record for each issuer route profile, acquirer route profile,
and transaction code (that is, the first two characters of a
processing code) combination in the system. Through
configuration of AIR records, a relationship between an
acquirer and an issuer is defined by specifying an allowed
transaction code.
The key to the AIRD is Issuer Route Profile, Acquirer Route
Profile, and Transaction Code. The Issuer Route Profile and
Acquirer Route Profile may be wildcarded with asterisks.
This table defines a set of transactions that is valid for every
Issuer/Acquirer Route Profile combination. The route code is
used as a key to the Route Table.
This table is populated from the Acquirer Issuer Relation
table, and is accessed as a read only table during online
transaction processing.
Acquirer_Route_Profile The Acquirer Route Profile table contains the description

associated with each acquirer route profile defined in the
system.
Acquirer_Txn_Allowed The Acquirer Transaction Allowed table contains one row for
each acquirer transaction profile, message category code
and processing code in the system that are allowed.
This table defines the set of transactions that are valid for an
acquirer.
Acquirer_Txn_Allowed_OLTP The Acquirer Transaction Allowed OLTP table contains one

row for each acquirer transaction profile, message category
code and processing code in the system that are allowed.
This table defines the set of transactions that are valid for an
acquirer. This table is populated from the Acquirer Txn
Allowed table, and is accessed as a read only table during
online transaction processing.
Action_Code_Extrn_to_Intrn The Action Code External to Internal Table contains one

record for every external action code defined for the action
code profile. All external action codes for the message
profile must be the same number of bytes.
Action_Code_Extrn_to_Intrn_OLTP The Action Code External to Internal OLTP table contains

one record for every external action code defined for the
action code profile. All external action codes for the message
This table is built from the Action Code External to Internal
table.

Action_Code_Intrn_to_Extrn The Action Code External to Internal Table contains one
record for every external action code defined for the action
code profile. All external action codes for the message
Action_Code_Intrn_to_Extrn_OLTP The Action Code External to Internal OLTP Table contains

one record for every external action code defined for the
action code profile. All external action codes for the message
This table is built from the Action Code Internal to External
table.
Active_Script_Statistics The Active Script Statistics Table contains one record for
each authorizer and message type which is being
monitored. The records contain the counts for the script and
whether the script is enabled or not. Once the number of
transactions processed with the script exceeds the minimum
transaction count, the percentage of approvals, denials and
referrals are compared to their corresponding threshold
limits to determine whether the script should continue to be
used or not. If not, the script is disabled. This table is
updated during online processing for scripts configured to be
monitored.
This table should not be audited.
Admin_Card Admin Card Data Source
Audit_Store_and_Forward The Audit Store and Forward (SAF) table contains all the
records which are stored to be forwarded to the EAE at a
later time.
Audit_Store_and_Forward_Config This table contains SAF Configuration information for

Auditing UI functions. It contains SAF limits and time-out
values. There should only be one record in this table.
Authorization_Script The Authorization Script Table contains one record for each
Authorization Script defined in the Script Configuration data
source. It defines whether the script is enabled or not and
whether to monitor the script to ensure thresholds are met
for the script.
Authorization_Script_OLTP The Authorization Script OLTP Table contains one record for
each Authorization Script defined in the Script Configuration
data source. It defines whether the script is enabled or not
and whether to monitor the script to ensure thresholds are
met for the script.
This table is populated from the Authorization Script table.
Banknet_Configuration The Banknet Configuration Table contains a record with

additional configuration required for each Banknet Interface.

Banknet_Configuration_OLTP The Banknet Configuration OLTP Table contains a record
with additional configuration required for each Banknet
Interface. The Bankenet Configuration OLTP Table is loaded
from the Banknet Configuration Table.
Banknet_Group_Timers The Banknet Group Timers Table contains a record for each
logon sent to keep track of duplicate MCI ids, and to handle
logon time-outs and responses.
Banknet_Institution_Id The Banknet Institution Identification Table contains an entry

for every institution identifier defined in the system which is
mapped to an MCI identification number.
Banknet_Institution_Id_OLTP The Banknet Institution Identification OLTP Table contains

an entry for every institution identifier defined in the system
which is mapped to an MCI identification number.
Banknet_Merchant_Program The Banknet Merchant Program Table contains a record for

each merchant program that a merchant may belong to
which maps to the MasterCard Promotion Code.
Banknet_Merchant_Program_OLTP The Banknet Merchant Program OLTP Table contains a

record for each merchant program that a merchant may
belong to which maps to the MasterCard Promotion Code.
This table is loaded from the Banknet Merchant Program
Table.
Card The Card Table. Contains one record for each card in the
network.
Card_Account The Card Account Table. Contains information for account

numbers belonging to the specified card.
Card_Verification Card Verification contains information required for the

secure management of cards in the system.
Card_Verification_OLTP Card Verification contains information required for the

secure management of cards in the system.
Chan_Profile_Instrm_Typ_Rel_OLTP The Channel Profile Instrument Type Relation OLTP table

contains one record for each Channel Profile and instrument
type combination allowed by the channel.
This table is populated from the Channel Profile Instrument
Type Relation table.
Chan_Profile_Instrm_Typ_Relation The Channel Profile Instrument Type Relation table contains

one record for each Channel Profile and instrument type
combination allowed by the channel.

Channel_Instrm_Type_Relation The Channel Instrument Type Relation table contains one
record for each instrument type that a channel id supports
that is not already specified in the Merchant Delivery
Channel table. The Debit, Credit and Adjustment counts and
amounts are maintained for each instrument type.
Channel_Public_Key_Security Contains the public key information of the Encryption PIN

Pad (EPP) of the channel device.
Channel_Security Channel Security Data Source
Compiler_Message The Compiler Message table contains one record for each
message which is generated by the script compiler for errors
and warnings. The record contains the severity of the
message, the message text and message detail.
Context This Context table stores context for a client component. If

client context exceeds the limits of a single record, the
context is stored in multiple records.
A context record has an expiration time stamp after which a
clean-up program will delete the record if the client has not
already done so. The assigns for the context tables are
defined in the Context Configuration table to allow a number
of Context tables to be shared or kept separate by different
clients.
Context_Configuration The Context Configuration table contains one record for

each context group that is defined in the system. The record
defines the assign name to use for a Context table. The
assign name represents a logical data source which allows
multiple Context tables to be defined based on the clients of
Context Component.
Context_Configuration_OLTP The Context Configuration OLTP table contains one record

for each context group that is defined in the system. The
record defines the assign name to use for a Context table.
The assign name represents a logical data source which
allows multiple Context tables to be defined based on the
clients of Context Component.
This table is populated from the Context Configuration table,
and is accessed as a read only table during online
Contingency_Interface The Contingency Interface configuration table contains a

record for each contingency interface defined in the system.
Contingency_Interface_OLTP The Contingency Interface OLTP configuration table

contains a record for each contingency interface defined in
the system. The Contingency Interface OLTP table is loaded
from the Contingency Interface table.

Currency_Conversion_Rate The Currency Conversion Rate data source contains the
currency rate to be used to convert an amount from a source
currency to a base currency
Currency_Conversion_Rate_OLTP The Currency Conversion Rate data source contains the

currency rate to be used to convert an amount from a source
currency to a base currency.
Diebold_Number_Table The Diebold Number Table contains information specific to

the Diebold Number Table (DNT) PIN verification method
Diebold_Number_Table_OLTP The Diebold Number Table contains information specific to

the Diebold Number Table (DNT) PIN verification method.
EMV_Script_Context This data source contains information pertinent for EMV

scripts.
EMV_Security The EMV Security Data. It contains the keys required for
EMV security processing and the scheme used.
EMV_Security_OLTP The EMV Security Data. It contains the keys required for
EMV security processing and the scheme used.
Environment The Environment table contains one record for each attribute
name/value pair in the system which overrides the default
value of the attribute defined by the application. Any attribute
not defined in the Environment table will use an appropriate
default value as defined by an application.
EPP_Serial_Number The EPP Serial Number data source is used to verify that the
EPP serial number received in a message from a device is a
valid/known serial number.
Euro_Opted_Currency The Euro Opted Currency data source is used to determine

if a currency is Euro opted. When a EU country decides to
stop using its local currency and begins using the Euro
(“opts in”), there is a transitional period in which the local
currency becomes “fixed” to the Euro.
Euro_Opted_Currency_OLTP The Euro Opted Currency data source is used to determine

if a currency is Euro opted. When a EU country decides to
stop using its local currency and begins using the Euro(“opts
in”), there is a transitional period in which the local currency
becomes “fixed” to the Euro.
Event The Event table contains one row for each event which may
be logged. Each event contains up to 20 tokens which can
be substituted into the message text. Event suppression is
also defined in the Event table.

Event_Document The Event Document table contains documentation for event
messages regarding the sub-system information for each
SSID group, the probable cause of the event, action taken by
the process when the event happened, the remedy for the
situation that caused the event, and token descriptions for
the event.
Exception_Log The Exception Log table contains one record for each
exception condition encountered by an endpoint which
requires the external message to be logged.
Exported_Operator_Documentation This table contains information about the registered

exported operators on a system. The operators’ description,
prerequisites, return values and parameter values are
configured in this table
Fees The Fees Table contains the information needed to calculate

a fee for a transaction for a specified fee profile and fee
name.
Fees_OLTP The Fees Table contains the information needed to calculate

a fee for a transaction for a specified fee profile and fee
name.
Holiday The Holiday table contains a record for each holiday date
specified as a holiday belonging to a holiday profile.
Holiday_OLTP The Holiday OLTP table contains a record for each holiday
date specified as a holiday belonging to a holiday profile.
This table is loaded from the Holiday table.
Host_Public_Key_Security_Primary This contains the primary public key and secret key
information generated for the Host.
Host_Public_Key_Security_Second This contains the secondary public key and secret key
information generated for the Host.
IBM_DES IBM DES contains information specific to the DES

(IBM 3624) PIN verification method.
IBM_DES_OLTP IBM DES contains information specific to the DES

(IBM 3624) PIN verification method.
Identikey The Identikey data source contains configuration

parameters required for the Identikey PIN verification
method.
Identikey_OLTP The Identikey data source contains configuration

parameters required for the Identikey PIN verification
method.

IMT_ATM_Tran_Code_Proc_Code The IMT ATM Transaction Code to Processing Code table
contains an entry for every BASE24 ATM transaction code
which may be received. The ATM transaction code is
mapped to a corresponding processing code.
IMT_ATM_Tran_Code_Proc_Code_OLTP The IMT ATM Transaction Code to Processing Code OLTP

table contains an entry for every BASE24 ATM transaction
code which may be received. The ATM transaction code is
This table is loaded from the ATM Transaction Code to
Processing Code table.
IMT_Interface This is the configuration for the IMT Interface.
IMT_Interface_OLTP This is the configuration for the IMT Interface.

This table is loaded from the IMT Interface Table.
IMT_POS_Tran_Code_Proc_Code The IMT POS Transaction Code to Processing Code table

contains an entry for every BASE24 POS transaction code
which may be received. The POS transaction code is
IMT_POS_Tran_Code_Proc_Code_OLTP The IMT POS Transaction Code to Processing Code table

contains an entry for every BASE24 POS transaction code
which may be received. The POS transaction code is
This table is loaded from the IMT POS Transaction Code to
Processing Code table.
Inbound_Station_OLTP The Inbound Station table contains one record for each
station an inbound message may be received on. The
outbound station is its station pair that a corresponding
message will be sent out on.
The inbound and outbound station names may be the same
value. The interface they belong to is identified.
This table is populated from the Interface Station table, and
is accessed as a read only table during online transaction
processing.
Institution The Institution table contains one record for each institution
or retailer in the system acting as a terminal owner or
card-issuing authorization entity.
Institution_ID_OLTP This is a read-only table built from the Institution table for use
in routing to acquirers.

Institution_OLTP The Institution OLTP table contains one record for each
institution or retailer in the system acting as a terminal owner
or card-issuing authorization entity.
Institution table for read-only memory tables used by online

processes. This table is populated from the Institution table
Instrum_Typ_Interchange_Rel The Instrument Type to Interchange Relation table maps

known instrument types to a known interchange.
Instrum_Typ_Interchange_Rel_OLTP The Instrument Type to Interchange Relation OLTP table

maps known instrument types to a known Interchange.
This table is loaded from the Instrument Type to Interchange
Relation table.
Interchange_Prefix The Interchange Prefix table contains one record for each
unique prefix and pan length combination identifying an
issuer. The instrument type is optional and defaults to
spaces. If the instrument type is spaces, the instrument type
from the System Prefix table is used.
Interchange_Prefix_OLTP The Interchange Prefix OLTP table contains one record for
each unique prefix and pan length combination identifying
an issuer. The instrument type is optional and defaults to
spaces.
If the instrument type is spaces, the instrument type from the
System Prefix table is used.
This table is populated from the Interchange Prefix table,
Interchange_Security Interchange Security Data Source holds keys that interfaces

use to exchange keys, process PINs, and perform message
authentication.
Interchange_Security_OLTP Interchange Security Data Source holds keys that interfaces

use to exchange keys, process PINs, and perform message
authentication.
Interface The Interface table contains one record for each interface in
the system.
Interface_Acquirer_OLTP The Interface Acquirer OLTP table contains one record for
each interface in the system.
This table is populated from the Interface table.

Interface_Issuer_OLTP The Interface Issuer OLTP table contains one record for
each interface in the system.
This table is populated from the Interface table
Interface_Manager Interface Manager Data Source, used to store status and

performance information for interfaces. It contains one
record for each interface defined in the system. The record
is updated during online transaction processing.
Interface_OLTP The Interface OLTP table contains one record for each
interface in the system.
This table is built from the Interface table.
Interface_Station Interface Station table contains one record per input station.
The station record stores status information for stations.
There can be multiple stations per interface.
ISO_Field The ISO Field table contains one record for each message
profile which defines the characteristics of the fields.
ISO_Field_OLTP The ISO Field OLTP table contains one record for each
message profile which defines the characteristics of the
fields.
This table is loaded from the ISO Field table.
ISO_Message The ISO Message table contains one record for each
message profile and message type combination in the
system.
ISO_Message_OLTP The ISO Message OLTP table contains one record for each
message profile and message type combination in the
system.
This table is loaded from the ISO Message table.
Issuer_Profile_Map The Issuer Profile Map table contains an entry for each
issuer in the system. An issuer may be an institution
identifier or an interface name. The table assigns a primary
and alternate journal profile for each entry.
Issuer_Profile_Map_OLTP The Issuer Profile Map OLTP table contains an entry for
each issuer in the system. An issuer may be an institution
identifier or an interface name. The table assigns a primary
and alternate journal profile for each entry.
This table is populated from the Issuer Profile Map table, and
processing.
Issuer_Route_Profile The Issuer Route Profile table contains an entry for each
issuer route profile in the system.

Issuer_Txn_Allowed The Issuer Transaction Allowed table contains one record for
each issuer transaction profile, message category code and
processing code in the system that are allowed. This table
defines the set of transactions that are valid for an issuer for
on-us and not-on-us transactions.
Issuer_Txn_Allowed_OLTP The Issuer Transaction Allowed table contains one record for
each issuer transaction profile, message category code and
processing code in the system that are allowed. This table
defines the set of transactions that are valid for an issuer for
on-us and not-on-us transactions.
This table is populated from the Issuer Txn Allowed table
Journal This contains a record for each journaled transaction.
Journal_Continuation This contains 1 to n records when a transaction exceeds the

storage capacity of the Journal file. The num_jrnl_rec
represents the actual number of additional records required
to journal the transaction.
Journal_Continuation_Assign The Journal Continuation Assign table contains one record

for each journal continuation index entry. There should be, at
a minimum, four assign names (one for previous, one for
current, one for future and one for an empty file). The actual
number of assigns is defined in the Journal Continuation
Configuration table by adding the retention period
(previous), future count plus two for the current and empty
file.
Journal_Continuation_Assign_OLTP The Journal Continuation Assign OLTP table contains one

record for each journal continuation index entry. There
should be, at a minimum, four assign names (one for
previous, one for current, one for future and one for an empty
file). The actual number of assigns is defined in the Journal
Continuation Configuration table by adding the retention
period (previous), future count plus two for the current and
empty file.
This table is loaded from the Journal Continuation Assign
table.
Journal_Continuation_Config The Journal Continuation Configuration table contains a row

describing the journal continuation.
Journal_Data_Suppression The Journal Transaction Data Element Suppression table

contains a record by Institution, message type, transaction
code for each TDE Object ID NOT to journal. The message
type and transaction code may be wildcarded with asterisks.

Journal_Data_Suppression_OLTP The Journal Transaction Data Element Suppression table
contains a record by Institution, message type, transaction
code for each TDE Object ID NOT to journal. The message
type and transaction code may be wildcarded with asterisks.
This table is populated from the Journal Data Suppression
Journal_Prfl_Group_Assign_OLTP The Journal Profile Group Assign OLTP table contains one
record for each journal profile and index entry. There should
be at minimum four assign names for each journal profile
(one for previous, one for current, one for future and one for
an empty file). The actual number of assigns is defined in the
Journal Profile Group table by adding the retention period
file. This table is populated from the Journal Profile Group
Assign table and is accessed as a read only table during
Journal_Profile_Group The Journal Profile Group table contains a row for each
journal profile in the system.
Journal_Profile_Group_Assign The Journal Profile Group Assign table contains one record
for each journal profile and index entry. There should be, at
a minimum, four assign names for each journal profile (one
for previous, one for current, one for future and one for an
empty file). The actual number of assigns is defined in the
Journal Profile Group table by adding the retention period
file.
Journal_Profile_Group_OLTP The Journal Profile Group table contains a row for each
journal profile in the system.
This table is populated from the Journal Profile Group table,
Journal_Query_Config The Journal Query Configuration table contains a record for

each Journal Batch Query that can be performed.
Journal_Query_Config_Output The Journal Query Configuration Output table contains a

record for each Output location for a specific Journal Batch
Query that can be performed. The Journal Query Output
Profile specifies a a group of output files that are
round-robined to determine the next group of files to write to.

Journal_Query_Output_Assign The Journal Query Output Assign table contains one record
for each journal profile and index entry. There should be at
minimum one assign name for each journal profile. This
table allows a journal query to exceed the capacity of a
single file, and records will continue to be written to the next
file if the first file fills up.
Journal_Query_Output_Map The Journal Query Output Map table contains an entry for
each journal query output profile in the system. The table
defines the ring of journal query profiles to use for a journal
query.
Journal_Query_Position_Info The Journal Query Position Information table contains a

record for each specific Journal Batch Query and Journal
Profile combination that has retained position information
used to continue a journal batch query.
Limits This table contains limits that are associated with a payment
instrument (for example, card).
Limits_OLTP This table contains limits that are associated with a payment
instrument (for example, card).
Linked_Currency The Linked Currency data source is used to determine if the

transaction acquirer currency and cardholder billing
currency are linked. Currencies are linked if the conversion
rates and minor units for both currencies are the same.
Linked_Currency_OLTP The Linked Currency data source is used to determine if the

transaction acquirer currency and cardholder billing
currency are linked. Currencies are linked if the conversion
rates and minor units for both currencies are the same.
MDS_Interface The MasterCard Debit Interface configuration data.
MDS_Interface_OLTP The MasterCard Debit Interface configuration data.
Merch_Instr_Typ_Relation The Merchant Instrument Type Relation table contains one

record for each merchant and instrument type combination
allowed that are not already specified in the Merchant Table.
Merch_Instr_Type_Relation_OLTP The Merchant Instrument Type Relation OLTP table contains

one record for each merchant and instrument type
combination allowed that are not already specified in the
Merchant Table.
This table is populated from the Merchant Instrument Type
Relation table, and is accessed as a read only table during

Merchant The Merchant table contains one record for each merchant
in the system. Each record contains information for use in
validating transactions and defining cutover times.
Merchant_Delivery_Chan_Prfl_OLTP The Merchant Delivery Channel Profile OLTP table contains

one record for each channel profile.
This table is populated from the Merchant Delivery Channel
Profile table.
Merchant_Delivery_Channel The Merchant Delivery Channel table contains one record

for each controlled channel id in the system. The
characteristics of the channel are defined, including
attributes, counts and amounts.
Merchant_Delivery_Channel_Profil The Merchant Delivery Channel Profile table contains one

record for each channel profile.
Merchant_OLTP The Merchant OLTP table contains one record for each
merchant in the system. Each record contains information
for use in validating transactions and defining cutover times.
This table is populated from the Merchant Table.
NCR_PIN_Verification This data source holds the keys used in NCR PIN
Verification. It is accessed using a Pin Verification Profile
and a NCR PIN Version Number.
NCR_PIN_Verification_OLTP This read only data source holds the keys used in NCR PIN
Verification. It is built from the NCR_PIN_Verification table.
Outbound_Station_OLTP The Outbound Station table contains one record for each
station an outbound message may be sent out on. The
inbound station is its station pair that a corresponding
message will be received on. The inbound and outbound
station names may be the same value. The interface they
belong to is identified.
This table is populated from the Interface Station table, and
processing.
Perusal_Script The Perusal Script table contains one record for each user
profile, query type, and query name in the system. The
executable script name is defined.
Perusal_Script_OLTP The Perusal Script OLTP table contains one record for each
user profile, query type and query name in the system. The
executable script name is defined.
This table is populated from the Perusal Script table, and is
accessed as a read only table during online transaction
processing.
PMX This table is used as output for extract scripts.

Positive_Balance The Positive Balance table contains current balance and
status information for an account.
Pre_Auth_Hold The Pre-Auth Hold table contains pre-auth transactions for a

specified card.
Prefix The Prefix Data Source defines each prefix that can be
processed. One record must exist in the Data Source for
each prefix to be processed. Note that if one prefix is used
with multiple account number lengths, multiple records must
exist in the Data Source—one for each account number
length.
Prefix_OLTP The Prefix Data Source defines each prefix that can be
processed. One record must exist in the Data Source for
each prefix to be processed. Note that if one prefix is used
with multiple account number lengths, multiple records must
exist in the Data Source—one for each account number
length.
Route The Route table contains one record for each issuer route
profile, route code, account 1 type, and account 2 type
defined in the system. Through configuration of Route
records, issuers define the routing and authorization
parameters for their transactions.
Issuers may share common routing and authorization
configurations through the specification of a common issuer
route profile. Processing codes share common routing and
authorization configurations through the specification of a
common route code. A wildcard value of asterisks is
supported for the issuer route profile, route code, account 1
type, and account 2 type.
Route_OLTP The Route table contains one record for each issuer route
profile, route code, account 1 type, and account 2 type
defined in the system. Through configuration of Route
records, issuers define the routing and authorization
parameters for their transactions.
Issuers may share common routing and authorization
configurations through the specification of a common issuer
route profile. Processing codes share common routing and
authorization configurations through the specification of a
common route code. A wildcard value of asterisks is
supported for the issuer route profile, route code, account 1
type, and account 2 type.
This table is populated from the Route table, and is
processing.

Script The Script table contains one record for each compiled script
statement. The Script Component writes records to this
table when compiling scripts
Script_Configuration The Script Configuration table contains one record for each
executable script in the system. An executable script may be
a high level script or a subscript. The script compiler updates
the param types and return values.
Script_OLTP The Script OLTP table contains one record for each
compiled script statement.
This table is populated from the Script table, and is accessed
as a read only table during online transaction processing.
Script_Source The Script Source table contains one record for each line of
a script. All lines together define the flow of a script. The
script source is compiled and the output is placed in the
Script table and updated in the Script Configuration table.
Security_Device_Config Security Device Configuration table. This table defines

parameters for hardware security devices. One record will
exist for each security device
SPDH_act_cde_descr The SPDH Action Code Description table contains a text

description of the external action code value. The table also
identifies optional data to be returned in the response
message.
SPDH_act_cde_descr_OLTP The SPDH Action Code Description table contains a text

message. This table is read-only.
SPDH_Message The SPDH Message table defines the required Field

Identifiers (FIDs) and sub-FIDs by transaction code for
SPDH request and response messages. FIDs and sub-FIDs
to be included in Message Authentication (MAC)
calculations are also defined. Each field is defined as a
binaryf 8 data type. Bits 1 through 26 represent FIDs A
through Z. Bits 27 through 42 represent FIDs a through z.
Bits 53 through 62 represent FIDs 0 through 9. Bits 63 and
64 are not used.

SPDH_Message_OLTP The SPDH Message table defines the required Field
Identifiers (FIDs) and sub-FIDs by transaction code for
SPDH request and response messages. FIDs and sub-FIDs
to be included in Message Authentication (MAC)
calculations are also defined. Each field is defined as a
binaryf 8 data type. Bits 1 through 26 represent FIDs A
through Z. Bits 27 through 42 represent FIDs a through z.
Bits 53 through 62 represent FIDs 0 through 9. Bits 63 and
64 are not used. This table is read-only
SPDH_opt_resp_data The SPDH Action Code Description table contains a text

message.
SPDH_opt_resp_data_OLTP The SPDH Action Code Description table contains a text

message.
SPDH_txn_cde_descr The SPDH Transaction Code Description table contains the

text description of the transaction code that is returned in the
SPDH response message.
SPDH_txn_cde_descr_OLTP The SPDH Transaction Code Description table contains the

text description of the transaction code that is returned in the
SPDH response message.
Store_and_Forward The Store and Forward (SAF) table contains all the records
which are stored to be forwarded to the interface at a later
time.
Store_and_Forward_Outstanding The Store and Forward (SAF) Outstanding table contains all
the records which have been stored and have been sent to
the interface. A response is pending for the request and thus
the SAF message is outstanding.
Stream The Stream table contains one record per line to be written
to a write only data source. The line is composed of a single
variable length string.
Surcharge This is the surcharge data source. There is a surcharge

relationship between the acquirer and the issuer. An
acquirer might support surcharging but not for their own
cards for example. Also, different surcharges can be
implemented by processing code and transaction currency
codes.

Surcharge_OLTP This is the surcharge datasource. There is a surcharge
relationship between the acquirer and the issuer. An
acquirer might support surcharging but not for their own
cards for example. Also, different surcharges can be
implemented by processing code and transaction currency
codes.
System_Algorithm_Config The System Algorithm Configuration table contains one

record for each prefix/prefix range and pan length based on
the algorithm name.
System_Algorithm_Config_OLTP The System Algorithm Configuration OLTP table contains

one record for each prefix/prefix range and pan length based
on the algorithm type.
This table is populated from the System Algorithm Config
System_Prefix The System Prefix table contains one record for each
unknown prefix to search based on Source Logical Network
and Acquirer Route Profile to determine the correct issuer.
The Acquirer Route Profile may be wildcarded.
System_Prefix_OLTP The System Prefix OLTP table contains one record for each
unknown prefix to search based on Source Logical Network
and Acquirer Route Profile to determine the correct issuer.
The Acquirer Route Profile may be wildcarded.
This table is populated from the System Prefix table, and is
processing.
TMX This table is used as output for extract scripts.
Usage This table contains the usages associated with a card,

account or customer. The usages are checked against limits
which are defined in the Limit Table.
User_Perusal_Prfl_Relation The User Perusal Profile Relation table contains one record
for each alias which maps specific users to a user profile.
User_Perusal_Prfl_Relation_OLTP The User Perusal Profile Relation OLTP table contains one
record for each alias which maps specific users to a user
profile.
This table is populated from the User Perusal Profile
Relation table, and is accessed as a read only table during
Visa_Interface The Visa Interface configuration data.
Visa_Interface_OLTP The Visa Interface configuration data.

Visa_PVV Visa PVV contains information specific to the Visa PVV PIN
verification method.
Visa_PVV_OLTP Visa PVV contains information specific to the Visa PVV PIN
verification method.
VisaII_act_cde_descr The VISA II Action Code Description table contains a text

description of the external action code value. The action
code description is limited to 16 bytes.
VisaII_act_cde_descr_OLTP The VISA II Action Code OLTP Description table contains a

text description of the external action code value. The action
code description is limited to 16 bytes.
VisaII_Confirmation_Pending The Visa II Confirmation Pending data source contains a

record for each transaction that a confirmation is still
pending. When the confirmation is received, the record is
deleted. There is no UI access to this file.
4.2.11 Financial transaction flow

BASE24-es card transactions may be broadly categorized as on-us (transactions
involving a card issued by the institution running the local copy of BASE24-es),
and not-on-us (transactions involving a card issued by some other institution).
On-us transactions are generally authorized by BASE24-es, or are routed by

BASE24-es to a back-end host application for authorization. Not-on-us
transactions are generally switched by BASE24-es to a national or local card
network for authorization.
On-us transactions can be further categorized by authorization type.
Both on-us and not-on-us transactions can also be independently categorized by

acquirer type as switch-acquired transactions, device-acquired transactions, or
host-acquired transactions.
Authorization type
Authorization type impacts both the configuration of the AOR (the files required
by BASE24-es, the requirement or lack of requirement for a security module) and
the TOR (the requirement or lack of requirement for connectivity to an external
authorizer.)
Switched transactions
Switched transactions are routed by BASE24-es to a local or national card
network for authorization. The card networks commonly prefer that the
BASE24-es system connect to them as a TCP/IP client (see “TCP/IP client

considerations” on page 51), though existing protocols provided in the z/OS
implementation of BASE24-es by the VTAM reader/writer programs (see “VTAM
considerations” on page 54) are commonly still supported.
Offline authorization
Offline transactions are transactions authorized by BASE24-es on the basis of
data configured in its own database. They require no connectivity in the TOR to
an external authorizer.
Online authorization
Online transactions are transactions routed by BASE24-es to a back-end host
application for authorization. On other platforms, connectivity to the back-end
host application is generally accomplished by some variety of TCP/IP data
communications.
Though this is also possible in the z/OS implementation of BASE24-es, it is

generally preferable with a CICS-based host application to take advantage of the
fact that BASE24-es and the back-end application are co-located, and use some
CICS facility to communicate. BASE24-es supports several options natively; one
desirable option is the use of a CICS LINK from the BASE24-es AOR to the
back-end host AOR. However, connectivity is generally driven by the
requirements of the back-end host system, and is often customized to meet
those requirements.
With the Online authorization method, it is common for BASE24-es to perform

some pre-screening such as Personal Identification Number (PIN) validation
before passing the transaction to the back-end host application for authorization.
Online/Offline authorization
Online/Offline authorization is the authorization method in which BASE24-es
generally goes to a back-end host system for authorization, but is capable of
standing in and performing authorization when the back-end host application is
unavailable.
Considerations for communications with the back-end host application are the
same as for Online Authorization described in “Online authorization” on
page 88). Considerations for the authorization database in the AOR are similar to
the considerations for Offline Authorization described in “Offline authorization” on
page 88, although a less robust authorization method (perhaps declining hot
cards rather than checking available balances) is often selected.
Acquirer type
The Acquirer type generally has little direct impact on the configuration of the
BASE24-es AOR other than the specific module within BASE24-es to which the

external message is routed for parsing. It has more direct implications on the
configuration of the TOR.
Device-acquired transactions
Device-acquired transactions include transactions that originate from automated
teller machine (ATM) and point of sale (POS) devices. They may include a mix of
on-us and not-on-us transactions, and may be authorized by any of the
authorization types.
ATM-acquired transactions generally enter the BASE24-es system through the

BASE24-es TCP/IP Server Communications handler (see “TCP/IP server
considerations” on page 52), though it is also common for transactions from older
ATMs to enter the system through an existing protocol and the VTAM
reader/writer transactions (see “VTAM considerations” on page 54).
POS devices have a wide variety of protocol requirements. In the z/OS

implementation of BASE24-es, transactions from these devices commonly flow
through an external protocol adapter and enter the BASE24-es system through
the BASE24-es TCP/IP Server Communications handler as described in “TCP/IP
server considerations” on page 52.
Switch-acquired transactions
Switch-acquired transactions are transactions acquired from local or national
card networks and routed to BASE24-es for authorization. They generally
represent on-us transactions only and are authorized as one of the Online,
Offline, or Online/Offline authorization types.
Connectivity to the card networks is generally through the BASE24-es TCP/IP

Client Communications Handler (see “TCP/IP client considerations” on page 51),
although existing protocols provided in the z/OS implementation of BASE24-es
by the VTAM reader/writer tasks (see “VTAM considerations” on page 54) are
commonly still supported.
Host-acquired transactions
Host-acquired transactions are usually transactions acquired by devices or card
networks connected to a pre-existing host application and routed to BASE24-es
for authorization or routing. Host-acquired transactions are most generally
not-on-us transactions (since on-us transactions are often processed by the
pre-existing host application itself), but may be a mix of switched and offline
transactions.
On other platforms, connectivity to the acquiring host application is generally

accomplished by some variety of TCP/IP data communications. Though this is
also possible in the z/OS implementation of BASE24-es, it is generally preferable
with a CICS-based host application to take advantage of the fact that BASE24-es

and the back-end application are co-located, and use some CICS facility to
communicate. Connectivity is generally driven by the requirements of the
acquiring host system, and is almost always customized to meet those
requirements.
4.2.12 BASE24-es logical architecture on z/OS

Figure 4-18 is the original picture of the BASE24-es logical architecture, with
components re-labeled to reflect the implementation-specific component that
provides the logical services.
External Endpoints
Automated Point of Sale Host Public Card

Teller Devices Acquiring or Networks
Machines Authorization
Protocol &
Middleware CICS Sockets/VTAM
Base24-es Communication Handlers
CICS TDQ/WebSphere MQ
Memory-Based Queues
Application
Integrated Server Integrated Server Integrated Server

System Interface System Interface System Interface
Services Services Services
Database
CICS/SMSVSAM Manager
VSAM/RLS Database
Figure 4-18 BASE24-es logical architecture
4.2.13 User interface region architecture

The BASE24-es User Interface consists of an HTTP client, an HTTP server, and
CICS-based server components. The HTTP server runs on a Windows© or
pSeries® front-end, and communicates with the CICS-based servers over
TCP/IP.

There are two different kinds of file servers, each with different characteristics.
One communicates with the HTTP server using the BASE24-es TCP/IP Server.
The other one communicates with the HTTP server using facilities provided by
the IBM CSKL TCP/IP Listener.
The majority of file access is accomplished using the CSKL-based CICS file
server SIDC. SIDC uses both the TCP/IP stack and the BASE24-es file
definitions, and must be run in a region containing both sets of resources. This
may be an AOR with TCP/IP services configured, a TOR with file definitions
configured, or a CICS region dedicated to BASE24-es UI processing.
SIDC UI Server
UI CICS Region
BASE24-es CSKL BASE24-es

UI Client Security
Validation
Program
BASE24-es UI Server SIDC
...
CICS RLS File Definitions
VSAM/RLS Database
Figure 4-19 BASE24-es UI File Server (SIDC)
The CICS-based SIDC UI server cannot work with the BASE24-es TCP/IP
Communications handlers and asynchronous message queuing because it must
maintain a transactional session with the UI Server.
In Figure 4-16 on page 66, the BASE24-es UI Server opens a session with the
IBM CSKL listener. CSKL links to the BASE24-es Security Validation Program to
verify the UI Server’s credentials. The Security Validation Program validates the
UI Server, so CSKL starts SIDC. CSKL calls GIVESOCKET and SIDC calls

TAKESOCKET, so SIDC becomes the owner of the socket established by the UI
Server. SIDC then sends a good response to the UI Server.
The UI Server now sends a BEGINTRANSACTION message to establish

transactional boundaries around the database updates to follow. SIDC combines
all subsequent updates into a single database transaction until the UI Server
calls ENDTRANSACTION, at which time SIDC performs a sync point.
The SIDC task persists until the client connection is lost. If the connection is lost
within the boundaries of a database transaction, SIDC rolls back any database
updates in progress before terminating.
XMLS UI server
The CICS-based XMLS UI Server is a special instance of the BASE24-es
Integrated Server used to route BASE24-es Process Control commands to the
Integrated Server, and to provide C++ facilities to access certain
specially-formatted data in some BASE24-es data files.
It has no special database transactional considerations and operates similarly to

the BASE24-es Integrated Server; the user interface connects with the
BASE24-es TCP/IP Server, which places messages destined for the XMLS
server on an associated TDQ.

BASE24-es
UI Client BASE24-es UI Server
TOR
BASE24-es
IP Server IP
TDQ
AOR
XDYR IS
XMLS
TDQ
XMLS IS
TDQ
Figure 4-20 BASE24-es UI Application Server (XMLS)
In Figure 4-20 the UI Client sends a BASE24-es Process Control command to

the UI Server. The UI Server sends the request to a BASE24-es TCP/IP Server
running in the TOR.
The TCP/IP server starts the AOR Routing Program XDYR and XDYR places the
transaction on the XMLS TDQ.
XMLS reformats the UI message as an IS Process Control message and places

the message on the IS TDQ. (Note that in BASE24-es Version 06.2, all Process
Control messages can be processed equivalently by any member of a server
class.)
The Integrated Server processes the request and places a response on the
XMLS TDQ (see “TCP/IP connected security module” on page 61 for details or
synchronous messages between BASE24-es components.) XMLS reformats the
response and places it on the TCP/IP Communication Handler’s TDQ for delivery
to the client.

4.2.14 Scalability considerations
The following discussion deals with replication for scalability. Replication for
availability receives a separate discussion in a subsequent chapter.
Replicating TORs
BASE24-es TCP/IP Communications Handlers may be replicated within a TOR,
and TORs may be replicated within a CICSPlex or other MRO environment.
TCP/IP Client Communications Handlers may be freely replicated. For example,

though it would rarely be necessary from a throughput perspective, multiple
TCP/IP Clients in multiple TORs might be configured to communicate with a
single remote endpoint, say a large public card network. Each TCP/IP handler is
configured with its own symbolic name, and is defined to the application as a
separate routable endpoint.
TCP/IP Server Communications Handlers may be replicated within limits

imposed by the sockets API. Specifically, the sockets API does not permit two
listeners at the same IP address to bind to and accept connections on the same
TCP port. Network routers, z/OS Shared Sockets or z/OS Virtual IP Addressing
may be used to permit multiple TCP/IP Servers to share the same IP address
and TCP port.
BASE24-es IP Communications Handlers (Client or Server) may be distributed

across any number of TORs and LPARs subject to the above considerations.
Replicating AORs
BASE24-es AORs may be freely replicated. As BASE24-es will attempt to
distribute work evenly across available AORs, consideration should be given to
defining more AORs in LPARs with more available resources.
Replicating tasks
BASE24-es Integrated Server tasks can and should be replicated within an AOR.
The MAXSERVERS parameter in the BASE24-es System Interface Services
Message Delivery SDMF configuration file controls the number of tasks created
in each AOR.
There is little overhead associated with configuring an ample number of tasks. As

a rule of thumb, there is no excessive overhead if half the instances of a server
class are in an ECK_WAIT state during a period of relatively high activity.

Partitioning files
BASE24-es provides several mechanisms for logically partitioning some of the
most intensively-used files for purposes of scalability. This is provided in addition
to file-system level partitioning.
Financial Journal File partitioning

Financial Journal (JRNL) files can be partitioned across institutions. It is also
possibly to further partition the JRNL files associated with a given institution by
card range, minimizing the possibility that inserts to the JRNL will become a hard
bottleneck.
Card and Usage file partitioning

Card and usage files can be logically partitioned by card range, minimizing
contention on these files.
Interface file partitioning

The SAF, Interface Manager, and Interface Station files may be partitioned by
interface, minimizing lock contention in these files.
Replicating LPARs
BASE24-es may be distributed across multiple Logical Partitions (LPARs) in a
Parallel Sysplex. The Parallel Sysplex may be distributed across multiple physical
processing units. The only requirement is that all the context-free servers in a
BASE24-es system must have a common database in which to hold their shared
context.
Benchmark observations
ACI Worldwide and IBM conducted a benchmark from November 7-18, 2005, at
the IBM Washington Systems Center to demonstrate the performance
characteristics of BASE24-es on the latest IBM System z9. The objective was to
achieve a processor consumption rate competitive with other solutions of this
type, and to maintain virtually constant MIPS per transaction rates as volumes
increased to a throughput of over 1000 transactions per second. The effort met
or exceeded these objectives.
򐂰 Performance and scalability were close to linear, while the cost per
transaction remained low and constant.
򐂰 The test achieved more than 1000 transactions per second on 8 CPUs
(limited only by I/O capacity). This represents twice the daily peak volume
requirements of the world’s largest transaction processors.
򐂰 System z9 performance can satisfy any customer’s requirements with near
linear MIPS per transaction. TCO per transaction becomes lower with higher
transaction volumes as support costs are fairly constant as volume grows.

In addition, BASE24-es clients who run the solution on the IBM System z9
mainframe can take full advantage of the other benefits of the mainframe
platform, such as superior security, utilization rates of over 90%, dynamic
workload balancing for improved efficiencies, virtually unlimited scalability,
superior performance and a system designed for continuous operations.
WSC Benchmark scalability test – single LPAR

Test configuration:
z/OS 1.6
VSAM 1.6
CICS TS CPSM 2.3
Storage subsystem 2105-800 8 LSS
LPAR configurations 2094-701 1 CPU
2094-702 2 CPU
2094-704 4 CPU
2094-708 8 CPU
Coupling Facility (2) 2094-701 1 CPU each
CICS statistics on (all), interval 5 minutes
RMF statistics on sys(notype(032,089,099),interval(500), nodetail)
subsys(stc, type(000:018,020:255, interval(500))
Results:
Test results are summarized in Figure 4-21 on page 97. Throughput
(transactions per second) and cost (MIPS per transaction) data was captured for
each of four LPAR configurations: 1, 2, 4 and 8 CPUs. MIPS per transaction
remained constant through approximately 1018 transactions per second, where
the I/O subsystem became saturated.
The solid upper line shows transactions per second, while the dotted lower line
shows that MIPS per transaction remained nearly constant as the transaction
rate increased. (Absolute CPU utilization and MIPS cost are not shown. For
accurate sizing information, contact ACI.)

CPSM - Single LPAR
CPU Busy
1200
Transactions per Second

1018
MIPS per Transaction

1000
764 Transactions
800 per second
600
431 MIPS per
400 transaction
230
200
0
1 2 4 8
Number of CPUs
Figure 4-21 Benchmark result
WSC Benchmark scalability test – dual LPAR

Test configuration:
z/OS 1.6
VSAM 1.6
CICS TS 2.3, CPSM 2.3
Storage subsystem 2105-800 8 LSS
LPAR configuration (2) 2094-704 4 CPU each
Coupling Facility (2) 2094-7011 CPU each
CICS statistics on (all), interval 5 minutes
RMF statistics on sys(notype(032,089,099),interval(500), nodetail)
subsys(stc, type(000:018,020:255, interval(500))
Results:
Test results are summarized in Figure 4-22 on page 98. Two 4-CPU LPARs
processed approximately the same throughput as a single 8-CPU LPAR (again
limited by I/O capacity), but the MIPS per transaction was slightly higher and
non-linear (15% at 1072 TPS).
The solid upper line shows transactions per second, while the dotted lower line
shows that MIPS per transaction increased only slightly as the transaction rate
increased. (Absolute CPU utilization and MIPS cost are not shown. For accurate
sizing information, contact ACI.)

CPSM - Two LPAR
LPAR2 - CPU Busy
Transactions per Second

1200
MIPS per Transaction

1072
1000 Transactions Per
800 Second
667
600 MIPS Per
298 Transaction
400
200
0
LPAR2 - CPU Busy
Figure 4-22 Benchmark result

5
Chapter 5. Designing the system layout

Most customers consider the implementation of e-payment systems like
BASE24-es to be among their mission-critical applications. And for
mission-critical applications, availability is a very significant business
requirement.
In this chapter, we explain the high availability characteristics of each BASE24-es

component. We also show you four possible options for configuring BASE24-es
when running under CICS and z/OS.

5.1 Pain versus gain
Availability for this type of a system is typically measured by the number of 9’s
that a system can achieve. Referring to Table 5-1, the first row would be
considered two 9’s and the last row would be considered seven 9’s of availability.
With each 9 comes a higher degree of cost and complexity. Customers must
weigh the cost of downtime against the cost of avoiding downtime to determine
the high availability model that best fits their business requirements.
Table 5-1 illustrates the relationship between the 9 and the actual business
service downtime.
Table 5-1 Downtime and availability relationship

Availability Downtime per year
99% Less than 88 hours
99.9% Less than 9 hours
99.99% Less than 53 Minutes
99.999% Less than 5.3 minutes
99.9999% Less than 32 seconds
99.99999% Less than 3.2 seconds
5.1.1 Planned versus unplanned outages

When designing a system for high availability, you need to consider both planned
and unplanned events that could make the system unusable. Planned events
include moving to new versions of the application software or operating system,
introducing new LPARs or CICS regions into the system, maintaining the
hardware or introducing new hardware components, and so on. Unplanned
events include application, operating system, or hardware failures and the like.
In a properly configured and maintained system, both planned and unplanned

events can be managed to meet the customer’s availability requirements.
5.1.2 Single site versus dual site

A single site can be designed to provide very high availability, but it cannot
withstand a catastrophic event that takes the site completely offline. To survive
this type of event, BASE24-es can be configured in a dual site model, with one
site being the Active site and the other being a Passive backup site.

In this model, there are two copies of the BASE24-es data and data replication
software. They are used to keep the data synchronized. If a failure occurs at the
primary site, the backup site can be activated to take over the processing. This
model is also useful during large-scale planned events, such as moving a data
center.
5.2 High availability considerations

In the following sections, we list considerations to keep in mind to enable the
BASE24-es solution to achieve a high level of availability.
5.2.1 Terminal Owning Regions (TORs)

To provide high availability in a BASE24-es system, two or more Terminal
Owning Regions (TORs) are required. Redundant TCP/IP Server connections
should be configured across multiple TORs. BASE24-es is designed to use
round robin processing across all available TCP/IP Server connections. It detects
when connections are lost and re-established. TCP/IP Clients utilize Virtual IP to
connect to BASE24-es.
Virtualization
There are two virtualization strategies available when configuring BASE24-es.
The first takes advantage of an ATM or point of sale (POS) device’s capability to
have a primary and alternate destination configured. The second strategy is to
provide Virtual IP Addressing (VIPA), a router, Sysplex Distributor, or a
combination of those.
Primary/Alternate configuration
Figure 5-1 on page 102 illustrates a primary/alternate configuration. In the figure,
note that half of the devices are configured with a primary destination of 1.2.3.4,
port 8000 and a secondary destination of 1.2.3.5, port 8000. The other half of the
devices are configured with primary destination of 1.2.3.5, port 9000 and a
secondary destination of 1.2.3.4, port 9000.
There is a separate TOR responsible for listening on each of the ports. The TORs
are spread across the two LPARs. The TORs are configured such that the
primary and secondary destination for all terminals will reside in different LPARs.
Chapter 5. Designing the system layout 101

LPAR 1 LPAR 2
CICS TOR CICS TOR CICS TOR CICS TOR
1.2.3.4 1.2.3.4 1.2.3.5 1.2.3.5

Port 8000 Port 9000 Port 8000 Port 9000
Figure 5-1 Primary/Alternate configuration
If one of the TORs fails, the devices connected to it before the failure would utilize
their secondary destination. This is illustrated in Figure 5-2 on page 103.

LPAR 1 LPAR 2
CICS TOR CICS TOR CICS TOR
1.2.3.4 1.2.3.5 1.2.3.5

Port 9000 Port 8000 Port 9000
Figure 5-2 Alternate route
BASE24-es can utilize the z/OS Automatic Restart Manager (ARM) to detect the
TOR failure and restart it. After it returns to service, customers have two choices:
either allow the connections to remain on the alternate TOR, or issue a command
to terminate those connections.
Terminating the connections will cause the devices to reestablish connections

with the primary TOR and balance the connections across the LPARs.
(Customers may choose to perform this action during off-peak hours, because it
results in a slight interruption of service for the affected devices.)
Virtual IP
In this model, the IBM Virtual IP Address (VIPA) or a router is utilized to provide
virtualization of the TORs. TCP/IP Client connections are distributed across the
available TORs by VIPA or the router.
As you can see in Figure 5-3 on page 104, balancing of the connections is done
dynamically. This is unlike the previous model, in which the balancing of
connections was done through static configuration at the device.

VIPA
LPAR 1 LPAR 2
CICS TOR CICS TOR
Figure 5-3 VIPA
In the case of a TOR failure, shown in Figure 5-4 on page 105, the devices that
were connected to the TOR will reconnect to the same IP address and port. The
connection will be distributed to one of the surviving TORs, as depicted.
After the failed TOR returns to service, it is difficult to rebalance the connections
across all TORs and LPARs. However, this should not cause a problem because
the majority of the work is done in the AOR, and the TORs distribute the load
across all available AORs.
Over time, connections will rebalance as devices lose and reestablish their
connections. Customers can choose to expedite rebalancing by terminating all
connections and forcing the all devices to reconnect. (However, this action should
be done during off-peak hours only.)

VIPA
LPAR 1 LPAR 2
CICS TOR
Figure 5-4 Failover scenario
Maintenance - TOR
When maintenance is required on some aspect of the TOR, the IP Handlers
within the TOR can be shut down in a controlled manner. During a controlled
shutdown, the IP Handler will terminate all connections. Any messages sent to IP
Handler after the connections have been terminated will be failed back to the
AOR. Typically, this will result in the message being rerouted or reversed
In addition, an IP Server will notify the application that the connections have been
terminated and the AOR will go into Network Management Mode and periodically
monitor the status of the connection.
5.2.2 Application Owning Regions (AORs)

Redundancy is the key high availability consideration for Application Owning
Regions (AORs). Multiple copies of the Integrated Server (IS) processes should
run in every AOR, multiple AORs should be configured in each LPAR, and the
system should have at least two LPARs.

Integrated Server failure
In the unlikely event that an IS fails, the surviving IS tasks will continue to deal
with the messages. The BASE24-es Task Monitor is responsible for detecting an
IS failure and restarting it.
AOR failure
In the event of an AOR failure, the TORs rely on the Dynamic Transaction
Routing Program in a CICSPlex environment, and routing user exit in a
non-CICSPlex environment, to detect and route around the AOR failure. The few
transactions that were on the TDQ at the time of the failure will be lost.
BASE24-es relies on the end-to-end protocols built into the electronic payment
industry to deal with the transaction appropriately.
z/OS Automatic Restart Manager (ARM) can be used to quickly restart the AOR
after a failure. Refer to 4.2.10, “AOR architecture” on page 56 for an in-depth
discussion about starting the AOR.
Maintenance - AOR
Occasionally, maintenance will be required on the AOR. This could take two
forms, either maintenance to the Integrated Servers, or maintenance to the AOR
itself. This section describes how maintenance can be performed on the AOR
without disrupting service.
Integrated Server maintenance

BASE24-es provides a mechanism to stop and restart all IS processes in turn.
This mechanism is useful when introducing new IS software or when bringing in
new configuration data.
AOR maintenance
When AOR maintenance is necessary, the AOR can be shut down in a controlled
manner. When a shutdown is required, the AOR will notify the TOR that it should
no longer be considered in the AOR selection process. This will stop new work
from being added to the TDQ. After all messages on the TDQ have been
processed, the Integrated Servers will be stopped and the AOR will be
terminated.
After maintenance is complete, the AOR can be restarted. Refer to 4.2.10, “AOR
architecture” on page 56 for an in-depth discussion of starting the AOR.
5.2.3 File access

In the following sections, we discuss the high availability considerations for the
files environment.

Data sharing
Data sharing is the concept that applications running in different CICS regions
can share access to data, no matter which z/OS image in a sysplex they happen
to be running on. Ideally, all CICS AORs should be able to access all of the data
needed to run any transaction. If you have only one system image, you should
find it easy to achieve this access.
If you have two or more z/OS system images in your sysplex, however, you may
have to implement specific levels of database or access method software to
enable all AORs to access the same data at the same time, no matter which
z/OS system the AOR is running in. DFSMS (SMSVSAM), in conjunction with
CICS TS, provides similar facilities for VSAM data sets.
Multiple CICS/ESA® applications in a CICSplex can share VSAM data sets with
integrity by using function shipping to a single FOR. This approach has
limitations, though. It does not solve the problem of the FOR being a single point
of unavailability. Other CICS applications outside the Parallel Sysplex can also
use the FOR by function shipping over ISC links. DFSMS provides RLS access
for VSAM files. CICS TS allows CICS users to exploit this function. RLS is
designed to allow VSAM data to be shared, with full update integrity, among
many applications running in many CICS regions in one or more z/OS system
images in a Parallel Sysplex.
The SMSVSAM server runs in its own address space and handles all VSAM
requests made in RLS mode. If this address space fails, it can be restarted
automatically up to six times, assuming RLSINIT=YES is specified in the
IGDSMSxx member that the host z/OS image uses. The most likely reason for an
SMSVSAM server failure is loss of connectivity to the Coupling Facility.
The SMSVSAM address space fails if it cannot rebuild a Coupling Facility lock
structure, if it loses connectivity to the Coupling Facility lock structure, or if the
Coupling Facility lock structure fails. The SMSVSAM address space also fails
when it loses the last active SHaring Control Data Sets (SHCDS) and has no
spare available.
When the SMSVSAM address space fails, any I/O against a VSAM data set open
in RLS mode, or any attempt to open a VSAM data set in RLS mode, receives an
error indicating that the SMSVSAM is unavailable. CICS detects that SMSVSAM
address space is unavailable when one of the transactions attempts an RLS
access.
When this event occurs, CICS issues message DFHFC0153: File control RLS
access is being closed down. The effect that this has varies, depending on
where individual tasks are in their life cycle.

A typical transaction accesses several records in VSAM data sets and then
performs either an implicit or explicit sync point. Statistically, a task is most likely
to discover the failure of the SMSVSAM address space when it next attempts to
read from or write to an RLS mode VSAM data set. It is less likely that a task
discovers that SMSVSAM has failed when it tries to sync point and releases its
locks in the VSAM data set.
Because SMSVSAM fails if it loses connectivity to the Coupling Facility lock

structure, consider having more than one Coupling Facility. The possibility of
SMSVSAM server failure is minimized if you have two or more Coupling
Facilities.
Coupling Facilities
A Coupling Facility (CF) is a special LPAR that provides high-speed caching, list
processing, and locking functions in a Parallel Sysplex. The LPAR can be
configured to run within a CP that runs other operating systems (it is often
referred to as an ICF in this case), or it could be configured in a stand-alone
processor that only runs Coupling Facility Control Code (CFCC).
It is important to understand the difference between a CF and an operating

system LPAR. If an operating system LPAR dies, all work that was running in that
system is dead and must be restarted.
On the other hand, depending on which structures are in a CF that suffers a

failure, it is possible that only a short pause in processing will be encountered as
the CF contents are rebuilt in an alternate CF.
As a result, it is possible that a CF will not be impacted by a failure to the same

degree as an operating system. More importantly, from an availability
perspective, is whether the structures in the failed CF support rebuild and
duplexing, and whether the failure impacted any connected systems or not.
Maximizing availability of CF service

In this section we provide a set of recommendations to maximize the availability
of the CF service. We also touch on performance, because good performance is
especially important for a CF, and poor CF performance can be disruptive for the
systems connected to it.
In order to be able to deliver acceptable response times, you must have enough
capacity in your CFs. In addition to CF CPU utilization, you must also consider
response times. It is possible for CF CPU utilization to be low, but with CF
response times so long that the cost of using the CF becomes unacceptable (this
is especially the case if there is a large disparity in speed between the CF CPC
and the CPCs that the operating systems are running on).

Regardless of the type of link being used, you need to configure for the following
qualities:
򐂰 Availability
You need at least two links between each pair of CPCs (at a minimum). CF
links can be shared using Multiple Image Facility (MIF). However, there can
only be one CF LPAR sharing a given “end” of a link, and the operating
system LPARs sharing a link must all be in the same sysplex. Remember that
if you share a link between multiple LPARs, the loss of that link will impact all
those LPARs.
򐂰 Performance
In the early days of Parallel Sysplex, two CF links provided sufficient capacity
for most users. Now the intensiveness with which CFs are being used has
increased to the point that in many sites, two CF links are no longer sufficient.
However, the rule of thumb for PATH BUSY and SUBCHANNEL BUSY
remains the same: both PATH BUSY and the number of requests delayed
because of SUBCHANNEL BUSY should not exceed 10 percent of the
number of requests.
In addition, subchannel utilization (calculated by multiplying the number of
requests on the subchannel by the average response time, divided by the
number of seconds in the interval) should not exceed 30%. Ensure that you
configure sufficient links so that these thresholds are not exceeded during
peak processing windows.
򐂰 Avoid single points of failure
There should be as few single points of failure as possible in common
between the links used to connect a set of LPARs. When you order CF links,
the configuration provided will normally do the best to avoid single points of
failure.
Failure isolation
One of the most important characteristics of a CF is its location in relation to the
systems that are connected to it. There are some advantages to it being in a
stand-alone processor. However, the most important question is whether a single
failure can impact both the CF and one or more systems connected to it.
A CF in a CPC where none of the other images in that CPC are connected to that
CF can provide nearly as good availability as one running in a stand-alone
processor. Many structures can be rebuilt even if there is a double failure.
However, some of the most important structures in a data sharing environment

cannot recover from such a failure without a data sharing group restart (unless
the structure is duplexed). To determine your need for a failure-isolated CF,
review the structures you use to see if any of them fall into that category.

Structure placement
Structures that have a failure isolation requirement should either be placed in a
failure-isolated CF (normally a stand-alone CF), or else they should be duplexed.
Another thing to beware of is structures ending up in a CF other than the one you
intended. This can happen, for example, following CF maintenance if the correct
procedures are not followed to return the structures to their intended CF.
To avoid this situation, regularly verify that all structures are actually in the first
available CF in their PREFLIST. One way to do this is by using the z/OS Health
Checker, which has a check for this condition. In addition, consider issuing a
SETXCF START,REBUILD,POPCF=cfname for each of your CFs on a regular
basis. This will ensure that all structures are in the first CF in their PREFLIST (all
other things being equal).
Never use a REBUILDPERCENT greater than 1. REBUILDPERCENT can be

used to stop a structure being moved in case of a partial connectivity failure.
However, with the speed of modern CF technology, it would be very rare that you
would not want to move a structure following a connectivity failure. For simplicity,
the best thing is just to not use that keyword at all—the default is to always
rebuild a structure following such a failure.
Recovering from CF failure

As stated previously, CFs are different from operating system LPARs in that if a
CF fails, it is possible that the CF contents can be recreated in an alternate CF,
allowing operations to continue with just a pause in processing. When planning
for CF recovery, you need to consider whether the structures in the failed CF
support recovery from a CF failure, and whether the remaining CFs have
sufficient capacity to take over from the failed CF.
Structure recovery Whether a structure can recover from a CF failure or not
depends on the exploiter that uses that structure, and
whether it supports structure rebuild or duplexing.
Structure rebuild There are two flavors of structure rebuild: User-managed
rebuild and system-managed rebuild.
Exploiters that support user-managed rebuild can
generally recreate the structure contents following the
failure of a CF or one of the connectors. However, in some
cases they cannot recover from a double failure.
Exploiters that only support system-managed rebuild
cannot recreate the structure contents following a CF
failure.
User-managed rebuild, which is supported by most of the
earlier CF exploiters, means that the address spaces
connected to a structure take responsibility for moving a

structure in case of a planned move, or of recovering the
structure in case of an unplanned move (failure of a CF,
for example). Most exploiters that support user-managed
rebuild maintain an in-storage copy of their data from the
structure, meaning that even if the CF containing the
structure is lost, they can still create a new version by
merging all their in-storage information.
System-managed rebuild, introduced with OS/390® 2.8
and CF level 8, makes it significantly easier for products
to avail themselves of the facilities provided by a CF.
Whereas the connector is responsible for moving a
structure when using user-managed rebuild, the operating
system takes over that role when using system-managed
rebuild. In this case, the connected operating systems
work together to copy the contents of a structure from one
CF to another. The systems have no knowledge of the
meaning of the contents of the structure; they just copy
the structure in its entirety.
Structure duplexing
Similarly, there are two flavors of structure duplexing, which are User-managed
duplexing and System-Managed CF Structure Duplexing.
User-managed duplexing is only used by DB2, and then only for its Group Buffer
Pool structures. System-Managed CF Structure Duplexing is supported for any
structure whose connectors support system-managed rebuild.
Both types of structure duplexing provide the ability to continue processing

following a failure (assuming there is no single point of failure between the CFs
containing the two copies of the structure). We describe both types in more detail
in the following sections.
User-managed structure duplexing

User-managed structure duplexing was introduced in OS/390 R3 with APAR
OW28460, and was integrated into OS/390 R6. This capability is used by DB2 for
its group buffer pool structures.
With user-managed duplexing, setting up the second structure, maintaining

synchronization between the structures, and recovering from structure-related
failures is the responsibility of the connector (DB2, in this case). Turning on
user-managed duplexing for a DB2 GBP structure introduces very little additional
overhead, but delivers significant benefits in terms of reduced recovery time if the
GBP structure is lost.

System-Managed CF Structure Duplexing
System-Managed CF Structure Duplexing was announced as part of z/OS 1.2,
with support in CF level 11 and later. With System-Managed CF Structure
Duplexing, the operating system takes responsibility for setting up a duplex copy
of a CF structure, maintaining the duplex relationship, and reverting back to
simplex mode in the event of a CF or CF link failure.
Assuming that the two copies of a duplexed structure are failure-isolated from
each other, the use of System-Managed CF Structure Duplexing for a structure
allows you to place a structure in the same failure domain as one or more of its
connectors, but still be protected from a single failure that would otherwise cause
a data sharing group restart. However, while System-Managed CF Structure
Duplexing provides substantial availability benefits, it can also have a significant
impact on the response time for update requests to the duplexed structures.
Number of Coupling Facilities

As a general rule, you should always have at least two Coupling Facilities. There
are a small number of structures whose loss does not have a significant impact
on the sysplex (OPERLOG and LOGREC are two examples), but in most cases,
the loss of a structure will have some negative or disabling effect on the
connected systems. Therefore, it is vital to always have an alternative CF
available for a structure to rebuild into in the case of a planned or unplanned CF
outage. This is why we recommend that every Parallel Sysplex has at least two
CFs, even if you are only exploiting the Resource Sharing capability of the
sysplex.
Some larger customers, especially those doing extensive exploitation of data

sharing and those that have very high availability requirements, are now starting
to deploy three CFs in their production Parallel Sysplexes. This provides greater
capacity to cope with unexpected workload spikes, and ensures that there is still
no single point of failure, even if one CF is unavailable for some reason.
Note: Whatever number of CFs you have, make sure that all of them are
specified on the PREFLIST for all critical structures.
CF maintenance procedures
There are different approaches to emptying Coupling Facilities.
One method that some customers use is to maintain three sets of CFRM
policies: One that contains both CFs, one that contains just one CF, and one that
contains just the other. To empty a CF, they switch to the policy containing just
the CF that will remain in service, and then rebuild all the structures that now
have a POLICY CHANGE PENDING status.

Another method is to simply issue a SETXCF
START,RB,CFNM=cfname,LOC=OTHER for the CF that is to be emptied. If you
do this, you will need to move any XCF structures out of the CF with individual
SETXCF START,RB commands.
After the work has been completed and the CF LPAR is activated and available to
all systems, always use the SETXCF START,RB,POPCF=cfname command to
repopulate the CF.
Whatever method you use, ensure the following:

򐂰 Maintain a well-documented and understood procedure for emptying the CF
in question and then re-populating it following the changes, as it may be
necessary from time to time to take a CF offline to implement hardware or
microcode changes.
򐂰 Before handing over the CF to the service representative, verify that the CF is
empty. Then configure off the CF links from all systems to the affected CF,
and deactivate that CF LPAR on the Hardware Management Console (HMC).
Direct Access Storage Device (DASD)

In order to protect your data environment, you might consider using DASD
designed for the most demanding, mission-critical environments requiring
extremely high availability, performance, and scalability. Any DASD subsystem
supporting the RAID-5 and RAID-10 architecture can provide the needed
infrastructure for the DASD subsystem.
For example, the IBM DS8000™ series is designed to avoid single points of
failure and provide outstanding availability. With the additional advantages of
IBM FlashCopy®, data availability can be enhanced even further; for instance,
production workloads can continue execution concurrent with data backups.
Metro Mirror and Global Mirror business continuity solutions are designed to
provide the advanced functionality and flexibility needed to tailor a business
continuity environment for almost any recovery point or recovery time objective.
The addition of IBM solution integration packages spanning a variety of
heterogeneous operating environments offers even more cost-effective ways to
implement business continuity solutions.
Business continuity means that business processes and business-critical

applications need to be available at all times. Therefore, it is very important to
have a storage environment that offers resiliency across both planned and
unplanned outages.
The DS8000 supports a rich set of Copy Service functions and management
tools that can be used to build solutions to help meet business continuance
requirements. These include IBM TotalStorage® Resiliency Family Point-in-Time

Copy and Remote Mirror and Copy solutions, which are currently supported by
the Enterprise Storage Server.
FlashCopy can help reduce or eliminate planned outages for critical applications.
FlashCopy is designed to provide the same point-in-time copy capability for
logical volumes on the DS6000™ series and the DS8000 series as FlashCopy
V2 does for ESS, and it allows access to the source data and the copy almost
immediately.
5.2.4 External system connectivity

Another consideration when designing a BASE24-es system for high availability
is the connectivity to external systems such as the card association networks,
regional or international switches, and external host systems. Connectivity loss to
any one of these systems could result in the perception of a system outage or a
failure to meet service level agreements.
BASE24-es overcomes this by simply allowing multiple connections to be

configured to each of the external systems. The interfaces to the external
systems use round robin processing across all available connections, resulting in
load balancing across all connections.
If a connection is lost, the interface will mark the connection as down and will no
longer send messages to it. The interface will periodically check the connection
to see if it has been re-established. Once re-established, the interface will begin
sending messages to the connection.
5.2.5 Back-end authorization system connectivity

When a CICS-based Host Application is available to provide authorization
services to the payment engine, BASE24-es provides a Host Interface that can
perform CICS LINKs to the Host Application. There are two methods used
communicate with such an application. The first is to use distributed links, and
the other is to statically configure the links. We discuss these methods in the
following sections.
Distributed links
In a CICSPlex environment, a customer can achieve high availability by taking
advantage of the CICSPlex Workload Manager (WLM) to distribute the CICS
LINK from BASE24-es across multiple Host Application target regions. In order to
do this, the BASE24-es Host Interface must be configured in Synchronous mode.
There should be one entry configured in the Station Table for the Host Interface.
That entry should correspond to an entry the SYDMF that defines the CICS

program name of the Host Application. The CICS program should be configured
as DYNAMIC, and the CICSPlex workload must be defined accordingly.
Statically configured links

In a CICSPlex or other Multi-Region Operations (MRO) environment, an
alternative approach to high availability is to rely on BASE24-es to distribute the
CICS LINK across the available target regions. To accomplish this, once again,
the BASE24-es Host Interface must be configured in Synchronous mode.
There should multiple entries in the Station Table, each corresponding to an

entry in the SYDMF. There should be one SYDMF entry defining the CICS
program name of the Host Application in each target region, which must be
unique. Each remote program definition in the routing region must be associated
with the corresponding target region.
The BASE24-es Host Interface will utilize the stations in turn, which will provide a
load balancing effect. If a target region is unavailable, BASE24-es Alternate
Routing can be configured to re-route the request to a second or third target
region for authorization.
5.3 BASE24-es configurations

IBM supports many different ways to configure CICS applications. When
configuring CICS applications, however, both technical configuration issues and
business considerations must be taken into account. Among those business
considerations is impact to users, cardholders, and account holders when the
system is unavailable.
In this section we examine the possible options to configure BASE24-es when

running under CICS and z/OS:
򐂰 Single LPAR
򐂰 Multiple LPARs, single CPC
򐂰 Multiple LPARs, multiple CPCs,
򐂰 Multiple LPARs, multiple CPCs, dual site
These options are not meant to be all-encompassing. Instead, they highlight

details to consider when planning a BASE24-es configuration.
Of course, there are many other considerations (cost of hardware, costs

associated with ongoing maintenance and support of the configuration,
requirements of other CICS applications, and so on) to be aware of when
planning such a configuration, but they are beyond the scope of this redbook.

The configuration options we describe progress from the most simplistic to the
more complex. As mentioned, while the technical issues are a consideration
when determining the appropriate configuration for BASE24, the business
requirements cannot be ignored. Customers typically consider the
implementation of an e-payment systems like BASE24-es to be a mission-critical
application, and availability is a very significant business requirement. Therefore,
the primary focus of the options is the availability of BASE24-es to process
transactions
Each option includes a description of the configuration and a discussion of the

single points of failure associated with the option.
5.3.1 Single LPAR

The first option is BASE24-es running in a single LPAR. The diagram in
Figure 5-5 on page 117 shows a typical configuration for this option with one
TOR, and with two AORs utilizing a single set of data files.
The number of both TORs and AORs can be expanded to provide greater
resilience and throughput. This option uses VIPA to provide virtualization of the
data communications.

CPC 1
LPAR 1
CF
CICS Routing Region CF
VIPA
VIPA
MSG Route
VSAM
VSAM
CICS Target Region

T
T
BASE24-es
Figure 5-5 Single LPAR configuration
Single points of failure

This option has several points of failure and is not typically recommended for
production use. This option is better suited for a test environment.
Single CPC
The single CPC is the next point of failure. While a CPC failure is rare, planned
maintenance will cause a system outage for the duration of the maintenance.
Single LPAR
The first point of failure is the single LPAR. If there is a planned or unplanned
outage of the LPAR, the whole system becomes unavailable for the duration of
the outage.
Single copy of data

The final point of failure in this model is the file subsystem. Since there is only
one subsystem accessing the data, a catastrophic failure of the of SMSVSAM

will result in a system outage. In this configuration, usually the data is also kept in
single copy, so a failure on the DASD subsystem will cause an outage for the
duration required to rebuild the VSAM data set. Refer to 5.2.4, “External system
connectivity” on page 114 for a description of how to configure these subsystems
for high availability.
5.3.2 Multiple LPARs/Single CPCs

This option has BASE24-es running in multiple LPARs. Figure 5-6 shows a
typical configuration for this option, with one TOR and two AORs in each LPAR.
The number of both TORs and AORs can be expanded to provide greater
resilience and throughput.
As with the first option, there is only one copy of the BASE24-es data. This option
uses VIPA to provide virtualization of the data communications.
CPC
LPAR CF LPAR
CICS Routing Region 1 CICS Routing Region 2
VIPACF
MSG MSG
Route Route
VIPA
VSAM
VSAM
CICS Target Region CICS Target Region
T T T T
BASE24 BASE24
-es -es
Figure 5-6 Multiple LPARs configuration

This option provides greater availability than the previous option. It has some of
the same points of failure, but nevertheless will satisfy the availability needs of
many BASE24-es customers.
Single CPC
The single CPC can represent a single point of failure. While a CPC failure is
rare, planned maintenance will cause a system outage for the duration of the
maintenance.
Single copy of data

The final point of failure in this model is the file subsystem. Because there is only
one copy of the data, a catastrophic failure of the DASD subsystem will result in a
system outage. Refer to 5.2.4, “External system connectivity” on page 114 for a
description of how to configure these subsystems for high availability.
5.3.3 Multiple LPARs/Multiple CPCs

This option has BASE24-es running in multiple LPARs across multiple CPCs.
Figure 5-7 on page 120 shows a typical configuration for this option, with one
TOR and two AORs in each LPAR, and one LPAR in each CPC. The number of
TORs, AORs, and LPARs can be expanded to provide greater resilience and
throughput.
As with the first two options, there is only one copy of the BASE24-es data. This
option uses VIPA to provide virtualization of the data communications.

CPC
LPAR 1 CF LPAR 2
CICS Routing Region CICS Routing Region
VIPACF
MSG MSG
Route Route
VIPA
VSAM
VSAM
CICS Target Region CICS Target Region
T T T T
BASE24-es BASE24-es
Figure 5-7 Multiple LPARs configuration

This option provides greater availability than the previous options. A
well-maintained and well-operated system using this configuration should be
able to withstand any event short of a catastrophic site failure.
This configuration provides the highest availability that can be achieved at one
site. It satisfies the requirements of many BASE24-es customers.
Single copy of data

The final point of failure in this model is the file subsystem. Because there is only
one copy of the data, use hardware facilities to be able to quickly recovery the
potential loss of data. Also keep in mind that the VSAM files used by the
BASE24-es application do not require forward recovery capability, since they are
easy to rebuild from static data.
5.3.4 Multiple LPARs/Multiple CPCs/Dual site

This option is a dual site model. With this option, one site is active and the other
site is a passive back-up. Unlike the previous options, this option has two copies

of the data. Asynchronous data synchronization tools are used to keep both
copies of the data synchronized.
Figure 5-8 illustrates this configuration. Each site houses a system similar to the
one described in the previous option, with tools to replicate the data. The number
of TORs, AORs, LPARs, and CPCs can be expanded to provide greater
resilience and throughput.
Site A Site B
CPC CPC
LPAR 1 CF LPAR 2 LPAR 1 LPAR 2

CICS Routing Region CICS Routing Region CF
CICS Routing Region CICS Routing Region
VIPACF VIPACF
MSG MSG MSG MSG
Route Route
VIPA Route
VIPA Route
VSAM VSAM
VSAM VSAM
CICS Target Region CICS Target Region CICS Target Region CICS Target Region
T T T T T T
T T
BASE24-es BASE24-es BASE24-es BASE24-es
Replication
Figure 5-8 Dual sites configuration

This option provides the greatest availability because there is no single point of
failure. A well-maintained and well-operated system using this configuration
should be able to withstand any event, including a catastrophic site failure. This
configuration will satisfy the requirements of the most demanding BASE24-es
customers.
For further details on concepts and implementation, refer to GDPS Family - An

Introduction to Concepts and Capabilities, SG24-6374.

6
Chapter 6. Installing BASE24-es on

z/OS
In this chapter, we describe our target configuration environment. We explain the
hardware/software requirements for BASE24-es, and describe the
pre-installation steps needed for a successful installation of BASE24-es on z/OS.
The chapter covers the following topics:

򐂰 The ITSO environment
򐂰 BASE24-es hardware and software requirements
򐂰 An overview of the pre-installation steps
򐂰 Installation considerations
򐂰 How to set up integration with the back-end system

6.1 The ITSO environment
Our hardware setup complied with the requirement for maximum availability of
the BASE24-es application on z/OS. We used a POS device simulator as a
device-acquiring endpoint to acquire the transaction from the POS, and a Visa
DPS networks simulator for acquiring transactions from Visa Debit Processing
Service (DPS). IP protocol was used for communication. We did not test
communication availability in this environment, but we used a Sysplex Distributor
for connection balancing and availability. The primary goal in the environment
was to provide high availability to the solution.
6.1.1 Hardware we used

Figure 6-1 on page 125 summarizes our hardware testing environment. The
environment consisted of three LPARs: two running on the same Central
Processor Complex (CPC), and one running on a different CPC within the same
Parallel Sysplex. On each LPAR, there was a Terminal Owning Region (TOR)
running with three cloned Application Owning Regions (AORs).
CICSplex System Manager (CP/SM) and BASE24-es were used to provide the
dynamic routing capability from each TOR to all the AORs.
All the AORs were capable of accessing the VSAM files in Record Level Sharing
(RLS) mode through the SMSVSAM address space. SMSVSAM allocates the
cache and lock structure in the two Coupling Facilities available to the Parallel
Sysplex.
The ACI desktop client was the user interface to BASE24-es.

ACI
Workload
Desktop
Drivers
Client Sysplex
Distributor
CPC CPC
SC30 SC30
SC31 SC32
CPSM
TOR TOR TOR
AOR AOR AOR
SMSVSAM SMSVSAM CF SMSVSAM

CF
VSAM data
Figure 6-1 Our hardware environment
6.1.2 Software we used

We used the following software in this environment:
򐂰 CICS TS Version 3.1
򐂰 z/OS Version 1.7
6.2 BASE24-es requirements

BASE24-es configurations may require communication controllers or switch
routers capable of a tunneling protocol for XOT or DLSw support. ATM acquiring
implementations will need a tool to create and maintain ATM state and screen
configuration files.
Chapter 6. Installing BASE24-es on z/OS 125

6.2.1 Hardware requirements
BASE24-es runs on any System z processor. It requires a Coupling Facility to
support the RLS environment.
6.2.2 Software requirements

The core BASE24-es application runs in one or multiple CICS regions with
Parallel Sysplex and CICSplex for maximum availability. Our recommendation is
to clone it in multiple CICS regions as a minimum to obtain sufficient availability.
The following tables list the minimum software prerequisite products and levels.
Table 6-1 summarizes the required z/OS software components.
Table 6-1 BASE24-es V06.2 application required z/OS software components

Application requirements Version
z/OS 1.6
DFSMS Part of z/OS
Communication Server Part of z/OS
z/OS C/C++ Compiler is required to link the delivered 1.4

software objects into an executable
Language Environment® (LE) Part of z/OS
CICS Transaction Server 2.3
Terminal Emulation Software Client selected
DFSORT™ Any level
JVM™ (32-bit) - optional 1.3.1
WebSphere MQ - optional 5.3.1
6.2.3 Desktop
The user interface to manage the BASE24-es application is controlled by
client/server architecture. The ACI desktop client is the user interface to
BASE24-es. Table 6-2 on page 127 summarizes the ACI desktop client
requirements. The ACI desktop client needs write access to the installation
location for file updates, logs, and configuration file storage.

Table 6-2 User workstation requirements
User Workstation
Processor type (minimum) Intel family or compatible
Processor speed 1GHz or greater
Memory 512 MB RAM
Monitor 1024x768 VGA
Disk space 128 MB (BASE24-es desk top application)
Operating system Choose only one of the following. All current service
packs applied, where applicable:
- Microsoft Windows 2000 Professional
- Microsoft Windows XP Home or Professional
6.3 Overview of pre-installation steps

A successful installation depends on careful pre-installation preparation. In the
following sections we provide information to be aware of and list the steps to
follow before performing the installation.
6.3.1 General considerations

򐂰 Even if you are planning to use a single region CICS solution, Parallel Sysplex
is required to run the BASE24-es application.
򐂰 BASE24-es requires VSAM Record Level Sharing (RLS) access. SMSVSAM
should be customized and active to support the installation. SMSVSAM
requires two structures in the Coupling Facility, IGWLOCK00 and cache. Be
sure you reserve this storage in the Coupling Facility and update your CFRM
policy to reflect the new structures. You also need to update your SMS
configuration to associate the VSAM files to the cache structures.
In our configuration, we defined two cache structures, one in each Coupling
Facility to balance the load.
򐂰 BASE24-es is installed into the HFS file system and PDS files. Make sure the
user ID used for the installation has superuser capability.
򐂰 If security is used (RACF or similar product) ensure that the CICS regions and
the TSO ids accessing the VSAM files have the correct authority.
򐂰 A file management utility is needed for debugging purposes (for example, File
Manager).

򐂰 Initially you can install the BASE24-es product on a single CICS region
functioning as TOR/AOR (although the ultimate configuration would be at
least a TOR on each LPAR routing transactions to cloned AORs that are
running on all LPARs). BASE24-es provides its own routing mechanism, but
we suggest that you use CP/SM for MRO definitions and operation.
CICS customization requires the use of System Logger. Make sure you have
defined the logging environment on the CICS definitions, as well as in the
CFRM and LOGR policies. Though CF logging is not required, it is the
suggested configuration.
Real Time Analysis and Monitoring CP/SM features are not required.
A TCP/IP socket is required, and it needs to be defined in all TOR regions and
in the AOR region that is used by the UI.
򐂰 Set up Automatic Restart Manager (ARM), which is a component of z/OS.
ARM is a sysplex-wide automatic restart mechanism that does the following:
– Restarts a z/OS subsystem in place, or on a different LPAR if it abends
– Restarts all the elements of a workload (for example, TORs and AORs)
򐂰 FTP Server is required for software installation.
򐂰 If Electronic Distribution is used, an FTP tool will be needed from a
LAN-based personal computer to the host system. Both binary and ASCII
data will be transferred. The installation uses the UNIX compress tools TAR
and GZIP; you can download them here:
http://www.ibm.com/servers/eserver/zseries/zos/unix/bpxalty1.html
򐂰 Provisioning of LAN access to all servers must be completed before
installation and IP addresses for servers must be provided.
6.3.2 Installing and implementing the workstation requirements

The onsite installation team's hardware required to install BASE24-es has two
options:
򐂰 The client provides a minimum of three terminals or personal computer
workstations.
򐂰 LAN access for the onsite team's mobile (laptop) computers.
Additionally, the installer will require two TSO ids with ISPF access.
6.3.3 Additional requirements

Test devices (ATMs, POS terminals, modems, interchange simulators,
interchange connections, and so on) must be available and connected prior to

BASE24-es installation. In our environment we used a POS device simulator as a
device-acquiring endpoint to acquire transactions from POS, and a VISA DPS for
acquiring transactions from VISA DPS. Acquirer interfaces are used when the
transaction has been forwarded to BASE24-es by a switch or an interchange.
6.3.4 New time library

The SIS_TZ environment variable was introduced with the 64-bit time library
(BASE24-es version 03.2 onwards). Because not all operating systems that
BASE24-es was running on supported a 64-bit time library at that time, there was
a potential for customers to run into issues related to data processing for dates
later than the year 2038.
Therefore, a 64-bit time library was introduced into the System Interface Services
layer of BASE24-es to circumvent this potential issue. Additionally, an
environment variable to accurately represent time zone offsets based on
POSIX 1.1 standards was introduced: the SIS_TZ environment variable. Some
activities are needed for setting the SIS_TZ environment variable in CICS.
6.4 Installation considerations

We followed the steps described in BASE24-es v06.2 IBM Mainframe Installation
Guide to install the product in our test environment. BASE24-es-specific CICS
consideration instructions are included as part of the product installation steps,
even though the initial CICS setup is not included.
򐂰 The BASE24-es Electronic Software Distribution (ESD) contains the object
code for BASE24-es, sample “make files” to link edit the objects, files used to
load the CICS definitions, and BASE24-es configuration data and JCL to
define the BASE24-es data files. The ESD also contains reference
publications for BASE24-es.
򐂰 A software installation tool utility is provided to help load the BASE24-es files
to the Hierarchical File System (HFS) on z/OS and to install the User Interface
(UI) for BASE24-es onto a properly configured personal computer. The UNIX
System Service (USS) facility for z/OS is the target for loading the BASE24-es
files, as this method simplifies the installation process on different platforms.
򐂰 The host side installation is executed using parameters provided by the tool.
Before running the script, perform a careful review.This script will install
BASE24-es to z/OS. Note that it is possible to restart the script at any step, or
from the beginning if needed. The installation script file will prompt for
verification of default naming conventions and other variables for the system.
The script also performs all links of System Integration Services (SIS) and
Engine Services (ES). Examine the logs and error files for the link.

򐂰 If compile/link edit steps are needed for the Message Delivery Services
(MDS) configuration screens, procedures are provided to be executed.
򐂰 At this point, we were ready to define the files used by BASE24-es. This
function is performed with the IDCAMS utility. The job is supplied with initial
default file size allocations of 100 records. (Note that these sizes may be
sufficient for initial system testing, but are unlikely to suffice in the long term.)
򐂰 CICS resource definitions are performed with the CICS DFHCSDUP utility.
JCL jobs are provided. A sample CICS start procedure is included, but it
needs to be customized for the site standard.
Note: The CICS ESDA space should be set to at least 600 MB for the ES
application. A System Initialization Table (SIT) is provided for sample
parameters. The CICS region is started with cold or initial to cause the
new resource definition to be installed.
򐂰 We logged on to the CICS system and confirmed basic functionality:

– Using CEDA or CEDC (Online RDO access utility)
EXPAND LIST(ESLIST)
ENTER COMMANDS
NAME TYPE LIST
ESFIL GROUP ESLIST
ESJRNL GROUP ESLIST
ESPGM GROUP ESLIST
ESTDQ GROUP ESLIST
ESTXN GROUP ESLIST
SIDALCI GROUP ESLIST
SIDTR GROUP ESLIST
SIGROUP GROUP ESLIST
SITCPIP GROUP ESLIST
– Using CEMT (basic operator controls), “new copy” a sample program as
follows and verify the resulting display:
CEMT I PROG(SIRSTR)
STATUS: RESULTS - OVERTYPE TO MODIFY
Prog(SIRSTR ),Len(0000162032),Le3,Pro,Ena,Pri, ,Ced, NORMAL,
Res(000),Use(0000000000),Any,Uex,Ful,Qua
򐂰 The RSTR transaction must be executed in order to initialize the Data Access
Layer (DAL) for BASE24-es programs to correctly access the VSAM
database.
򐂰 The event log entries are created as the application encounters significant
operational, informational, and error situations. They provide valuable

information sources for problem analysis, and often will confirm significant
operational status changes.
The event log can be viewed with the R300 CICS mapped terminal
transaction. The other primary error and status information sources reside
within the CICS spool output, where the files CEEOUT and CEEMSG
correspond to the C++ streams stdout and stderr.
򐂰 When restarting the CICS region again, examine the CEEOUT CICS job
output file to confirm that the RSTR transaction completed successfully and
that all OLTP memory tables loaded successfully.
At this point, the BASE24-es is installed in a single CICS region; refer to
Figure 6-2.
Figure 6-2 CICS log output showing a successful installation
You can use the provided user interface to verify that the BASE24-es installation
in a simple CICA region has been completed correctly. Figure 6-3 on page 132
displays a sample from the BASE24-es user interface.

Figure 6-3 Sample from BASE24-es user interface
6.4.1 Expanding the BASE24-es installation to a CICSplex

The following sections describe how to expand the configuration from using a
single CICS region to a CICSplex configuration.
The CICS System Definitions (CSDs) generated in the single region installation
were used as input to the CICSplex to provide resource definitions for the various
MASs. The CICS utility DFHCSDUP was used to extract specified groups from
the CSD. A CPSM-provided user exit (EYU9BCSD) formatted the extracted data
into CPSM resource definition records that can be used as input to the Batched
Repository Update Facility (BATCHREP).
CICSplex considerations
򐂰 The starting point is a sample BASE24-es CICSplex using the CICSplex
System Management (CPSM) facility.

򐂰 CICS resources definitions are extracted from the single region CSD and are
prepared to be input into CPSM using the Batched Repository Update Facility
(BATCHREP).
򐂰 Sample resource definitions are reviewed and will be provided for establishing
the basic BASE24-es CICSplex under the CICSplex Business Application
Services (BAS) component of CPSM that is responsible for managing CICS
resource definitions and the initialization process in the CICSplex.
򐂰 The resource definitions are then loaded into the CICSplex using the
BATCHREP facilities.
Note: ACI provides procedures to accomplish these tasks. For further details
on the CICSplex setup, refer to DTR BASE24-es IBM CICSplex Installation
Guide.
6.5 Setting up integration with the back-end system

Our system was integrated with a CICS-based simulator that acted as a
back-end host authorization system. The list of implementation tasks is as
follows:
1. Determine the message format supported by the back-end system.
BASE24-es offers both network message and Host ISO message support.
2. Determine the host type.
BASE24-es can communicate with either an external host (different platform)
or an internal host (same platform) application.
3. Determine the communications protocol.
– BASE24-es can connect to external host system using TCP/IP services.
– BASE24-es offers both synchronous and asynchronous connections to an
internal host system.
4. Determine whether network management is supported.
Network management supports logon, logoff, key exchange, and echo type
messages.
5. Determine the host configuration. In our case:
– We selected the Host ISO 93 message format for this project.
– We simulated an internal host system.

– We used synchronous communications to the internal host utilizing the
CICS LINK API for this project.
• The host simulators were running in separate host-only regions.
• Remote CICS LINK commands were issued to access the host
simulators from the routing regions running BASE24-es.
– Network management was disabled for this test, due to limitations of the
host simulator.
Configuring the host

For this project, we configured our host as follows:
1. Using the BASE24-es User Interface Host Interface Configuration windows,
we added three stations to the Host Interface.
Each station represents a copy of the host simulator running in a separate
CICS region.
Example: Station name was Host_Station_01.
2. You need to supply a BASE24-es Dynamic Destination Map file (DDMF) entry
for each host station configured to BASE24-es.
The service name is the key to this file. It must be unique, and will be used
within BASE24-es as a reference point to a component.
Example: Service name was Host_Station_01. In the BASE24-es user
interface, select the Stations tab from the Host Interface Configuration
window.
3. You need to supply a BASE24-es Synchronous Destination Map file (SYDMF)
entry for each host station configured to BASE24-es.
Example: Host station Host_Station_01 was configured to run host program
SIMLTR01.
4. CICSplex System Manager (CPSM) was used to configure the host simulator
programs for remote execution.
To allow for remote execution, we created three copies of the simulator
program: SIMLTR01, SIMLTR02, and SIMLTR03. They needed to be defined
with the remote attributes for system and name on the program definition for
each simulator program.
Note: A program-in-group definition is also needed for each simulator

program.

5. The BASE24-es User Interface was used to configure the Host
Communication Type on the ISO Host tab on the Network Host window so
messages are synchronous.
6. The BASE24-es User Interface Routing Configuration Destinations window
was used to configure the primary, alternate 1 and alternate 2 destinations in
order to utilize all three configured host stations.

7
Chapter 7. Setting up and running the

initial workload
In this chapter we describe the tasks we performed to set up and run the
workload for our projects. We cover the following topics:
򐂰 Setting up the workload
򐂰 Our ITSO system layout
򐂰 Verifying that the installation is operational
򐂰 Setting up monitoring

7.1 Setting up the workload
In order to configure BASE24-es to run the workload, we had to configure the
following elements:
򐂰 The platform-dependent infrastructure, which is specific to the CICS
implementation
򐂰 The platform-independent business components, which are common to all
implementations
7.1.1 Configuring the BASE24-es infrastructure

We performed these tasks to configure the BASE24-es infrastructure:
򐂰 Configure the TCP/IP communication subsystem and define the end-points
managed by that subsystem in the System Interface Service (SIS) Message
Delivery Service routine files.
򐂰 Configure the BASE24-es Integrated Server in the SIS Message Delivery
Service routing files.
򐂰 Configure the simulators routed to in the same test scenarios as a
CICS-based back-end authorization system.
򐂰 Configure the SIS Workload Management subsystem.
In the following sections, we describe these tasks in more detail.
TCP/IP configuration
1. We needed to determine whether the BASE24-es TCP/IP process would be
acting as a client or a server.
2. We needed to reserve the TPC/IP address and the port number.
3. The BASE24-es TCP/IP Control File (TCPIPCFG) screens allowed us to set
these parameters for all processes:
– Service Name is the key to this file. It must be unique, and will be used
within BASE24-es as a reference point to a component.
Example: Service name Switch_Station_01. In the BASE24-es user
interface, you would select the Stations tab from the Switch Interface
Configuration window. The Service name is then entered as one of the
stations to be used to send and receive message traffic from a Switch.
– Port number that this BASE24-es TCP/IP process will communicate on.

– The Destination is a BASE24-es component that will be used as either the
acquirer or issuer. These values are bound into the BASE24-es Integrated
Server.
– Maximum number of sockets that this BASE24-es TCP/IP task will
support.
– TCP/IP Header Type used by the BASE24-es TCP/IP process.
Configuration allows for two-byte and four-byte headers.
In our project, we used two-byte headers.
– TCP/IP server type that this BASE24-es TCP/IP will act like. Configuration
allows for client and server processes.
In our project, we used both types of processes.
򐂰 The BASE24-es Static Destination Map (SDMF) requires a corresponding
entry for each TCP/IP Control File service defined.
– Service Name is also the key to this file, and must match one defined in
the TCP/IP Control File.
– Transaction ID is the CICS transaction that this BASE24-es TCP/IP
process will use during execution.
– The Destination Type will need to be in sync with the type of BASE24-es
TCP/IP process that is configured in the TCP/IP Control File.
• For TCP/IP client and server processes, the Destination Type will
indicate TCP/IP sockets are being used.
򐂰 The BASE24-es Dynamic Destination Map (DDMF) requires a corresponding
entry for each TCP/IP Control File (TCPIPCFG) destination defined.
– The TCPIPCFG destination is entered into the DDMF Service Name,
which is also the key to the DDMF file.
– Transaction ID is the CICS transaction that this BASE24-es TCP/IP
process will use during execution. This is also the name of a CICS
Transient Data Queue that a TCP/IP client and server process will access.
– The Destination Type will need to be in sync with the type of BASE24-es
TCP/IP process that is configured in the TCP/IP Control File.
򐂰 The BASE24-es SOCKRECS File requires a corresponding entry for each
TCP/IP Control File service that is defined as a TCP/IP client process.
– Port number that this BASE24-es TCP/IP process will communicate on.
– TCP/IP Address that this client process will connect to.
– The Destination will need to be in sync with a BASE24-es TCP/IP process
that is configured in the TCP/IP Control File.
Chapter 7. Setting up and running the initial workload 139

– Disconnect Type is used to determine when a disconnect should be sent
from the TCP/IP client to the TCP/IP server.
In our project, we set configuration to not send disconnects.
– TCP/IP Header Type used by the BASE24-es TCP/IP process.
Configuration allows for two-byte and four-byte headers.
In our project, we used two-byte headers.
Integrated Server (IS) configuration

򐂰 Each BASE24-es Integrated Server transaction has its own configuration.
In our project, we used a single configuration.
򐂰 A BASE24-es Static Destination Map (SDMF) entry is required for each
BASE24-es Integrated Server.
– Service Name is the key to this file, and must be unique.
– Transaction ID is the CICS transaction that this BASE24-es Integrated
Server process will use during execution.
– The Destination Type used for our project was set so that a CICS START
Transaction command will be issued to pass data from a BASE24-es
TCP/IP process to this BASE24-es Integrated Server process.
– The Maximum Servers is the number of concurrent BASE24-es Integrated
Server processes that will run when a CICS region is started.
– The Dynamic Service Flag allows you set the amount of time the
BASE24-es Integrated Server will run, before it automatically stops and
restarts itself.
• In our project, we used the BASE24-es Integrated Servers in long
running mode.
• They start when a CICS region starts, and run the entire time the
region is up without recycling themselves.
򐂰 The BASE24-es Dynamic Destination Map (DDMF) requires an entry for each
CICS transaction configured to execute as a BASE24-es Integrated Server.
– Service Name is also the key to this file. This should represent the
Integrated Server.
– Transaction ID is the CICS transaction that this BASE24-es Integrated
Server process will use during execution.
– The Destination Type will need to be in sync with the communications
method used by this BASE24-es Integrated Server.
• In our project, we set the configuration to start a CICS transaction.
BASE24-es used CICS transaction XDYR for this purpose.

Host configuration
The BASE24-es Dynamic Destination entry is required for each host station
configured to BASE24-es.
򐂰 Service Name is the key to this file. It must be unique, and is be used within
BASE24-es as a reference point to a component.
Example: Service name Host_Station_01. In the BASE24-es User Interface,
you would select the Stations tab from the Host Interface Configuration
window. The Service name is then entered as one of the stations to be used
to send and receive message traffic from this host.
򐂰 Transaction ID is the CICS transaction that a BASE24-es Integrated Server
process will issue a CICS START Transaction command for when sending
data to the host in asynchronous mode.
򐂰 In our project, we used synchronous mode (CICS LINK).
򐂰 The BASE24-es User Interface will be used to configure the Host
Communication Type on the ISO Host tab on the Network Host window so
messages are asynchronous.
Dynamic routing configuration

BASE24-es provides a dynamic transaction routing configuration and user exit
that, when used in combination with CICSPlex System Manager (CPSM), allows
for dynamic routing of financial transactions.
򐂰 During the installation process program DTR6002 will be compiled, then
linked as EYU9WRAM.
򐂰 CICS System Initialization Table parameters are set to use the IBM-supplied
dynamic routing program, EYU9XLOP.
򐂰 A BASE24-es Dynamic Transaction Routing Control File entry is needed for
each CICS Transaction to be dynamically routed.
– Application Key is the key to the file.
In our project, we used the BASE24-es-supplied transaction of XDYR.
This transaction must be defined as dynamic and routable in all target
regions.
– Workload name must match what was defined in CICSPlex System
Manager during the installation process.
This is the value that is used by BASE24-es routing and target regions to
register with CICSPlex System Manager workload when CICS regions are
started.

– The API flag indicates whether the CICSPlex System Manager API is
being used.
In our project, we used the API.
– The Handshake Flag indicates whether the routing regions will perform
handshake operations to all target regions.
– TOR Scope is a list of CICS APPLIDs for the routing regions that will
register with the CICSPlex System Manager TOR Workload.
– AOR Scope is a list of CICS APPLIDs for the target regions that will
register with the CICSPlex System Manager AOR Workload.
7.1.2 Business configuration

The BASE24-es User Interface can be used to manually add the data necessary
for the initial workload.
The BASE24-es installer can build this data ahead of time. By working with the
customer beforehand, the installer can build the necessary data in a project
environment at ACI before bringing it to a customer site.
7.2 Our ITSO system layout

In this section, we list the components we used to set up our ITSO environment:
Performance data
Acquirers
򐂰 Visa DPS interface
򐂰 SPDH POS devices
Issuers
򐂰 BASE24-es Scripting Component
򐂰 ISO 93 Host Simulator
Journal files
򐂰 Ten journal files, each with two alternate indexes.
Merchant
򐂰 Eight thousand merchants, each supporting four terminals.
򐂰 Thirty-two thousand terminals

Prefix
򐂰 One prefix record was defined.
Routing
򐂰 Offline authorization using the BASE24-es Scripting component.
򐂰 Online authorization using the ISO 93 Host Simulator.
Script
򐂰 Offline authorization used a base script with a series of subscripts to
complete the authorization process, and accessed the Card file and the
Usage file in the process.
򐂰 Online authorization used a base script with a series of subscripts to
complete pre-screening of the transactions before being sent to the issuer for
authorization, and accessed the Card file in the process.
Institution
򐂰 One institution record was defined.
File partitions
򐂰 Card
򐂰 3.2 million card records defined.
򐂰 Partition order is card number
򐂰 Two partitions each with 1.6 million card records.
Usage
򐂰 Partition order is card number.
򐂰 Five partitions.
򐂰 This starts out as an empty file.
򐂰 Usage records are added at first use and updated after that.
7.3 Setting up monitoring

The following section describes the tasks you need to perform in order to set up
monitoring.

7.3.1 Routing region
BASE24-es allows you to see how each routing region is distributing financial
transactions across the registered target regions by using the Dynamic
transaction routing AOR Status Menu. Follow these steps:
1. Log into the routing region.
2. Access the Dynamic Transaction Routing Record List.
3. Edit the Dynamic Transaction Routing Application record.
4. At the command line, type: STAT. Then press Enter.
This will display the Dynamic Transaction Routing AOR Status Menu.
– Each registered routing region will appear in the column titled REGN.
– The column titled ALIV will have a value of either NO or YES.
• A value of NO indicates this routing region has not registered and will
not have transactions routed to it.
• A value of YES indicates this routing region has registered and will
have transactions routed to it.
– The column titled SUCCESS- is a numeric count of the number of financial
requests that this routing region has delivered to this target region.
This works for only a single routing region. If more that one routing region
is configured, separate access is required.
7.3.2 Target region

BASE24-es allows you to view the event messages produced by all components
by using the Event File Browser. Follow these steps:
1. Access any routing or target region in your configuration.
2. Access the BASE24-es Event File Browser. Your first access will take you to
the most recent events based on date and time.
7.4 Installation verification

We performed the tasks described in the following sections to verify that the
installation completed successfully and was fully operational.

7.4.1 Simulate single transactions
Before running any workload, verify through the list of tasks described here that
the CICS environment is operational. After that you can run transactions, initially
one at a time. The transactions are from each acquiring end-point and sent to
each issuing end-point.
Initialization
򐂰 Each target region has an initialization program that CICS executes as part of
its PLT process.
򐂰 Check the following items in each target region:
– TCP/IP task
• Use the CICS supplied CEMT command to verify that the expected
TCP/IP tasks are running.
• Use the NETSTAT GATE REPORT DSN 'data set name' command
from the ISPF command shell to verify that TCP/IP processes are in
the correct state. In our configuration, we issued:
NETSTAT REPORT DSN 'tsorwsa.netstat.cicsibm2’
The output data set ‘TSORWSA.NETSTAT.CICSIBM2’ should show the
port associated with the SPDH server process is listening for client
connections; see Example 7-1.
Example 7-1 Output data set showing port listening for client connections
EZZ2350I MVS TCP/IP NETSTAT CS V1R4 TCPIP Name: TCPIP 16:30:25
EZZ2585I User Id Conn Local Socket Foreign Socket State
EZZ2586I ------- ---- ------------ -------------- -----
EZZ2587I CI31IBM2 0033C4EA 0.0.0.0..7708 0.0.0.0..0 Listen
EZZ2587I CI31IBM2 0033C4E7 0.0.0.0..9829 0.0.0.0..0 Listen
EZZ2587I CI31IBM2 0033D558 172.17.4.153..9823 172.17.1.177..3149
Establsh
EZZ2587I CI31IBM3 0033EA4A 0.0.0.0..9850 0.0.0.0..0 Listen
EZZ2587I CI31IBM3 0033EA4B 0.0.0.0..9843 0.0.0.0..0 Listen
EZZ2587I CI31IBM3 0033EA49 0.0.0.0..9853 0.0.0.0..0 Listen
EZZ2587I CI31IBM3 0033EA4C 172.17.4.153..2934 172.21.200.37..1500
Establish

– AOR status
• Use the BASE24-es Dynamic Transaction Routing AOR Status menu
to verify that the target regions have registered.
򐂰 Items to be checked in each routing region:
– Use the BASE24-es Event File Browser to verify that all OLTPs were built
correctly.
– Use the CEMT command to verify that the configured number of
BASE24-es Integrated Server processes are running.
Transaction counts
From the workload driver, verify that the number of requests and the number of
responses match.
Approval codes
From the workload driver, verify that the responses have the proper approval
code.
Timeouts
From the workload driver, verify whether any timeouts have been recorded.
7.4.2 Increasing the workload

To increase the workload we needed to increase the size and number of records
for a few BASE24-es files, as described here.
Cards
The routing method used for our project requires read access to the Card file for
every transaction. To avoid having the Card file become a bottleneck, we
performed these tasks:
򐂰 We increased the number of records in the file. As a consequence, we
needed to increase the VSAM file allocated space by redefining the file.
򐂰 This number will vary, depending on the workload goals. To address this, we
wrote a program to build these records in a sequential file. Then we used IBM
utility programs to load them into the VSAM files.
The card numbers built by this program will need to kept in sync with those
delivered by the workload driver.

򐂰 A single Card file can be bottleneck. In our project, we used the BASE24-es
File Partitioning feature. This allowed us to create many physical files that
appear as one logical file to BASE24-es.
– In our project, we used two partitions to the Card file.
– The BASE24-es User Interface File Partition Configuration window is used
to configure the specific fields and values to partition the Card file with.
– The BASE24-es metadata configuration file CONFCSV will need to be
updated to incorporate any new Card file assign names that are added
during the file partitioning configuration. This is also where the new Card
files are connected to a CICS DDNAME.
– VSAM define statements are needed for any new files added.
– CICSPlex System Manager File Definitions and File in Group entries will
be needed for any files.
Journals
The BASE24-es Journal file can be configured with up to seven alternate
indexes. One alternate index (AIX®) is required for Online Transaction
Processing (OLTP). The other indexes may be optionally configured to provide
for online perusal of the Journal file according to various criteria.
򐂰 We used the required AIX and one optional AIX, which is a typical
configuration.
򐂰 Each financial transaction results in a record being written to the BASE24-es
Journal File. To avoid having this file become a bottleneck, use more than one
journal file.
򐂰 BASE24-es allows you to configure many Journal files per institution. This
allows you to create many physical files that appear as one logical file.
– Use the BASE24-es User Interface Journal Configuration windows to
define the number of journal files necessary to achieve the workload
goals.
– BASE24-es uses a hashing algorithm to determine to which of the
configured Journal files to write a financial transaction record.
Usage
The routing method used for our project requires a read and update access to
the Usage file for every transaction. To avoid having this file become a
bottleneck, we performed these tasks:
򐂰 We increased the number of records in the file. As a consequence, we
needed to increase the VSAM file allocated space by redefining the file.

򐂰 A single usage file can be a bottleneck. For our project, we used the
BASE24-es File Partitioning feature. This allowed us to create many physical
files that appear as one logical file to BASE24-es.
In our project, we used five partitions for the Usage file.
The BASE24-es User Interface File Partition Configuration window is used to

configure the specific fields and values to partition the Usage file with.
The BASE24-es metadata configuration file CONFCSV will need to be updated

to incorporate any new Usage file assign names that are added during the file
partitioning configuration. This is also where the new Card files are connected to
a CICS DDNAME.
VSAM define statements are needed for any new files added.
CICSPlex System Manager File Definitions and File in Group entries will be
needed for any files.
Merchant processing
Merchant processing requires the coordination between merchants and devices.
򐂰 In our project, we identified the number of merchants necessary to meet the
workload goals and the number of terminals each merchant will support.
򐂰 These numbers will vary depending on the workload goals. We addressed
this by writing a program to build these records in a sequential file. Then we
used IBM utility programs to load them into the VSAM files.
The terminal numbers built by this program will need to kept in sync with those
delivered by the workload driver.

8
Chapter 8. Tuning BASE24-es on z/OS

In this chapter we discuss considerations you need to be aware of when
migrating a BASE24-es installation from a single-transaction-at-a-time
development area into a production or performance area, where the financial
transactions per seconds goals will be increased. We cover the following topics:
򐂰 BASE24-es File Partitioning
򐂰 Basic considerations
򐂰 Single files as potential bottlenecks
򐂰 CICS logging impacts
򐂰 RMF Monitor III indicator

8.1 BASE24-es File Partitioning
The BASE24-es File Partitioning feature allows a data source to appear as a
single logical file to the BASE24-es application, while it is really two or more
separate physical files. Note these steps:
򐂰 BASE24-es User Interface File Partition Configuration window:
– Fields tab
• Select a file to partition.
• Select and order one or more fields from the selected file that will
determine the partitioning order.
– Catalog tab
• Enter the range upon which file partitions will be selected. This will be
based upon the fields selected on the Fields tab.
• Each range will need to be associated with a unique combination of
assign name root and an assign name suffix. This is the tie to a
physical file.
򐂰 BASE24-es Metadata configuration
– Metadata is maintained through the MataMan application.
– The CONFCSV file contains the necessary information to tie a data source
assign to a physical VSAM file using the CICS DDName.
– Each range defined on the Catalog tab will need a corresponding entry in
this file.
򐂰 CICS resource definitions
– To define the file as a resource to CICS, the CICSPlex System Manager
interface is used.
– Each range defined on the Catalog tab will require a file definition entry
and a file in Group entry.
8.2 Basic considerations

Here are ways to help you increase overall system performance:
򐂰 Place the VSAM index and data components on separate volumes.
򐂰 Spread the file across multiple volumes.
򐂰 Use multiple channel paths to access the DASD.
򐂰 Caching any of these potentially large files is beneficial.

8.3 Single files as potential bottlenecks
BASE24-es uses VSAM files during the processing of financial transactions.
Depending on the configuration, some files can have one or more access per
financial transaction. In this section we list files that at one time or another have
been identified as bottlenecks (or potential bottlenecks) in a performance
environment. We also describe how to remove the bottleneck.
Card file
򐂰 Access ranges from zero to two reads on every financial transaction,
depending on the authorization method configured.
򐂰 BASE24-es concepts
– The BASE24-es File Partitioning ITSO configuration:
• Fields tab
Catalog name is Card
Field partition order number 1
Field name Personal Account Number (PAN)
• Catalog tab
Range one
Low key range value of “9876500000000001”
Assign name root of “CARD_01”
Assign name suffix of “A”
Range two
Assign name root of “CARD_01”
Assign name suffix of “B”
– BASE24-es Metadata configuration CONFCSV file
• Logical Card file
Assign name of “CARD”
Table name of “file_part_catalog”
Physical name of CRDCAT
• Physical Card file 1
Assign name of “CARD_01A”
Table name of “card”
Chapter 8. Tuning BASE24-es on z/OS 151

Physical name of CRD01A
• Physical Card file 2
Assign name of “CARD_01B”
Table name of “card”
Physical name of CRD01B
Journal file
Each financial transaction results in a record being written to the BASE24-es
Journal file. To avoid having this file become a bottleneck, use more than one
Journal file. Also examine the number of alternate indexes being used.
Each alternate index adds requires additional CPU to process. One way to
reduce CPU consumption in a production or performance environment is to
remove the alternate indexes that are not required to meet a customer’s business
needs. The BASE24-es Journal file comes with seven alternate indexes:
򐂰 Account Key
򐂰 Channel Key
򐂰 Clerk Key
򐂰 Clerk Key2
򐂰 Merchant Key
򐂰 On-line Key
򐂰 PAN Key
In our project, we used the On-line Key and the PAN key.
BASE24-es allows you to configure many Journal files per institution, and it is
possible to view a group of physical files as a logical entity. In our ITSO
environment, we used the BK02 institution for the scenarios.
Use the BASE24-es User Interface Journal Issuer Profile Relationship

Configuration windows to define the journal files necessary to achieve the
workload goals.
򐂰 Primary Journal Profiles:
– JLF_BK02_P00
– JLF_BK02_P01
– JLF_BK02_P02
– JLF_BK02_P03
– JLF_BK02_P04
– JLF_BK02_P05
– JLF_BK02_P06
– JLF_BK02_P07

– JLF_BK02_P08
– JLF_BK02_P09
򐂰 Alternate Journal Profiles:
– JLF_BK02_A00
– JLF_BK02_A01
– JLF_BK02_A02
– JLF_BK02_A03
– JLF_BK02_A04
– JLF_BK02_A05
– JLF_BK02_A06
– JLF_BK02_A07
– JLF_BK02_A08
– JLF_BK02_A09
Use the BASE24-es User Interface Journal Profile Configuration windows to

configure the Journal file assigns. The Journal Profile Assign is associated to an
entry in the CONFCSV file. The CONFCSV entry is associated with the CICS
DDName, the link to the physical file.
򐂰 Primary Journal Profiles:
– JLF_BK02_P00 -
Data Source Assign Name is JLF_BK02_P00_0
CONFCSV
Assign name of “JLF_BK02_P00_0”
Table name of “Journal_key_seq2”
Physical name of P1JR00
– JLF_BK02_P01
Data Source Assign Name is JLF_BK02_P01_1
CONFCSV
Assign name of “JLF_BK02_P01_1”
Table name of “Journal_key_seq2”
Physical name of P1JR01
– All the remaining primary and alternate journal file configurations follow
the same pattern.
BASE24-es uses a hashing algorithm to determine to which of the configured

Journal files to write a financial transaction record.

Merchant Delivery Channel file
Merchant processing requires the coordination between merchants and devices.
򐂰 Identify the number of merchants your installation have.
򐂰 Identify the number of terminals your installation needs to configure.
In our project, we took into account the following performance considerations

when we built the merchant processing environment:
򐂰 8,000 merchants, each supporting four terminals, for a total of 32000
terminals.
򐂰 A batch program was written to build both merchant and terminal records into
separate sequential files. Then IBM utility programs were used to load them
into the VSAM files.
Positive Balance file

This file was not used during the ITSO project.
However, when it is used, then the BASE24-es File Partitioning described for the
Card file or the Usage file also applies to the Positive Balance file.
Usage file
Access is a read and update on every financial transaction, depending on the
authorization method configured.
򐂰 The BASE24-es File Partitioning ITSO configuration:
– Fields Tab
• Catalog name is Usage
• Field partition order number 1
• Field name ID
– Catalog Tab
• Range one
Assign name root of “USGD_ID_00”

• Range two

• Range three

• Range four

• Range five

򐂰 BASE24-es Metadata configuration CONFCSV file
– Logical Usage file
• Assign name of “USAGE”
• Table name of “file_part_catalog”
• Physical name of USGCAT
– Physical Usage file 1
• Assign name of “USGD_ID00A”
• Table name of “Usage”
• Physical name of USGD00

8.4 CICS logging impacts

CICS allows you to configure the percent full of the CICS log file before the file
needs to be trimmed. CICS also allows you to configure what percent the log can
be trimmed back to, when the trimming process begins.
When configured correctly, the trimming process has little or no effect on the
processing of transactions. When configured incorrectly, however, the trimming
process can cause additional overhead. That is, while the trimming process is
executing, additional CPU will be used and the customer will experience
increased response time. For more information about this topic, refer to the
presentation “Logger/CICS - Performance and Common Problems”, which you
can find at the following site:
http://www.ibm.com/support/techdocs/atsmastr.nsf/PubAllNum/PRS183
You can also use the DFH0STAT, SMF type88 record, IBM CICS Performance
Analyzer tools to investigate common CICS log stream problems.
If your installation is suffering error conditions because the configuration of the

parameters AKPFREQ, HIGHOFFLOAD, LOWOFFLOAD and LS_SIZE is not
optimized, refer to the following site for helpful information:
http://www.ibm.com/support/docview.wss?uid=swg21052014
8.5 RMF Monitor III indicator

On the RMF Monitor III Coupling Facility Activity report, the Async Change %
column can be an indicator that CPU utilization increase and response time
increases are possible.
When this value starts to get above 5%, observations have shown an increase in
CPU utilization coupled with an increase in response time.

To address this situation, you might want to investigate further in order to verify
that your Coupling Facility configuration is balanced, and possibly redistribute the
structure to avoid bottlenecks.

9
Chapter 9. Achieving high availability

on z/OS
In this chapter we document our test scenarios. First we describe the objective,
the injection, and the expected behavior for each scenario. Then we describe the
test we performed against each scenario, and detail the actual behavior. In cases
where the expected behavior differs from the actual test behavior, we provide a
solution.
The chapter covers the following topics:

򐂰 CICS, CP/SM
򐂰 Data environment
򐂰 BASE24-es application
򐂰 Hardware
򐂰 z/OS and relevant subsystems

9.1 Test scenarios
High availability is of prime importance to modern online transaction processing
(OLTP) systems. Improvements in modern hardware have made fault tolerance
of less relative importance—the loss of data in the communications networks far
exceeds the loss of information in the OLTP processing system itself, and
end-to-end protocols designed to recover from loss of data in the
communications networks also assure financial integrity during infrequent
hardware failures. However, contractual obligations and end-user requirements
continue to make high availability vitally important.
Our test configuration was designed to represent the highest possible level of
availability attainable at a single production site. All components were replicated
to eliminate any single point of failure. Scenarios were developed to assure that
the OLTP processing system remained available through a wide cross-section of
potential hardware and software failures.
Note: For simplicity, some of the failover scenarios are illustrated by graphics
which show only two LPARs. However, in our configuration we used three LPARs.
9.2 CICS and CP/SM

Here we describe the following CICS and CICSplex scenarios:
1. CMAS failure after workload has been activated
2. TCP/IP client - TOR failure
3. TCP/IP server - TOR failure
4. AOR failure
5. TCP/IP server - TOR maintenance
6. AOR maintenance
7. Dynamically adding a TOR
8. Dynamically adding an AOR
In the following sections we explain these scenarios in detail.
9.2.1 CMAS failure after workload has been activated
Objective and Injection

The objective of this test case was to verify that the workload keeps running even
when a CMAS region fails. To simulate this scenario, we issue the CANCEL
CMASname command from SDSF or the console.

Expected behavior
With the CMAS running on the same LPAR there would be no adverse effect on
either the current workload, or outgoing workload, due to the fact that ACI's
EYU9WRAM in the Routing Region (TOR) uses round robin processing to select
target regions.
Also, this is possible since there are no affinities within the BASE24-es
application. If there were affinity lifetimes of System or Permanent, the workload
would fail.
With the CMAS running on a different LPAR and the Routing Region (TOR)
routing to Target Regions (AOR) on the other LPAR, those Target Regions will
have an additional weight due to the CMAS being unavailable. Workload will still
be routed to the AORs, but CPSM will not be able monitor, operate, analyze or
otherwise manage these AORs.
Actual behavior
The workload continued to be processed without interruption.
The LPAR in which the CMAS was failed produced a series of messages in both
routing and target regions, as shown here:
10.53.44 STC03506 +EYUNL0902I CICSET01 LMAS LRT CMAS Requested termination initiated
10.53.49 STC03506 +EYUNL0999I CICSET01 LMAS LRT termination complete
10.53.49 STC03506 +EYUXL0011I CICSET01 LMAS shutdown in progress
10.53.49 STC03506 +EYUCL0005I CICSET01 ESSS Receive Link Task terminated
10.53.49 STC03506 +EYUXL0018I CICSET01 LMAS restart in progress
.
.
.
10.55.14 STC03506 +EYUCL0006I CICSET01 ESSS link to CMASSC30 established
10.55.14 STC03506 +EYUXL0007I CICSET01 LMAS Phase II initialization complete
10.55.14 STC03506 +EYUNL0099I CICSET01 LMAS LRT initialization complete
The other LPARs showed no signs of the failure.
Solution
The expected behavior was observed.
9.2.2 TCP/IP client - TOR failure

when a TOR region running the client fails. To simulate this scenario, we issue
the CANCEL TORname command from SDSF or the console.
Chapter 9. Achieving high availability on z/OS 161

Expected behavior
1. Responses to messages in flight in the AOR will be reversed when the send
to the TOR fails, and will be reversed.
2. TOR will be restarted by CICS Automatic Restart Manager (ARM).
3. ACI Task Monitor handler will be restarted by first ACI handshake transaction
from an available AOR, and will restart IP handler.
4. IP Handler will reconnect to the simulator. Message traffic will resume.
5. If stale messages on outbound TDQ are still accessible, IP handler will read
the TDQ and forward messages to the simulator.
6. Messages actually being processed by the IP handler at the time the TOR
was cancelled will be lost, as are messages on the outbound TDQ that could
not be recovered at region restart. An out-of-balance condition exists where
the host has authorized the transaction, but the acquirer has not received the
approval.
In the real world we would rely on end-to-end recovery. Though the simulator
does not support this functionality, a typical real world interface might
generate a reversal on the time-out, then stand in and generate a notification
of the stand-in approval.
Actual behavior
At 100 transactions per seconds (TPS), three routing regions were each
processing one-third of the workload. One of the routing regions was cancelled.
At this point the workload driver recognized one of its client connections was
down, and the workload was reduced by one-third. After the routing region
automatically recovered, the client connection was reestablished and the full
workload was resumed.
The routing region that was purged was restarted with a new started task number
by the Automatic Restart Manager. The routing region was restarted by ARM in
less than one second. The routing regions were fully initialized and the client
connection was established in 30 seconds, at which time the full workload was
resumed. During this period, the workload driver recorded 7 requests as timed
out.
Following are the system log messages showing when the cancel command was
issued and the ARM start command was issued:
0000000 SC30 2006143 10:32:02.27 ACIRWSA 00000290 C CICSET01,ARMRESTART
0000000 SC30 2006143 10:32:02.55 STC04052 00000090 IEF450I CICSET01 CICS -
ABEND=S222 U0000 REASON=00000000
.
.
.

4000000 SC30 2006143 10:32:02.64 00000290 IXC813I JOBNAME CICSET01,
ELEMENT SYSCICS_CICSET01 852
852 00000290 WAS RESTARTED WITH THE FOLLOWING
START TEXT:
852 00000290 S CICSET01
852 00000290 THE RESTART METHOD USED WAS
DETERMINED BY THE ACTIVE POLICY.
Following is the BASE23-es event message showing the IP client process has
been started:
06-05-2310:32:32TCPSERVNACI.8000000.10000 I1000 ET01T710 0
TCP/IP Main, long term task starting*
Solution
9.2.3 TCP/IP server - TOR failure

when a TOR region running a TCP/IP server fails. To simulate this scenario, we
issue the CANCEL TORname command either from SDSF or the console.
Expected behavior
1. Responses to messages in flight in the AOR will be reversed when the send
to the TOR fails.
2. Simulator will reconnect to virtual IP and be routed by Sysplex Distributor to
remaining IP handlers. Message traffic will resume.
3. The TOR will be restarted by CICS Automatic Restart Manager (ARM). ACI
Task Monitor handler will be restarted by first ACI handshake transaction from
an available AOR, and will restart IP handler.
4. If messages on outbound TDQ are still accessible, IP handler will read TDQ
and fail messages back to the IS process. The IS process will generate
reversal transactions. Verify queue empty with CECI and reversal transactions
with journal perusal. Subsequent connections will be distributed by VIP to all
available TORs.
5. Messages actually being processed by the IP handler at the time it was
cancelled will be lost, as are messages on the outbound TDQ that could not
be recovered at region restart. An out-of-balance condition exists where the

host has authorized the transaction, but the acquirer has not received the
approval.
In the real world we would rely on end-to-end recovery. Though the simulator
does not support this functionality, a typical real world interface might
generate a reversal on the time-out, then stand in and generate a notification
of the stand-in approval.
6. Rebalancing of connections across the TORs after recovery is an open issue.
It may not be necessary.
Actual behavior
At 100 transactions per second (TPS), three routing regions were each running,
processing one-third of the workload. One of the routing regions was cancelled.
At this point the workload driver issues a new client connection request and the
connection was established to one of the remaining IP servers. The workload
driver showed no interruption to the workload.
The routing region that was purged was restarted with a new started task number
by the Automatic Restart Manager. The routing region was restarted by ARM in
less than one second. The routing regions were fully initialized, and the IP server
process was available for new client connections in 20 seconds.
Following are system log messages showing when the cancel command was
0000000 SC30 2006143 11:10:31.53 ACIRWSA 00000290 C CICSET01,ARMRESTART
0000000 SC30 2006143 11:10:31.78 STC04072 00000090 IEF450I CICSET01 CICS -
.
.
.
0000000 SC30 2006143 11:10:31.87 00000290 IXC813I JOBNAME CICSET01,
ELEMENT SYSCICS_CICSET01 349
START TEXT:
349 00000290 S CICSET01
Following is a BASE24-es event message showing the IP client process had

been started:

Solution
The expected behavior was observed. To achieve this behavior, you need to have
a restarting mechanism in place for the TOR region. We accomplished that by
using ARM.
9.2.4 AOR failure

when a AOR region fails. To simulate this scenario, we issue the CANCEL AORname
command either from SDSF or the console.
Expected behavior
1. The TOR start of XDYR in the failed AOR will fail. The TORs will detect failure,
remove the failed AOR from their active list, and reroute the message to an
active AOR.
2. AOR will be restarted by CICS ARM.
3. A handshake program in AOR will register with the active TORs, and traffic
will be distributed across all active AORs.
4. Any message actually being processed by the IS processes in the AOR at the
time of the failure will be lost. Any associated changes will be backed out of
the database. No out-of-balance condition exists since the transaction has
been backed out, and the acquirer has not received an approval.
Though the simulator does not support this functionality, a typical real world
interface might generate a reversal on the time-out, then stand in and
generate a notification of the stand-in approval.
5. Messages on the IS inbound TDQ may or may not be lost.
6. Any messages that cannot be recovered from the IS inbound TDQ will be lost.
No out-of-balance condition exists since the transaction has been backed out,
and the acquirer has not received an approval.
7. Any messages that can be recovered from the TDQ will be processed. The
database will be updated, and responses will be sent to the acquirer. Most of
these responses will almost certainly be stale.
Though the simulator does not support this, in the real world the acquirer
would generate a late-response reversal to bring things back into balance.

Actual behavior
There were nine target regions sharing the workload. One of the target regions
was canceled. At this point the BASE24-es user-supplied dynamic routing exit
detected that AOR was no longer available, and removed it from the list of
available AORs. Workload continued to flow to the remaining AORs. After the
AOR is registered with the BASE24-es user-supplied dynamic routing exit,
transactions would again be routed to it.
The workload driver detected no interruption in workflow. In this case no

messages were lost. The target region that was cancelled was restarted with a
new started task number by the Automatic Restart Manager. The target region
was restarted by ARM in less than one second. The routing region was fully
initialized and registered with the BASE24-es dynamic routing exit in 30 seconds.
Following are the system log messages showing when the cancel command was
0000000 SC30 2006143 11:37:02.18 ACIRWSA 00000290 C CICSEA03,ARMRESTART
0000000 SC30 2006143 11:37:03.21 STC04099 00000090 IEF450I CICSEA03 CICS -
.
.
.
0000000 SC30 2006143 11:37:03.32 00000290 IXC813I JOBNAME CICSEA03,
ELEMENT SYSCICS_CICSEA03 200
START TEXT:
200 00000290 S CICSEA03
Following is the BASE24-es Dynamic Routing AOR Status Display showing the
failed AOR EA03 back in alive status:
REGN ALIV GLOB SUCCESS- ABEND--- ROUTERR-

EA01 YES NO 00000581 00000000 00000000
EA11 YES NO 00000581 00000000 00000000
EA21 YES NO 00000580 00000000 00000000
EA02 YES NO 00000580 00000000 00000000
EA12 YES NO 00000580 00000000 00000000
EA22 YES NO 00000580 00000000 00000000
EA03 YES NO 00000467 00000000 00000000
EA13 YES NO 00000580 00000000 00000000
EA23 YES NO 00000580 00000000 00000000

Solution
The expected behavior was observed. To achieve this behavior, you need to have
a restarting mechanism in place for the AOR region. We accomplished that by
using ARM.
9.2.5 TCP/IP server - TOR maintenance

when maintenance needs to be applied to a TOR. To simulate this scenario, we
issue the SIQM STOP ALL command from the CICS region.
Expected behavior
1. The ACI IP handlers will get termination notifications.
2. The IP handlers will close their sockets
3. Any messages on the outbound TDQs will be returned to the originators for
appropriate handling (for example, approved responses on the queue will be
backed out; financial advice will be placed in a store-and-forward file; and so
on).
4. After monitoring its associated TDQ for a few seconds, the IP handler will
terminate.
Actual behavior
Three routing regions were each running, processing one-third of the workload.
In one of the routing regions the IP server process was stopped. At this point the
workload driver issued a new client connection request, and the connection was
established to one of the remaining IP servers. The workload driver showed no
interruption to the workload.
Following are the tasks which were active before the command to stop the IP
process was issued. Task T701 is the BASE24-es IP server process:
Tas(0000026) Tra(CONL) Sus Tas Pri( 255 )
Sta(U ) Use(CICSUSER) Uow(BED98A8FA2977854)
Tas(0000028) Tra(COI0) Sus Tas Pri( 255 )
Sta(U ) Use(CICSUSER) Uow(BED98A911AFB4996) Hty(USERWAIT)
Tas(0000034) Tra(COIE) Sus Tas Pri( 255 )
Sta(U ) Use(CICSUSER) Uow(BED98A9C4A348494)
Tas(0000047) Tra(CSKL) Sus Tas Pri( 255 )
Sta(S ) Use(CICSUSER) Uow(BED98A9E80156856)
Tas(0000065) Tra(T701) Sus Tas Pri( 250 )
Sta(S ) Use(CICSUSER) Uow(BED98AA7CABAC8C4)

Following are the tasks which were active after the stop command completed
processing. Notice the IP task T701 is no longer active:
Tas(0000026) Tra(CONL) Sus Tas Pri( 255 )
Sta(U ) Use(CICSUSER) Uow(BED98A8FA2977854)
Tas(0000028) Tra(COI0) Sus Tas Pri( 255 )
Sta(U ) Use(CICSUSER) Uow(BED98A911AFB4996) Hty(USERWAIT)
Tas(0000034) Tra(COIE) Sus Tas Pri( 255 )
Sta(U ) Use(CICSUSER) Uow(BED98A9C4A348494)
Tas(0000047) Tra(CSKL) Sus Tas Pri( 255 )
Sta(S ) Use(CICSUSER) Uow(BED98A9E80156856)
The following shows the number of items in the transient data queue that the IP
server process uses when sending responses back to connected clients. The
stop command was issued in CICS system CICSET01.
Notice that the column Number Items is showing zero for CICS System
CICSET01, indicating that messages on the queue were returned to the
application. The other CICS systems have numbers greater than zero under this
column, showing they are still processing.
CMD Queue CICS Enabled Accesses ATI ATI Trigger Number Recovery
--- ID--- System-- Status--- -------- Tran Term Level--- Items-- Status-----
T701 CICSET01 ENABLED 4257 1 0 NOTRECOVABL
Solution
9.2.6 AOR maintenance

when maintenance needs to be applied to a AOR. To simulate this scenario, we
issue the SIQM STOP ALL command from the CICS region.
Expected behavior
1. The ACI WLM subsystem is invoked to unregister the AOR with the TORs.
New work will stop flowing from the TORs to the AOR.
2. The transaction will monitor the state of the ACI TDQs in the AOR. When all
queues are empty (which will probably take a couple of seconds or less under
normal circumstances), the transaction will notify the long-lived ACI
transactions to terminate.

Actual behavior
Nine target regions were configured in the BASE24-es dynamic routing
configuration and were active. One target region was being removed from active
status. Workload continued uninterrupted. The inactive target region was later
brought back into active status.
Following is the BASE24-es dynamic routing configuration before the removal of

target region EA03.
Note: The alive status of region EA03 is YES, indicating it is participating in

financial transaction processing. Also, verify the count in the SUCCESS-
column; this shows the number of requests sent to each region.
Application key: XDYR TOR:

EA01 YES NO 00000154 00000000 00000000
EA11 YES NO 00000153 00000000 00000000
EA21 YES NO 00000153 00000000 00000000
EA02 YES NO 00000153 00000000 00000000
EA12 YES NO 00000154 00000000 00000000
EA22 YES NO 00000154 00000000 00000000
EA03 YES NO 00000154 00000000 00000000
EA13 YES NO 00000154 00000000 00000000
EA23 YES NO 00000154 00000000 00000000
Following is the BASE24-es dynamic routing configuration after removal of target

region EA03.
Note: The alive status of region EA03 is NO, indicating it is not participating in
financial transaction processing. Also note that the number of requests sent to
region EA03 have halted.

EA01 YES NO 00000664 00000000 00000000
EA11 YES NO 00000664 00000000 00000000
EA21 YES NO 00000663 00000000 00000000
EA02 YES NO 00000663 00000000 00000000
EA12 YES NO 00000664 00000000 00000000
EA22 YES NO 00000664 00000000 00000000
EA03 NO NO 00000313 00000000 00000000
EA13 YES NO 00000664 00000000 00000000

EA23 YES NO 00000664 00000000 00000000
Following is the BASE24-es dynamic routing configuration after reactivation of

target region EA03.
Note: The alive status of EA03 is YES. Also note the number of requests sent
to region EA03, as it is again increasing.

EA01 YES NO 00001135 00000000 00000000
EA11 YES NO 00001135 00000000 00000000
EA21 YES NO 00001134 00000000 00000000
EA02 YES NO 00001134 00000000 00000000
EA12 YES NO 00001135 00000000 00000000
EA22 YES NO 00001135 00000000 00000000
EA03 NO NO 00000627 00000000 00000000
EA13 YES NO 00001135 00000000 00000000
EA23 YES NO 00001135 00000000 00000000
Solution
9.2.7 Dynamically adding a TOR

To add TOR capacity to a running system, follow these steps:
1. Add the new TOR to the routing control file using the Dynamic Transaction
Routing configuration screen (R311 transaction).
2. Reload the routing control file into memory. From a clear CICS screen in each
AOR, enter: R001 LONE,XDYR (assuming XDYR is your routing transaction).
3. Bring up the TOR.
Expected behavior
The expectation is to see the existing AOR regions register with the new TOR as
soon as the region is initialized. Consequently, the workload should be spread
evenly across the three TOR regions.

Actual behavior
Two routing regions were configured in the BASE24-es dynamic routing
configuration and were active with connections to the workload driver. The
workload driver had three IP servers available. BASE24-es had two IP clients
connected to the workload driver. The workload driver was at two-thirds capacity.
A third routing region was added. The IP client in the new routing region
connected to the workload driver, and the full workload was driven across all
routing and target regions.
Following is the BASE24-es dynamic routing configuration before the addition of

the third routing region:
Application key: XDYR ("****" for default application record)
Appl name: ESXDYR__ (application name)
Workload name: WLSPES1_ (CPSM workload name)
API Flag: Y (Y/N)
Default seq flg: Y (Y/N - whether to perform default seq)
Handshake flag: Y (Y/N - whether to perform handshake trn)
Initial trans: ________________
TOR Scope: ET01 ET11 ____ ____ ____ ____ ____ ____ ____ ____ ____ ____
____ ____ ____ ____ ____ ____ ____ ____
AOR Scope: EA01 EA11 EA21 EA02 EA12 EA22 EA03 EA13 EA23 ____ ____ ____
Note: There are only two regions in the TOR scope.
Following is the BASE24-es dynamic routing configuration after the addition of

the third routing region:
API Flag: Y (Y/N)
Initial trans: ________________
TOR Scope: ET01 ET11 ET21 ____ ____ ____ ____ ____ ____ ____ ____ ____
____ ____ ____ ____ ____ ____ ____ ____
Note: There are only three regions in the TOR scope.
Following is the BASE24-es dynamic routing AOR status; this is from the third
routing region that was added. It shows how it has distributed financial

transactions to all nine available target regions, as they register with the new
routing region:

EA01 YES NO 00000117 00000000 00000000
EA11 YES NO 00000105 00000000 00000000
EA21 YES NO 00000088 00000000 00000000
EA02 YES NO 00000119 00000000 00000000
EA12 YES NO 00000106 00000000 00000000
EA22 YES NO 00000088 00000000 00000000
EA03 YES NO 00000422 00000000 00000000
EA13 YES NO 00000101 00000000 00000000
EA23 YES NO 00000280 00000000 00000000
Solution
9.2.8 Dynamically adding an AOR

1. Add the new AOR to the routing control file using the Dynamic Transaction
Routing configuration screen (R311 transaction).
2. Reload the routing control file into memory. From a clear CICS screen in each
TOR, enter: R001 LONE,XDYR (assuming XDYR is your routing transaction)
3. Bring up the AOR.
Expected behavior
The new AOR should automatically register with the TORs.
Actual behavior
Eight target regions were configured in the BASE24-es dynamic routing
configuration and were active. A ninth target region was added. The new target
region registered the BASE24-es dynamic routing configuration and became
active to begin processing financial transactions. Workload was uninterrupted
during the process.
Following is the BASE24-es dynamic routing configuration before the addition of

the ninth target region:

API Flag: Y (Y/N)
Initial trans: ________________
TOR Scope: ET01 ET11 ET21 ____ ____ ____ ____ ____ ____ ____ ____ ____
____ ____ ____ ____ ____ ____ ____ ____
AOR Scope: EA01 EA11 EA21 EA02 EA12 EA22 EA03 EA13 ____ ____ ____ ____
____ ____ ____ ____ ____ ____ ____ ____
Note: The AOR scope does not include region EA23.
Following is the BASE24-es dynamic routing AOR status:


EA01 YES NO 00000088 00000000 00000000
EA11 YES NO 00000088 00000000 00000000
EA21 YES NO 00000088 00000000 00000000
EA02 YES NO 00000088 00000000 00000000
EA12 YES NO 00000088 00000000 00000000
EA22 YES NO 00000088 00000000 00000000
EA03 YES NO 00000087 00000000 00000000
EA13 YES NO 00000087 00000000 00000000
Note: The AOR column does not include region EA23.
Following is the BASE24-es dynamic routing configuration after the addition of

the ninth target region:
API Flag: Y (Y/N)
Initial trans: ________________
TOR Scope: ET01 ET11 ET21 ____ ____ ____ ____ ____ ____ ____ ____ ____
____ ____ ____ ____ ____ ____ ____ ____
____ ____ ____ ____ ____ ____ ____ ____
Note: The AOR scope includes region EA23.

Following is the BASE24-es dynamic routing AOR status:

EA01 YES NO 00000664 00000000 00000000
EA11 YES NO 00000672 00000000 00000000
EA21 YES NO 00001084 00000000 00000000
EA02 YES NO 00000664 00000000 00000000
EA12 YES NO 00000666 00000000 00000000
EA22 YES NO 00000933 00000000 00000000
EA03 YES NO 00000663 00000000 00000000
EA13 YES NO 00000662 00000000 00000000
EA23 YES NO 00000080 00000000 00000000
Note: The AOR column include regions EA23, and transactions are being
successfully routed to it.
Solution
9.3 Data environment

Here we describe the data environment scenarios:
1. SMSVSAM failure
2. Lock structure full - IGWLOCK00
3. R/O data table at initial time
4. R/O flat file at initial time
9.3.1 SMSVSAM failure

The objective of this scenario was to verify that the workload is still running even
when an SMSVSAM failure occurs, and to document if any recovery actions
need to be taken. To simulate the failure, we issue the V
SMS,SMSVSAM,TERMINATESERVER command.

Expected behavior
1. AORs that have an active workload will receive an AFCR abend when they try
to continue RLS requests. New workloads will receive an AFCS abend until
the SMSVSAM Address Space is restarted. SMSVSAM will start if
RLSINIT(Yes) has been set in IGDSMSxx parmlib member.
2. After SMSVSAM is restarted, the AORs will reopen the ACB as shown in the
AOR log messages:
DFHFC0153 The previous instance of the SMSVSAM server has failed.
File control RLS access is being closed down.
DFHFC0568I File control dynamic RLS restart has started.
DFHFC0562 The RLS control ACB has been successfully registered by

CICS.
DFHFC0570 File control RLS access has been enabled.
DFHFC0569I File control dynamic RLS restart has ended.

3. AORs can now reopen the RLS data sets and accept workload.
Actual behavior
Three LPARs were running, processing the workload, equally distributed. In one
of the LPARs, a force SMSVSAM command was issued. SMSVSAM was then
restarted automatically because of the RLSINIT(YES) definition in the IGDSMS
parmlib member.
While SMSVSAM was unavailable, all messages sent to this LPAR for processing
were queued. After SMSVSAM became available, the queued messages were
processed.
While SMSVSAM was unavailable, the IP client process was unavailable, which
was notified by the workload driver and the workload was reduced by one third.
After SMSVSAM became available, the IP client process connected to the
workload server and the full workload was again processed.
The workload drivers reported that 103 transactions timed-out during the time it
took to fully recover from this failure.
The following system log messages show SMSVSAM being closed and
restarting in one second, and the restart ending 10 seconds later. No BASE24-es
event messages were generated.
4020000 SC30 2006144 09:31:31.76 STC04531 00000090 +DFHFC0153 CICSET01 412

412 00000090 The previous instance
of the SMSVSAM server has failed. File control
412 00000090 RLS access is being
closed down.
.
.
.
8000000 SC30 2006144 09:31:32.91 00000090 *IGW415I SMSVSAM SERVER
ADDRESS SPACE HAS FAILED AND IS RESTARTING 996
.
.
.
4020000 SC30 2006144 09:31:42.12 STC04533 00000090 +DFHFC0569I CICSEA03
File control dynamic RLS restart has ended.
Solution
The expected behavior was observed. In order to automatically have SMSVSAM
restart after a failure, we specified the RLSINIT(YES) keyword in the IGDSMSxx
parmlib member.
9.3.2 Lock structure full - IGWLOCK00

This test measured the resilience of the system when the sysplex Coupling
Facility IGWLOCK00 structure fills.
A simple test driver program is written to retail locks within the Coupling Facility
lock structure. When the structure has only a few entries remaining, introduce
workload to the system to use the remaining entries as BASE24-es modifies its
recoverable RLS files.
Expected behavior
1. Syslog will show the following messages for the AOR:
DFHRM0205 An activity keypoint has been successfully taken.
DFHLG0760 Log stream xxx.xxx not trimmed by keypoint processing.

Number of keypoints since last trim occurred: 1. History point held
by transaction: yyyy task number: zzzzz.
DFHDU0203I A transaction dump was taken for dumpcode: AFDH,

DFHFC4701 Backout failed for transaction yyyy, VSAM file xxxxxxx,
unit of work X'BC98C1CDD57F7F08', failure code X'C4'.

2. Displaying the structure will show output that is similar to the following:
D XCF,STR,STRNAME=IGWLOCK00
IXC360I 19.49.56 DISPLAY XCF 130
STRNAME: IGWLOCK00
STATUS: ALLOCATED
TYPE: LOCK
POLICY INFORMATION:
POLICY SIZE : 8192 K
POLICY INITSIZE: N/A
SCROLL ===> CSR
POLICY MINSIZE : 0 K
FULLTHRESHOLD : 80
ALLOWAUTOALT : NO
REBUILD PERCENT: N/A
DUPLEX : DISABLED
PREFERENCE LIST: P1CF02
ENFORCEORDER : NO
EXCLUSION LIST IS EMPTY
ACTIVE STRUCTURE
----------------
ALLOCATION TIME: 11/14/2004 16:38:57
CFNAME : P1CF02
COUPLING FACILITY: 009672.IBM.51.000000061147
PARTITION: 0D CPCID: 00
ACTUAL SIZE : 8192 K
STORAGE INCREMENT SIZE: 256 K
ENTRIES: IN-USE: 23597 TOTAL: 23597, 100% FULL
LOCKS: TOTAL: 1048576
PHYSICAL VERSION: BC1EAD96 11947D49
LOGICAL VERSION: BC1EAD96 11947D49
Note: The CFRM POLICY for this lock structure has a maxsize of
8192 K, and the structure has reached this maxsize.
3. Rebuild the lock structure to a larger size to allow backout to continue.
In our scenario, the lock structure was already at the maximum size allowed in
the Coupling Facility Resource Manager (CFRM) policy. In such a case, you
have to perform these additional steps:
a. Define a CFRM policy that has a larger IGWLOCK00 structure size.

b. Make the new CFRM policy the active policy by issuing the following
command:
SETXCF START,POLICY,TYPE=CFRM,POLNAME=config3
c. Perform the Rebuild of the IGWLOCK00 structure with either of the
following commands:
SETXCF START,REBUILD,STRNAME=IGWLOCK00
SETXCF START,ALTER,STRNAME=IGWLOCK00,SIZE=xxx
d. To reinitiate the backout process, issue the following command:
CEMT SET DSN(your.datasetname) RETRY
If you are already prepared for a larger lock structure through your CFRM
policy, simply issue the rebuild command.
Actual behavior
Lock structure IGWLOCK00 was preloaded so that it was approximately 95% full.
Then workload was applied to increase the structure to 99% full.
At this point each financial transaction produces an AFDH abend. Each of these
abends produces a dump, so watch the spool so it does not become full. An alter
command was issued against the structure to give it more space. This changed
the structure from 99% full to 27% full. Financial transactions were then
processed normally, and the AFDH abends no longer occurred.

Example 9-1 System log messages - lock structure 0% full
SC30 2006145 14:17:45.34 -D XCF,STR,STRNAME=IGWLOCK00
SC30 2006145 14:17:45.40 IXC360I 14.17.45 DISPLAY XCF 155
STRNAME: IGWLOCK00
STATUS: ALLOCATED
TYPE: LOCK
POLICY INFORMATION:
POLICY INITSIZE: 40000 K
FULLTHRESHOLD : 80
ALLOWAUTOALT : NO
DUPLEX : DISABLED
PREFERENCE LIST: CF37 CF38 CF39
ENFORCEORDER : NO
ACTIVE STRUCTURE
----------------
ALLOCATION TIME: 05/25/2006 14:17:33
CFNAME : CF37
COUPLING FACILITY: 002084.IBM.02.000000026A3A
ENTRIES: IN-USE: 0 TOTAL: 73739, 0% FULL
PHYSICAL VERSION: BEDB1266 5BAEB999
LOGICAL VERSION: BEDB1266 5BAEB999
SYSTEM-MANAGED PROCESS LEVEL: 8
XCF GRPNAME : IXCLO00B
DISPOSITION : KEEP
ACCESS TIME : NOLIMIT
NUMBER OF RECORD DATA LISTS PER CONNECTION: 16
MAX CONNECTIONS: 4
# CONNECTIONS : 3
CONNECTION NAME ID VERSION SYSNAME JOBNAME ASID STATE

---------------- -- -------- -------- -------- ----
----------------
SC30 01 00010007 SC30 SMSVSAM 0049 ACTIVE
SC31 03 0003000E SC31 SMSVSAM 004E ACTIVE
SC32 02 0002000B SC32 SMSVSAM 0039 ACTIVE
Example 9-2 System log messages - DFHLOG has not been trimmed
DFHLG0760 05/25/2006 14:29:29 CICSEA01 Log stream SYSPROG.CICSEA01.DFHLOG not trimmed by
keypoint processing. Number of
keypoints since last trim occurred: 6. History point held by transaction: WMIL, task
number: 00299.
DFHRM0205 05/25/2006 14:29:38 CICSEA01 An activity keypoint has been successfully taken.

number: 00299.
number: 00299.
number: 00299.
Example 9-3 System log - AFDH abend during financial transaction processing - lock structure full
DFHDU0203I 05/25/2006 14:32:55 CICSEA01 A transaction dump was taken for dumpcode: AFDH,
Dumpid: 10/0001.
Dumpid: 10/0007.
Dumpid: 10/0006.
Dumpid: 10/0005.
Dumpid: 10/0004.
Dumpid: 10/0003.
Dumpid: 10/0002.
Example 9-4 CICS log - AFDH abend during financial transaction processing - lock structure full
DFHAC2236 05/25/2006 14:32:55 CICSEA01 Transaction IS01 abend AFDH in program ISLO term ????.
Updates to local
recoverable resources will be backed out.
Updates to local
Updates to local
Updates to local

Updates to local
14.33.56 STC05051 +DFHFC4701 CICSEA01 028

028 05/25/2006 14:33:56 CICSEA01 Backout failed for transaction IS01,
028 VSAM file USGD04, unit of work X'BEDB160CB5E4C446', task 01012, base
028 ACIRLS.USGD04, path ACIRLS.USGD04, failure code X'C4'.
14.36.55 STC05051 +DFHFC4701 CICSEA01 924
924 05/25/2006 14:36:55 CICSEA01 Backout failed for transaction IS01,
924 VSAM file USGD04, unit of work X'BEDB16BA41E77C16', task 02596, base
924 ACIRLS.USGD04, path ACIRLS.USGD04, failure code X'C4'.
Example 9-5 System log - lock structure 99% full

STRNAME: IGWLOCK00
STATUS: ALLOCATED
TYPE: LOCK
POLICY INFORMATION:
FULLTHRESHOLD : 80
ALLOWAUTOALT : NO
DUPLEX : DISABLED
ENFORCEORDER : NO
ACTIVE STRUCTURE
----------------
ALLOCATION TIME: 05/25/2006 14:17:33
CFNAME : CF37
ENTRIES: IN-USE: 73724 TOTAL: 73739,
99% FULL

DISPOSITION : KEEP
MAX CONNECTIONS: 4
# CONNECTIONS : 3
CONNECTION NAME ID VERSION SYSNAME JOBNAME ASID

STATE
---------------- -- -------- -------- -------- ----
----------------
SC30 01 00010007 SC30 SMSVSAM 0049
ACTIVE
SC31 03 0003000E SC31 SMSVSAM 004E
ACTIVE
SC32 02 0002000B SC32 SMSVSAM 0039
ACTIVE
Example 9-6 System log - increasing size of lock structure

SC30 2006145 14:43:13.75 -setxcf START,ALTER,STRNAME=IGWLOCK00,SIZE=800000
SC30 2006145 14:43:13.82 IXC530I SETXCF START ALTER REQUEST FOR STRUCTURE IGWLOCK00 ACCEPTED.
SC30 2006145 14:43:22.21 IXC533I SETXCF REQUEST TO ALTER STRUCTURE IGWLOCK00
COMPLETED. TARGET ATTAINED.
CURRENT SIZE: 86016 K TARGET: 86016 K
Example 9-7 System log - lock structure 27% full

STRNAME: IGWLOCK00
STATUS: ALLOCATED
TYPE: LOCK
POLICY INFORMATION:
FULLTHRESHOLD : 80
ALLOWAUTOALT : NO
DUPLEX : DISABLED

ENFORCEORDER : NO
ACTIVE STRUCTURE
----------------
ALLOCATION TIME: 05/25/2006 14:17:33
CFNAME : CF37
ENTRIES: IN-USE: 73739 TOTAL: 269207,
27% FULL
DISPOSITION : KEEP
MAX CONNECTIONS: 4
# CONNECTIONS : 3
CONNECTION NAME ID VERSION SYSNAME JOBNAME ASID

STATE
---------------- -- -------- -------- -------- ----
----------------
SC30 01 00010007 SC30 SMSVSAM 0049
ACTIVE
SC31 03 0003000E SC31 SMSVSAM 004E
ACTIVE
SC32 02 0002000B SC32 SMSVSAM 0039
ACTIVE
Solution
The expected behavior was observed. To avoid this issue you can use monitoring
tools such as RMF that will warn your installation before this threshold becomes
critical.

9.3.3 Read-only data table at initial time

The objective of this test was to observe the behavior of the system when one of
the VSAM files required to build a BASE24-es Online Transaction Processing
read-only memory table (OLTP) is not available when the CICS region is
initialized.
An error will be introduced reading the Journal Profile Group and Journal Profile
Group Assign data sources.
Expected behavior
An error will be generated during CICS region startup. The AOR will not be
functional and transactions routed to it will be declined. The AOR must be taken
out of service by the operator until the problem is resolved, at which time the
AOR can be restarted.
Actual behavior
During initialization, each routing region detected that the BASE24-es Journal
File configuration was not available. This was due to the VSAM files being either
not available or empty. All financial transactions were declined and returned to
the workload driver.
Following is the BASE24-es event message at installation time showing the

Journal File profile does not exist:
06-05-2518:54:07ISLO 1582 I20 EA21IS 0
JRNL: Profile = JLF_BK02_P08 does not exist in the JPGD data source.*
06-05-2518:54:08ISLO 1582 I20 EA23IS 0
06-05-2518:54:08ISLO 1582 I20 EA23IS 0
06-05-2518:54:08ISLO 1582 I20 EA03IS 0
06-05-2518:54:08ISLO 1582 I20 EA03IS 0
06-05-2518:54:08ISLO 1582 I20 EA02IS 0
06-05-2518:54:08ISLO 1582 I20 EA02IS 0
Solution

9.3.4 Read-only flat file at initial time

During CICS region initialization, BASE24-es obtains the schema for its VSAM
data sets and CICS memory tables from metadata configuration files that are
defined to CICS as Extrapartition Transient Data Queues. The objective of this
test was to observe the behavior of the system when one of the Extrapartition
TDQs required to determine the schema is not available at CICS region startup.
An error will introduced reading the CONFCSV and MDBCSV data sets, which
are mapped to corresponding Extrapartition TDQs during PLT processing.
Expected behavior
This is a catastrophic error and the application will be unable to generate
operator events or perform any financial processing. The application will abend.
The region will be inoperable until the operator corrects the error and restarts the
CICS region.
Actual behavior
During initialization, each routing region detected that the BASE24-es metadata
configuration file was not available and produced an event message. All financial
transactions were declined and returned to the workload driver. Each financial
transaction attempted to start a BASE24-es Integrated Server that will abend,
because it was uninitialized.
Following are the CICS CEEOUT messages indicating the failure during
initialization:
RSTR 20060525191115 SIRSTR: ES Region startup processing initiated.
RSTR 20060525191115 The DAL Bootstrap program for CICS has started.
RSTR 20060525191115 TDQ ACI1 Action = EXEC CICS SET TDQUEUE., Action Status =
failed, RESP = 17, RESP2 = 14. IOERR: An error
RSTR 20060525191115 occurred opening or closing the data set associated with
the queue.
RSTR 20060525191115 The DAL bootstrap program is halting due to errors.
RSTR 20060525191115 Subject: EMTBLD - DAL error: 48
RSTR 20060525191115 ÝSub-system: DAL Subject: CICS Action: EXEC CICS READQ TS.
Action Status: failed Error 44: (QIDERR: The
RSTR 20060525191115 queue specified can not be found in auxiliary or main
storage.)¨
Solution

9.4 BASE24-es application
Here we describe the following BASE24-es application scenarios:
1. Journal file full
2. IP client failure
3. IP server failure
4. Integrated server failure
5. Planned application upgrade
6. Distributed program link to back-end authorization failure
7. Usage file full
9.4.1 Journal file full

The objective of this test was to examine the behavior of the BASE24-es
Integrated Server when it encounters a file full condition on one of the
BASE24-es financial journal (JRNL) files in a switch-acquired offline-authorized
transaction environment; see Figure 9-1 on page 187.
To avoid a lengthy wait while writing a very large number of messages to fill one
or more JRNL files, we created the failure by modifying the JCL that creates the
JRNL to create artificially small JRNL files. We configured the workload driver to
send VDPS messages to be authorized offline by BASE24-es, and we observed
the behavior as records were added that exceeded the maximum capacity of the
primary JRNL file.
Test configuration
In the initial configuration, transactions flowed from the workload driver. They
were authorized by BASE24-es and logged to the primary JRNL file, and
responses were returned to the workload driver. There are primary and alternate
JRNL files associated with each card issuer configured in the system.

Workload
Drivers
LPAR LPAR
TOR IP TOR IP
Client Client
AOR AOR
IS IS
Primary JRNL
File Alternate
JRNL File
Event Log File
Figure 9-1 Journal file full scenario
Expected behavior
As each Primary JRNL file filled up, an error was returned to each Integrated
Server when it attempted to write to that file. The Integrated Server reports the
log in the Error Log File and in its associated CICS spooler, and the current and
subsequent financial transactions are logged by that Integrated Server to the
configured Alternate JRNL file associated with the failed primary. One event is
logged per BASE24-es Integrated Server per full JRNL file. Refer to Figure 9-2
on page 188 for an illustration of the journal file full expected behavior scenario.
A slightly longer response time may be experienced by the workload driver on the
request that actually encounters the failure; otherwise no impact to the workload
driver is expected.
Recovery occurs naturally in the course of the business cycle, as records are
extracted from the JRNL file and it is deleted and redefined before it is reused
some days later. In the intervening days the operator has the opportunity to
examine the error logs and modify the JCL to redefine a larger JRNL file if
necessary. (Recovery was beyond the scope of this test.)

Workload
Drivers
LPAR LPAR
TOR IP TOR IP
Client Client
AOR AOR
IS IS
Primary JRNL
File Alternate
JRNL File
Event Log File
Figure 9-2 Journal file full expected behavior scenario
Actual behavior
When the primary Journal File full condition was encountered, the financial
transactions were then written to the Alternate Journal File. An event message
was produced to indicate this action occurred for each file. During this process,
the workload continued uninterrupted.
Following is the BASE24-es event message showing the failure to insert to the
Primary Journal File:
06-05-2517:06:34ISLO 1046 E10 EA03IS 1
INSERTER: Insert failed for Kernel Inserter Class. Error TABLE FULL: The requ
est failed, no more records could be added to the table. Data Source: JLF_BK
02_P00_0 .*
06-05-2517:06:36ISLO 1046 E10 EA13IS 1
02_P05_5 .*
06-05-2517:06:36ISLO 1046 E10 EA13IS 1

02_P09_9 .*
06-05-2517:06:35ISLO 1046 E10 EA03IS 1
02_P01_1 .*
02_P04_4 .*
02_P06_6 .*
06-05-2517:06:34ISLO 1046 E10 EA03IS 1
06-05-2517:06:34ISLO 1046 E10 EA03IS 1
02_P03_3 .*
06-05-2517:06:35ISLO 1046 E10 EA03IS 1
02_P08_8 .*
06-05-2517:06:35ISLO 1046 E10 EA03IS 1
02_P07_7 .*
06-05-2517:06:36ISLO 1046 E10 EA13IS 1
02_P02_2 .*
Solution
9.4.2 IP client failure

The objective of this test was to examine the behavior of the BASE24-es system
under the spontaneous failure of the BASE24-es IP Client in a switch-acquired
transaction environment.
A spontaneous failure was simulated by using CICS CEMT to force purge the IP
Client task.

Workload
Drivers
LPAR LPAR
TOR TOR
IP IP
Client Client
BASE24-es BASE24-es
Task Task
Monitor Monitor
TDQ TDQ
AOR AOR
Figure 9-3 IP client failure scenario
Test configuration
Initially the workload driver was connected with two BASE24-es IP Clients in two
Terminal Owning Regions. Each IP Client distributed requests across the
available Application Owning Regions. Each AOR routed responses back to the
CICS TDQ in the TOR that originated the request; see Figure 9-3.
Expected behavior
When one BASE24-es IP client is force purged, the connection to the workload
driver is lost. Responses to transactions in flight may be on the TDQ associated
with the IP client, and responses may continue to be placed on the TDQ while the
IP Client remains down.
Within a few seconds, the IP Client will be restarted by the BASE24-es Task
Monitor task. At that time, connections with the workload driver will be
reestablished and new requests will start to flow.
Responses that were on the TDQ at the time of the failure, or that were placed on
the TDQ during the failure, will be passed on to the workload driver. However,
since the workload driver is not a fully functional switch, real world end-to-end
recoverability will not occur in this test scenario.

Actual behavior
Three IP clients were each running, processing one-third of the workload. One of
the IP clients was force purged. At this point the workload driver recognized the
client was down and the workload was reduced by one-third. After the IP client
automatically recovered, the full workload was resumed.
The IP process that was purged was restarted with a new CICS transaction ID by
the BASE24-es monitor process. Recovery took a few seconds (in our case,
seven seconds). During that time, three requests were indicated as timed-out at
the workload simulator. There were messages on the IP client queue to be
processed when it was purged, as listed here:
IP client task before failure:

Sta(S ) Use(CICSUSER) Uow(BED8505C40B0141)
IP client Failure:
05/23/2006 09:15:03 CICSET01 Transaction T710 abend AEXY in program EZACIC01
term ??? Updates to local recoverable resources will be backed out.
IP client recovery:
IP client Task after recovery:

Sta(S ) Use(CICSUSER) Uow(BED84B153BE63346)
Note: Task 0000058 had been replaced by task 0001592.
Solution
9.4.3 IP Server failure

under the spontaneous failure of the BASE24-es IP Server in a POS
device-acquired transaction environment.
A spontaneous failure was simulated by using CICS CEMT to force purge the IP
Server task.

Test configuration
Workload
Driver
Network
Router
LPAR LPAR
TOR TOR
IP IP
Server Server
BASE24-es BASE24-es
Transaction Transaction
Monitor Monitor
TDQ TDQ
AOR AOR
Figure 9-4 IP server failure scenario
Initially the workload driver was connected with two BASE24-es IP Servers in two
Terminal Owning Regions. Connections were distributed across the IP Servers
by a network router. Each IP Server distributed requests across the available
Application Owning Regions. Each AOR routed responses back to the CICS
TDQ in the TOR that originated the request; refer to Figure 9-4.
Expected behavior
At the time of the failure, connections between the workload driver and the IP
Server are lost. The workload driver will reestablish connections with the network
router, which will direct the connections to the remaining IP Server. Workflow will
resume.
Responses to transactions in flight may be on the TDQ associated with the IP

client, and responses may continue to be placed on the TDQ while the IP Client
remains down.
Within a few seconds, the IP Client will be restarted by the BASE24-es Task
Monitor task. If restart occurs quickly enough, some of the workload drivers
connections may be assigned by the network router to the restarted IP Server
rather than the surviving server in the other TOR.

Responses that were on the TDQ at the time of the failure, or that were placed on
the TDQ during the failure, will be returned to one of the available AORs for
failure processing, resulting in a financial reversal being logged to the appropriate
financial journal file. The workload driver will experience late or no responses.
However, since the workload driver is not a fully functional POS device, real
world end-to-end recoverability will not occur in this test scenario.
Actual behavior
Three IP servers were running, each processing one-third of the workload. One
of the IP servers was force purged. At this point the workload driver issued a new
client connection request and the connection was established to one of the
remaining IP servers. The workload driver showed no interruption to the
workload.
The IP process that was purged was restarted with a new CICS transaction ID by
the BASE24-es monitor process and was available to accept new connections
within few seconds (in our case, eleven seconds).
During that time, the following event was generated:

06-05-2309:43:49ISLO 2031 E180 EA11IS 0
SPDH: Failed message received for Channel ID S000020153 with insufficie
nt PAN/Amount/Sequence Number data to be processed.*
Due to a limitation in the workload driver, sequence number processing was

disabled in the BASE24-es SPDH component. Without sequence number
processing, the above event was generated. When sequence number processing
is enabled, this event would not be generated and this financial transaction will
result in a reversal.
IP client task before failure:

Sta(S ) Use(CICSUSER) Uow(BED8505D8D1F589B)
IP client Failure:
05/23/2006 09:43:38 CICSET01 Transaction T701 abend AEXY in program EZACIC01
IP client recovery:
IP client Task after recovery:

Sta(S ) Use(CICSUSER) Uow(BED8517C1A4B15D6)

Solution
9.4.4 Integrated Server failure

under the spontaneous failure of one BASE24-es Integrated Server (IS) task in a
switch-acquired offline-authorized transaction environment. The spontaneous
failure was simulated by using CICS CEMT to force purge one of the IS tasks in
one AOR.
Test configuration
At inception, IP clients in the TORs have connected to the workload driver, and
transactions are being distributed across the IS TDQs in the available AORs and
are being authorized. When one IS task is force-purged, the remaining IS tasks
in the AOR continue processing messages from the TDQ. (Sufficient IS tasks are
configured to assure the loss of any one task will not affect the throughput
capability of the AOR.)
Within a few seconds the BASE24-es Task Monitor will restart the failed IS task
and processing will return to normal.
If a message is being processed by the failed IS at the time of failure,

uncommitted database updates will be backed out by CICS, assuring financial
integrity within the BASE24-es system.
If a message is being processed by the failed IS at the time of failure, the

message will time out at the workload driver.Figure 9-5 on page 195 illustrates
the Integrated Server failure scenario.
Because the workload driver is not a fully functional switch, end-to-end recovery
will not occur. However, in the real world, end-to-end recovery would assure
financial integrity between the acquiring interface and BASE24-es.

Workload
Driver
LPAR LPAR
TOR TOR
AOR AOR
IS IS
IS IS
TDQ TDQ
TDQ
BASE24-es BASE24-es
Task Task
Monitor Monitor
Figure 9-5 Integrated Server failure scenario
Expected behavior
The expected behavior was as follows:
1. The workload will be taken up by the remaining IS tasks.
2. The IS task will be restarted by the ACI monitor program.
3. Any message actually being processed by the IP handler at the time it was
cancelled will be lost. Any associated database changes will be backed out.
No out-of-balance condition exists, since the transaction has been backed out
and the acquirer has not received an approval.
Actual behavior
Workload continued to be processed uninterrupted.
The IS process that was purged was restarted with a new CICS transaction ID by
the BASE24-es monitor process.

IS task before failure:
Tas(0000269) Tra(IS01) Sus Tas Pri( 254 )
Sta(S ) Use(CICSUSER) Uow(BED7685E432DB499) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7685E48D7AF19) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED768526B8749D6) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76820F1626C19) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7682A8A19DF59) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7682F6A492599) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7683445B607D9) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7683DFC57C89B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7685AEAA7AF59) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED768560010865B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7685AD633691B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7685AD66ABF1B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED767AE3E4FFD94) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED767AE3E536854) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED767AE3E767D41) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED767AE3E78CE41) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED767AE3E9DEAD4) Hty(EKCWAIT )
IS Failure:
5/22/2006 16:22:09 CICSEA01 Transaction IS01 abend AEXY in program ISLO

IS Task after recovery:
Sta(S ) Use(CICSUSER) Uow(BED7694EAD7544DB) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7694B5262EC1B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7694BEE72601B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED768A848D28546) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED768BB797F2759) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7694B86CA4DD9) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED768C0319C6696) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7694250833914) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED768C9E816C319) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76947A55DD256) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED768DD17E27719) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED768E1F090961B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED768E6B8E1DF1B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7690D296DBD99) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76916B4167E16) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7691B7C7917D4) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769250D551396) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7693D1462F0D9) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769420CB66396) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76947BBBA23DB) Hty(EKCWAIT )
Solution

9.4.5 Planned application upgrade

during a planned upgrade of the BASE24-es Integrated Server (IS) software in
an active system authorizing switch-acquired offline-authorized transactions.
Assuming the BASE24-es Integrated Server is referenced as CICS PROGRAM

(ISLO) and as CICS TASK (IS01), the sequence of commands (from a blank
CICS screen) would be:
CEMT SET PROGRAM (ISLO) PHASEIN
SIQM RESTART IS01
These commands would be executed in each AOR in which the IS program is to

be upgraded.
Notes:
򐂰 If there are multiple transactions associated with the ISLO program, it may
be preferable to use the 'SIQM RESTART ALL' command to restart all the
long-running BASE24-es tasks in the AOR, rather than issuing the 'SIQM
RESTART tran' command for each transaction.
򐂰 The RESTART command causes the rolling restart of the IS servers and it
should have no impact on processing.
Figure 9-6 on page 199 illustrates the Planned application upgrade scenario.

Test configuration
Workload
Driver
LPAR LPAR
TOR TOR
AOR AOR
IS IS
IS IS
TDQ TDQ
TDQ
BASE24-es BASE24-es
Task Task
Monitor Monitor
Figure 9-6 Planned application upgrade scenario
Expected behavior
Setting the ISLO program to PHASEIN specifies that a new copy of the program
is brought into CICS memory. Because the RESTART command causes the IS01
tasks to stop and restart, they are restarted with the new version of the ISLO
program.
The RESTART command causes a rolling restart so that instances of the task
are stopped and restarted sequentially outside of a financial transaction. Some
minor increases in response time will occur as a result of the RESTART
command. No impact to the financial integrity of the system is expected.
Actual behavior
The IS processes before the restart command was issued:

Sta(S ) Use(CICSUSER) Uow(BED76A030857CE99) Hty(EKCWAIT )

Sta(S ) Use(CICSUSER) Uow(BED76A02C9233299) Hty(FCRVWAIT)
Sta(S ) Use(CICSUSER) Uow(BED769FD7AB60D56) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769F93E56019B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769E34FD01F59) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76A003DA96316) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED7699AF53B0F9B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769F6DC16E314) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769A48CD0529B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769A9517DB894) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76A007D16155B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769B30EEDF596) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769BCA13098DB) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769B7D5C3EF99) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769C65665DA1B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769FB76A080D9) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED769CB28FEE659) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76A015C6B3214) Hty(EKCWAIT
Sta(S ) Use(CICSUSER) Uow(BED769E34F9B6659) Hty(EKCWAIT
Sta(S ) Use(CICSUSER) Uow(BED769FBB6C2CD9B) Hty(EKCWAIT
Restart commands:
CEMT SET PROG(ISLO) PHASEIN
SIQM RESTART IS01
Following are the event messages notifying that the IS stop/start process was
initiated by the commands:
IS01 20060522163011 SIS MDS: Task stopping, name = %EA02IS010000262

IS Task after restart:

Sta(S ) Use(CICSUSER) Uow(BED76B34D3D91719) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76B33C7741CD9) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76B327D9CCD5B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76B33F131EB99) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76ACECF9CD85B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76B0D6BD60C54) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76AD39CA3E456) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76ACECF5F491B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76ADD552C5C96) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76AD88A117514) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76AE233E78C9B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76B257A60A35B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76B1236A76914) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76AF566194094) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76AF0906B6316) Hty(EKCWAIT )

Sta(S ) Use(CICSUSER) Uow(BED76AFF04629019) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76AFA3613671B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76B34099C945B) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76B2F31994059) Hty(EKCWAIT )
Sta(S ) Use(CICSUSER) Uow(BED76AD392BC8E59) Hty(EKCWAIT )
Note: All task IDs have changed.
Solution
9.4.6 Remote program link to back-end authorization failure

under the failure of a CICSLINK to a customer’s back-end host system.
The assumption was that there will be at least two AORs running the customer's
back-end host system, and that BASE24-es will route online authorizations to
those systems via remote LINK. At least two BASE24-es stations will be
configured, each associated in the BASE24-es synchronous destination map file
(SYDMF) with a separate remote program definition in a separate authorizing
region. Three regions running a host simulator are available. The failure will be
injected by canceling one of the back-end AORs. Figure 9-7 on page 203
illustrates this scenario.
Note: We chose to test this scenario using the alternate routing capabilities of
the BASE24-es application, rather than using the equally valid approach of
configuring a CIVSplex distributed link.

Test configuration
TOR TOR
BASE24-es BASE24-es
AOR AOR
SAF
TDQ SAF
TDQ
IS IS
CPSM
Backend Authorization Backend Authorization

AOR AOR
Figure 9-7 Distributed program link to back-end authorization failure scenario
Expected behavior
BASE24-es would use round robin processing across the available stations. After
the AOR failure, messages to the associated station would experience an error
on the link. BASE24-es would stand in and authorize those messages, adding
financial notifications to the store-and-forward file.
After a configurable number of link errors, BASE24-es would mark the station as
down and continue processing using the remaining stations
Actual behavior
BASE24-es continued to attempt to route transactions to the cancelled host
region and produced a BASE24-event message. Alternate routing selected an
active host region. The transaction was then returned to the workload driver as
authorized. Workload continued to be processed uninterrupted throughout this
process.

Note: If you do not wish to see the BASE24-event message produced when an
attempt is made to route to a cancelled host region, use the BASE24-es Event
management system to control the suppression of this message.
Following is the BASE24-es event message showing inability to link to host:

06-05-2611:38:35ISLO 1042 I110 EA11IS 1
SENDMSG: Unable to send and /or receive a message from the SIS Send
Class.
Result 12 returned.ÝSub-system: CICS Subject: SIMIATT0 Action: EXEC
CICS LINK
Action Status: failed Error 53: (SYSIDERR: 28 - The remote system
specified in SYSID is not in service.)¨*
Solution
9.4.7 Usage file full

The objective of this test was to examine the behavior of the BASE24-es
Integrated Server when it encounters a file full condition on the BASE24-es
Usage files in a switch-acquired offline-authorized transaction environment.
The Usage file contains one record per card per usage period. The error was
created by an extended high-volume run with random cards.
Test configuration
In the initial configuration, transactions flow from the workload driver. They are
authorized by BASE24-es, logged to the primary JRNL file, and responses are
returned to the workload driver. There are primary and alternate JRNL files
associated with each card issuer configured in the system. Figure 9-8 on
page 205 illustrates the Usage file full scenario.

Workload
Drivers
LPAR LPAR
TOR IP TOR IP
Client Client
AOR AOR
IS IS
USGD File
Event Log File
Figure 9-8 Usage file full scenario
Expected behavior
When a file full error is encountered on the Usage file, BASE24-es logs an
operator event indicating the error but continues to authorize transactions.
Updates to existing records continue. The fact that BASE24-es is unable to write
a new record to the USGD does not mean the initial transaction it is trying to
record should be retried.
Recovery will occur naturally as usage periods elapse and records are removed
from USGD.
Actual behavior
After the Usage file was full, Updates to existing Usage File records were
processed normally. Each attempt to add a new record resulted in the following
message:
EERR 20060517190454 Kernel Inserter Class. Error TABLE FULL: The request
failed, no more records could be added to the table.

EERR 20060517190454 Data Source: USAGE
.ÝSub-system: FILE_PARTITION Subject: USAGE
EERR 20060517190454 Action: Row Insert Action Status: FAILED Error
Assign Entry Data Source: (insertion failed)
EERR 20060517190454 ¨
Solution
If this error is encountered frequently, it may be necessary to take an outage to
resize the system, so it is good practice to assure that the usage file is
adequately sized initially.
9.5 Hardware
This section covers the following hardware failure scenarios:
1. Central Processor (CP) failure
2. CPC failure/LPAR failure
3. Coupling Facility (CF) failure
9.5.1 Central Processor (CP) failure

The objective of this scenario was to demonstrate that the application
BASE24-es is able to continue working even if a CP failure occurs.
Because our LPARs are symmetrical, we can perform this test on any LPAR and
the behaviors should be similar on all of them. In our case, we used SC31. We
initiate this test by removing a CP using the CF CP(0),offline command.
Configuration
When we issue the command D M=CPU on SC31, we can see in the response that
the LPAR is running on two processors (CPs). For this reason, taking one
processor offline should not affect the functionality of the application. If this was
the only or last processor in the LPAR, a failure of the processor will be presented
similarly as an LPAR failure or CPC failure.
IEE174I 13.52.53 DISPLAY M 101
PROCESSOR STATUS
ID CPU SERIAL
00 + 24991E2094
01 + 24991E2094
02 -
03 -

CPC ND = 002094.S18.IBM.02.00000002991E
CPC SI = 2094.712.IBM.02.000000000002991E
CPC ID = 00
CPC NAME = SCZP101
LP NAME = A24 LP ID = 24
CSS ID = 2
MIF ID = 4
+ ONLINE - OFFLINE . DOES NOT EXIST W WLM-MANAGED

N NOT AVAILABLE
To initiate the test, we issued the following command from SC31:

CF CPU(0), offline
This will result in a behavior that is very similar to a real CP failure.
Expected result
We are expecting a message on the console informing us that CP0 is having a
problem, but the application should continue to run without failure. The CPU
sparing feature handles this type of failure in a way that will be transparent to the
application and operations.
Actual behavior
Three LPARs were each running processing one-third of the workload. Each
LPAR had two online CPUs. In one of the LPARS, CPU(0) was taken offline.
Workload continued uninterrupted. CPU(0) was later brought back online.
Following are the system log messages showing CPUs 00 and 01 online:
RESPONSE=SC31
PROCESSOR STATUS
ID CPU SERIAL
00 + 24991E2094
01 + 24991E2094
02 -
03 -
Following is the SDSF display showing LPAR is 16% busy across two CPUs:
SDSF DA SC31 SC31 PAG 0 CPU/L 16/ 16 LINE 1-4 (4)
COMMAND INPUT ===> SCROLL ===> CSR
NP JOBNAME ame ProcStep JobID Owner C Pos DP Real Paging SIO CPU%
CICSEA11 A11 CICS STC04571 SYSPROG NS FE 49T 0.00 325.17 3.05

CICSET11 T11 CICS STC04568 SYSPROG NS FE 14T 0.00 0.00 1.52
Following is the system log messages showing only CPU 01 is online:

RESPONSE=SC31 IEE505I CPU(0),OFFLINE
RESPONSE=SC31 IEE712I CONFIG PROCESSING COMPLETE
RESPONSE=SC31
PROCESSOR STATUS
ID CPU SERIAL
00 -
01 + 24991E2094
02 -
03 -
Following is the SDSF display showing LPAR is 26% busy across one CPU:
SDSF DA SC31 SC31 PAG 0 CPU/L 26/ 26 LINE 1-4 (4)
COMMAND INPUT ===> SCROLL ===> CSR
NP JOBNAME ame ProcStep JobID Owner C Pos DP Real Paging SIO CPU%
CICSET11 T11 CICS STC04568 SYSPROG NS FE 14T 0.00 116.18 3.02
Following is the system log messages showing CPUs 00 and 01 are online:
RESPONSE=SC31 IEE504I CPU(0),ONLINE
RESPONSE=SC31 IEE712I CONFIG PROCESSING COMPLETE
RESPONSE=SC31
PROCESSOR STATUS
ID CPU SERIAL
00 + 24991E2094
01 + 24991E2094
02 -
03 -
Solution

9.5.2 CPC failure/LPAR failure

The objective of this test was to demonstrate that the workload will continue to
flow and execute successfully in the case of either an LPAR or CPC failure. We
assumed the workload is spread across multiple LPARs/CPCs.
Because the behavior of this scenario is similar for either an LPAR or CPC loss,
we will perform the test by failing the LPAR. To initiate the test, we need to access
the HMC of the CPC and deactivate the partition, which was SC32 in our case.
For this test, we are also planning to have the ARM function enabled to
automatically restart the CICS regions on one of the surviving LPARs.
Configuration
The test took place while the workload was spread across the three LPARs:
SC30, SC31, and SC32. One of the LPARs, SC32 in our case, is going to be
deactivated to simulate a failure.
Expected result
As soon as the deactivation task is completed, the icon representing the LPAR on
the HMC should change from green to red. The CICS regions and the
SMSVSAM server will be down, as a consequence.
When CICS terminates, there might be active locks left that are converted to
retained locks. Locked records cannot be accessed by transactions in other
CICS regions. Transactions in other CICS regions will wait for an active lock. For
this reason, the CICS regions need to be brought back on a running system in
the sysplex to be able to resolve the potential retained lock condition.
When the CICS region is brought back, CICS will analyze the logging data,
perform rollback, and release the locks. We expect to see the CICS regions
restarted automatically by ARM.

Note from a logging point of view: In our exercise, we fail an LPAR. In this
configuration, when the CICS regions restart on a different system, they will be
able to retrieve the logging information from the Coupling Facility where the
log streams are located.
If instead of failing the LPAR, we suffered an error in the CEC—and the

Coupling Facility was located in the same CEC—then the CICS region will be
using the logging staging data sets to retrieve the information needed for a
restart. The logging configuration is a key point in the LPARs and Coupling
Facility configuration
There is no need to restart the SMSVSAM server, because there is one

infrastructure running on each system and the lock information is shared through
the Coupling Facility.
After the CICS (AORs) regions are brought back and initialized, then unless they
required for the workload, these regions can be taken down. The only reason for
restarting these regions is to be able to clear potential locks held on RLS records.
From the workload driver perspective, when the connection with the TOR is lost,
the driver will reconnect automatically after 5 seconds using the same IP
address. If the network has routing capability, this connection will be routed and
established with the surviving TORs. In this case there will no loss from the
client’s side.
If the network does not have routing capability, the driver will keep pinging the
same IP address until the TOR is back on the same LPAR.
Actual behavior
Three LPARs were each running processing one-third of the workload. Then one
LPAR was stopped.
This caused the IP client processes and the CICS mirror transactions on the
surviving LPARs to remain in an IRLINK state for several seconds. During this
time, the workload simulator did not send new workload.
After the LPAR was removed from the sysplex by an active SFM policy, the
workload simulator again sent workload to the connected IP client processes.
One observed exception was that our workload simulator did not accept
connections from one of the IP client processes. This was resolved by recycling
the workload simulator.

You can see that LPAR SC32 is not longer part of the sysplex:
RESPONSE=SC30
SYSPLEX COMMPLEX: SC30 SC31
Solution
9.5.3 Coupling Facility (CF) failure

Objective of this test was to verify that the workload continues to flow and
execute successfully in case of a Coupling Facility failure. A Coupling Facility
failure could occur as a result of a hardware failure (probably a power failure),
improper allocation of log stream storage, or loss of connectivity to it.
In this scenario, we demonstrate that, if a Coupling Facility fails, any allocated

structures residing in that Coupling Facility can be rebuilt into the alternate
Coupling Facility by the structure owner. In our configuration, IXGLOGR owns the
CICS log structures and SMSVSAM owns the lock and cache structures for
VSAM RLS.
Note: This scenario makes sense only if the installation is planning to have at
least two Coupling Facilities in the configuration. With one Coupling Facility,
some of the elements might not be able to recover the single point of failure.
Configuration
Following is the configuration of the Coupling Facilities used for our scenarios:
CF37 contains the following structures:

IGWLOCK00 ISGLOCK ISTGENERIC
ISTMNPS IXC_DEFAULT_1 IXC_DEFAULT_4
LOG_DFHLOG_001 RLSCACHE01 SYSTEM_OPERLOG
SYSZWLM_6A3A2084

IXC_DEFAULT_2 IXC_DEFAULT_3 LOG_DFHSHUNT_001
RLSCACHE02 SYSIGGCAS_ECS SYSTEM_LOGREC
SYSZWLM_WORKUNIT SYSZWLM_991E2094

Failing either one of the Coupling Facilities will have a similar effect on our
workload since the involved structures (logging and SMSVSAM) are equally
distributed between the two Coupling Facilities.
To initiate the test, we need to access the HMC of the CEC and deactivate the
partition, CF39 in our case.
Expected result
Coupling Facility failure in a CICS RLS system will disable access to all data sets
being accessed in RLS mode and, more importantly, disable all CICS system
logging unless a suitably configured alternate Coupling Facility is available. With
a suitably configured system, the only impact of a Coupling Facility failure should
be the time to recognize the error and rebuild the structures into the alternate
Coupling Facility.
The Coupling Facility failure messages are issued to all systems in the Parallel
Sysplex at the time of the failure. The subsystems that own the structures will
then drive the rebuild of the structures. Within seconds, the structures should
recover successfully into the new Coupling Facility, and, in the case of log stream
data, CICS logging should resume automatically.
Actual behavior
We had two Coupling Facilities active. We halted one of the Coupling Facilities.
The structures that were active in the halted Coupling Facility were rebuilt in the
surviving Coupling Facility. During this time workload continued uninterrupted
and BASE24-es did not log any events.
Following are the system log messages before the failure of CF37:
Note: The following is the output of the D CF command issued against CF37
and CF39 showing the CONNECTED SYSTEMS and STRUCTURES for each CFs.

CFNAME: CF37
COUPLING FACILITY : 002084.IBM.02.000000026A3A
SITE : N/A
POLICY DUMP SPACE SIZE: 2048 K
ACTUAL DUMP SPACE SIZE: 2048 K
CONNECTED SYSTEMS:
SC30 SC31 SC32

STRUCTURES:
SYSZWLM_6A3A2084

CFNAME: CF39
COUPLING FACILITY : 002094.IBM.02.00000002991E
PARTITION: 2F CPCID: 00
SITE : N/A
CONNECTED SYSTEMS:
SC30 SC31 SC32
STRUCTURES:
Following are the system log messages after failure:

4040000 SC30 2006144 17:29:49.24 00000090 IFB100E LOGREC LOG
STREAM ERROR ON SYSTEM SC30 - RC=0008-0864 815
815 00000090 UNABLE TO WRITE TO
LOG STREAM - LOG STREAM IS NOT AVAILABLE
815 00000090 LOG STREAM NAME:
SYSPLEX.LOGREC.ALLRECS
815 00000090 STRUCTURE NAME:
SYSTEM_LOGREC
Note: The following is the output of the D CF command issued against CF37
and CF39 after the failure, showing the CONNECTED SYSTEMS and STRUCTURES for
each CFs.
CFNAME: CF37
SITE : N/A

CONNECTED SYSTEMS:
SC30 SC31 SC32
STRUCTURES:
IXC_DEFAULT_3 IXC_DEFAULT_4 LOG_DFHLOG_001
LOG_DFHSHUNT_001 RLSCACHE01 RLSCACHE02
SYSIGGCAS_ECS SYSTEM_LOGREC SYSTEM_OPERLOG
SYSZWLM_WORKUNIT SYSZWLM_6A3A2084 SYSZWLM_991E2094
Following are the system log messages after failure for CF37:
Note: No systems are shown in the CONNECTED section of the message.
CFNAME: CF39
COUPLING FACILITY : 002094.IBM.02.00000002991E
PARTITION: 2F CPCID: 00
SITE : N/A
NO SYSTEMS ARE CONNECTED TO THIS COUPLING FACILITY
STRUCTURES:
IXC_DEFAULT_2(PND) IXC_DEFAULT_3(PND) LOG_DFHSHUNT_001(PND)
RLSCACHE02(PND) SYSIGGCAS_ECS(PND) SYSTEM_LOGREC(PND)
SYSZWLM_WORKUNIT(PND) SYSZWLM_991E2094(PND)
Solution
9.6 z/OS and relevant subsystems

The following scenarios highlight best practices regarding z/OS and the relevant
subsystems:
򐂰 z/OS maintenance

򐂰 CF maintenance
9.6.1 z/OS maintenance - best practices

Objective of this test was to show that the workload from the simulator will
continue to flow and execute successfully even when we are removing a z/OS
system from the configuration.
To initiate the test, we will perform an orderly shutdown of all the subsystems
running on one of the z/OS images (for example, SC32), by following a normal
shutdown procedure. After the subsystems have been closed, we can deactivate
the LPAR from the HMC (optional).
Configuration
The test will take place while the workload is running across the three systems:
SC30, SC31, and SC32. One system, SC32 in our case, is going to be shut down
to simulate a normal maintenance procedure.
Expected result
There should be no impact to end users.
When the connection with the TOR is closed, the workload simulator will
reconnect automatically after 5 seconds, using the same IP address. If the
network has routing capability, this connection will be routed and established with
the surviving TORs. In this case there will no loss from the client’s side.
If the network does not have routing capability, the driver will keep pinging the
same IP address until the TOR is back on the same LPAR.
Actual behavior
The procedure to maintain a z/OS system has been tested within IBM and
documented in multiple sources. We do not consider this scenario to be specific
to the BASE24-es application, and implicitly it is covered in 9.5.1, “Central
Processor (CP) failure” on page 206.
9.6.2 CF maintenance - best practices

The objective of this test was to verify that the workload from the simulator
continues to flow and execute successfully when it is necessary to shut down a
Coupling Facility for maintenance. A Coupling Facility might need to be removed

from a sysplex for maintenance, to have a new level of Coupling Facility control
code installed, or to be replaced.
When such a situation occurs, you must remove any structures contained in the
Coupling Facility before it is shut down and removed from the sysplex. We will
demonstrate that it is possible to move all the structures from one Coupling
Facility to the alternate one without interrupting the workload.
This scenario makes sense only if the installation is planning to have at least two
Coupling Facilities in the configuration. With one Coupling Facility, some of the
subsystems will not be able to operate.
Configuration
Following is the configuration of the Coupling Facilities used for our scenarios:
SYSZWLM_6A3A2084

Since the structures are equally distributed among the CFs, the procedure to
apply maintenance would be similar for either one of the CFs. In our scenario, we
use CF37.
The following procedure will remove CF37 from the sysplex:

1. Create a new CFRM policy that does the following (optional step):
– Deletes any reference to the Coupling Facility to be removed.
– Includes the name of the Coupling Facility where all structures are to be
rebuilt in the preference list.
Use a name for the new CFRM policy that is different from the name of the
original CFRM policy so that when Coupling Facility maintenance is
complete, the original policy can be restored.
2. To determine the CHPIDs in use by the Coupling Facility to be removed, issue
the following command on each system connected to the Coupling Facility: D
CF,CFNAME=CF37

3. Start the new CFRM policy. SETXCF
START,POLICY,TYPE=CFRM,POLNAME=newpolicyname (only if step1
executed)
4. To obtain the names of all allocated structures in the Coupling Facility to be
removed, issue the following command: D XCF,CF,CFNAME=CF37
The following response should be received. At the bottom is a list of the

structures that need to be moved to the alternate Coupling Facility before
shutting down CF37:
CFNAME: CF37
SITE : N/A
CONNECTED SYSTEMS:
SC30 SC31 SC32
STRUCTURES:
SYSZWLM_6A3A2084
5. At this point, the structures resident on CF37 need to be moved to CF39. All
of the structures currently allocated in CF37 support the rebuild function.
Rebuilding a structure allows the application to remain active during the
reconfiguration.
Rebuilding a structure is the preferred way to move a structure from a
Coupling Facility. If you are unsure of whether a structure supports rebuild,
you can attempt a rebuild of the structure. If rebuild is not supported, the
system will indicate that the rebuild operation cannot occur, and you must use
an alternative procedure for moving the structure.
The following commands will remove the structures from CF37:
SETXCF START,REBUILD,CFNAME=CF37,LOCATION=OTHER
Note: The LOCATION=OTHER specification indicates that the structures

are to be rebuilt in another coupling facility in each structure's preference
list other than the coupling facility in which the structure now resides.
SETXCF START,REBUILD,STRNAME= IXC_DEFAULT_1,LOCATION=OTHER

SETXCF START,REBUILD,STRNAME= IXC_DEFAULT_4,LOCATION=OTHER
Note: The XCF signalling structures must be rebuilt one at a time. After all
structures have been removed from CF37, it can be removed from the
configuration for maintenance.
Configure all CHPIDs offline to the Coupling Facility that you are removing.
6. Power off the Coupling Facility. In our case, we will simple deactivate the CF
LPAR. When the Coupling Facility is deactivated, any remaining structure
data is lost.
Expected result
We expect to see the rebuild messages for the structures allocated in CF37. After
the moves are completed, there should be no structures left in CF37. We expect
to see no variation, from a workload point of view. Within seconds, the structures
should be successfully moved into the new Coupling Facility, and in the case of
log stream data, CICS logging should resume automatically.
Actual behavior
The maintenance procedure for a Coupling Facility has been tested within IBM
and documented in multiple sources, for example, in Setting up a Sysplex. We do
not consider this scenario to be specific to the BASE24-es application and it is
implicitly covered in 9.5.3, “Coupling Facility (CF) failure” on page 211.

10
Chapter 10. Conclusion

In the payment transaction industry, it is crucial to have an IT infrastructure that is
designed and deployed with sufficient redundancy to support the uptime
requirements of its users, because failures occur in hardware, system software,
application software, and networks.
In this chapter, we summarize the high availability requirements of a payment

transactions enterprise and reiterate how the BASE24-es installation on z/OS
and the test scenarios described this redbook demonstrate the ability to achieve
the highest level of availability. You can use these scenarios to verify correct
product installation. We also present a minimum setup for high availability of
BASE24-es on z/OS.

10.1 Summary
Failures in a payments environment are a highly visible customer service issue.
The entire payments cycle must be conducted in near real-time. Outages at any
point will negatively affect customer loyalty. A payments environment is an online
business that must be up and running when customers expect the system to be
there, and have the capacity to handle the heaviest workload.
In this environment, continuous availability from end to end is not just a server
requirement, but a requirement for all components in the IT complex. The
weakest link determines availability as perceived by customers.
Maintaining availability for ATM and POS networks has been characterized by
predictable workloads for many years. Application changes have been few and
the business model has been stable. A stable environment also requires
comparatively few personnel to support it and only limited changes in their skill
sets.
10.2 The payment environment requirements

The new-generation payments environment, however, will be characterized by
higher levels of front-end and back-end integration. The ability to operate
transparently across all channels, including home and mobile banking via the
Internet, will be critical. This will mean not only that a range of new technologies
must be accommodated, but also that workloads will become less predictable
and more subject to unexpected peaks and fluctuations.
The cost of maintaining high availability is of great importance to the payments

industry. The technical challenge facing this environment is how to modernize
processes while at the same time radically reduce switching costs and minimize
the risks of change.
In-house developed solutions account for more than half the systems in the total
payments environment. These legacies, often old designed applications, consist
of low-level programming languages which require a specialist staff to maintain.
The design often consists of inflexible software that cannot be easily modified to
keep pace with market changes.
BASE24-es uses modular design/object-orientated application development

tools and scripting. The scripting component is based on open systems
architecture and enables users to rapidly change specific authentication
requirements on transactions from delivery channels. Benefits of this scripting
flexibility are shorter response times to changing business needs, reduced risk,

and improved customer loyalty. The enhanced configuration capability offers a
high degree of flexibility to address individual service needs.
The BASE24-es program code stream is by and large generic for all platforms, so
that the differences in the implementations on these platforms can be attributed
largely to the Quality of Service (QoS) of the platforms themselves and their
underlying basic software infrastructure.
IBM System z is designed for, and employed in, a broad range of

business-critical roles. Mixed workload execution, as well as data and system
management capabilities, are superior to any server counterparts. System z also
implements functions such as logical partitioning and virtual Linux™ server
hosting, and the workload management facilities are significantly better
optimized for volatile Web transaction workloads than those of other servers.
System z, when deployed in a Parallel Sysplex cluster, offers continuous

availability, an extraordinarily high-performance I/O handling capability, and near
linear scalability for BASE24-es on z/OS. A Parallel Sysplex cluster can be
engineered with no single point of failure by employing redundant I/O paths with
transparent, dynamic switching, triple mirroring, highly-available networking, and
automatic and transparent failure handling and restart capabilities.
10.2.1 Conclusions
Our BASE24-es installation on z/OS and the tests performed in the scenarios
clearly demonstrate the ability to achieve the highest level of availability.
For this project, we tested numerous failure scenarios for availability on

hardware, system software, and application software. The scenarios are
described in detail in this redbook, and they can be used as reference for the
verification of correct product installation. We used BASE24-es release 06.2 in a
beta version, and achieved the expected behavior.
A payment transaction system is a highly specialized application. When moving

an existing in-house developed application or EFT application from another
platform, the staff who know and understand the application can apply their
knowledge and experience to migrate the software to the mainframe platform.
Minimum configuration
In this section we list a minimum setup for high availability of BASE24-es on
z/OS:
򐂰 Two LPARs distributed in a Parallel Sysplex cluster, one LPAR per CPC. Be
sure your configuration has at least two Coupling Facilities.
Chapter 10. Conclusion 221

򐂰 On each LPAR, at least one Terminal Owning Region (TOR) running with at
least two cloned Application Owning Regions (AORs).
򐂰 On each AOR, at least two BASE24-es Integrated Server tasks.
򐂰 CICSplex System Manager (CP/SM), together with BASE24-es workload
management facilities, are used to provide the dynamic routing capability
from each TOR to all the AORs.
򐂰 AORs accessing the VSAM files in Record Level Sharing (RLS) mode through
the SMSVSAM address space. SMSVSAM will allocate the cache and lock
structure to the Coupling Facilities available to the Parallel Sysplex cluster.

Glossary
Application-Owning Region (AOR). In multiregion Coupling Facility Resource Management

operation (MRO) or intersystem communication (CFRM). A set of definitions for the use of a Coupling
(ISC), a CICS address space whose primary Facility. CFRM provides the services to manage
purpose is to manage application programs. It Coupling Facility resources in a sysplex.
receives transaction routed requests from a
terminal-owning region (TOR). See also CICS Complex (CICSPlex). A collection of related
terminal-owning region, file-owning region. and connected CICS regions, which helps to
address the efficiencies in having multiple,
Application Programming Interface (API). The full-function CICS systems processing a single
interface by which an application program accesses OLTP workload.
operating system and other services. An API is
defined at source code level and provides a level of Client (TCP/IP). A TCP/IP client establishes a
abstraction between the application and the kernel connection, typically using the connect() sockets
(or other privileged utilities) to ensure the portability API call. In the world of short-lived sockets, the client
of the code. usually also speaks first, sending the initial
application-level request, receiving a response, and
Automatic Restart Manager (ARM). A z/OS disconnecting. In the world of long-lived sockets
recovery function that can automatically restart common in the world of financial online transactions,
batch jobs and started tasks after they or the system it is common for a TCP/IP client to be an
on which they are running terminate unexpectedly. application-level server. The BASE24-es TCP/IP
Client Communications Handler is a TCP/IP client.
Asynchronous request. A request that is issued See Server, Listener.
without immediately awaiting a response. If a
response is expected, the issuing application CICSPlex SM Address Space (CMAS). A
typically starts a timer and performs other work until CICSPlex SM component that is responsible for
a response is received or a timer expiration occurs. managing CICSplexes. A CMAS provides the
single-system image for a CICSplex by serving as
Coordinating Address Space (CAS). The function the interface to other CICSplexes and external
that sets up the CICSPlex SM component topology programs. There must be at least one CMAS in each
and that supports the MVS/TSO ISPF graphic user MVS image on which you are running CICSPlex SM.
interface to CICSPlex SM. CAS is used in A single CMAS can manage CICS systems within
CMAS-to-CMAS links. one or more CICSplexes.
Coupling Facility (CF). A special LPAR that Central Processor (CP). The part of the computer
provides high-speed caching, list processing, and that contains the sequencing and processing
locking functions in Parallel Sysplex. facilities for instruction execution, initial program
load, and other machine operations.
CFRM policy. The allocation rules for a Coupling
Facility structure that are declared by a z/OS
administrator.

CICSPlex System Manager (CP/SM). A Hardware Security Module (HSM). A secure
system-management tool that enables you to Crypotographic facility for card holder PIN
manage multiple CICS systems as though they were transaction, verification and message
one. CICSPlex SM can manage independent, authentication.
full-function CICS systems running on one or more
connected central processor complexes (CPCs) just Listener (TCP/IP). A TCP/IP listener waits for a
as easily as it can manage multiple, interconnected connection, typically using the bind(), listen() and
CICS systems functioning as a CICSplex, also on accept() sockets API calls. A TCP/IP server typically
one or more connected CPCs. does not perform other productive work, handing the
established socket off to another task to accept
Central Processor Complex (CPC). In a z/OS or requests and send responses while the listener
OS/390 environment, a physical collection of awaits more connections. The IBM-provided CSKL
hardware that consists of main storage, one or more transaction is a TCP/IP listener, which hands the
Central Processors, timers, and channels. established socket off to another task such as the
BASE24-es SIDC database server. See Client,
Data Access layer (DAL). A BASE24-es System Server.
Interface Services component that provides the
BASE24-es application with a common API to Logical partition (LPAR). A subset of a single
access various databases and file systems, such as system that contains resources (processors,
VSAM. memory, and input/output devices). A logical
partition operates as an independent system. If
Direct Access Storage Device (DASD). A device hardware requirements are met, multiple logical
that allows storage to be directly accessed, such as partitions can exist within a system.
a disk drive.
Managed Address Spaces (MAS). A CICS system
Dynamic Destination Map File (DDMF). A routing that is being managed by CICSPlex SM.
configuration file used by the BASE24-es System
Interface Services Message Delivery Service to Multiregion Operation (MRO). Communication
route asynchronous messages. between CICS systems in the same processor
without the use of SNA network facilities. This allows
Distributed Program Link (DPL). A function of several CICS systems in different regions to
CICS intersystem communication that enables an communicate with each other, and to share
application program to ship LINK requests to resources such as files, terminals, temporary
another application program on a different instance storage, and so on.
of CICS.
Not-on-us card. A card which was not issued by the
Electronic Software Distribution (ESD). financial institution and which is usually routed to a
public card network for authorization. See on-us.
File-Owning Region (FOR). A CICS address space
whose primary purpose is to manage files and Offline authorization. An authorization type
databases. wherein BASE24-es is configured to authorize on-us
Also known as data-owning region. transactions using its own database and scripts. See
on-us, online, online/offline.
Hierarchical File System (HFS). A system for
organizing files in a hierarchy, as in a UNIX system. Online Analytical Processing (OLAP).
z/OS UNIX System Services files are organized in
an HFS.

Online Transaction Processing (OLTP). A type of Server (TCP/IP). A TCP/IP server waits for a
interactive application in which requests submitted connection, typically using the bind(), listen() and
by users are processed as soon as they are accept() sockets API calls. A TCP/IP server typically
received. Results are returned to the requester in a also performs productive work, accepting requests
relatively short period of time. and sending responses. In the world of short-lived
sockets, the client usually also speaks first. The
Online authorization. An authorization type server receives a request, sends a response, and
wherein BASE24-es is configured to route on-us awaits either a new request or a client disconnect. In
authorization requests to a back-end host the world of long-lived sockets common in the world
authorization system. If the back-end system is of financial online transactions, the TCP/IP server
unavailable, requests may be configured to be may be an application-level client, and send the
routed to an alternate authorizer. See on-us, offline, initial request. The BASE24-es TCP/IP Server
online/offline. Communications Handler is a TCP/IP server. See
Client, Listener.
Online/offline authorization. An authorization type
wherein BASE24-es is configured to route on-us System Interface Services (SIS). The group of
authorization requests to a back-end host BASE24-es components that provide a
authorization system. If the back-end system is platform-dependent implementation of a
unavailable, requests may be configured to be platform-independent system services API, allowing
routed to an alternate authorizer. If no alternate the rest of the BASE24-es code to run without
authorizer is configured, or if the alternate is modification on various hardware and middleware
unavailable, BASE24-es may stand in and perform configurations. SIS includes DAL, Time, Timer,
authorization using its own database and scripts. Message Delivery, and similar services.
See on-us, offline, online/offline.
System Initialization Table (SIT). A table
On-us card. A card that is issued by the financial containing parameters used by CICS upon startup.
institution and which is usually to be authorized
locally. See not on-us. SMSVSAM. The name of the VSAM server that
provides VSAM record-level sharing (RLS). See
Record Level Sharing (RLS). See VSAM also VSAM record-level sharing.
record-level sharing.
Synchronous Destination Map File (SYDMF). A
System Authorization Facility (SAF). A z/OS routing configuration file used by the BASE24-es
facility through with programs communication with System Interface Services Message Delivery
an external security manager such as RACF. Service to route synchronous messages.
Static Destination Map File (SDMF). A Synchronous request. A request that is issued
configuration file used by the BASE24-es System followed by an immediate wait for response.
Interface Services Message Delivery Service to Generally acceptable only for lower-latency,
define attributes of various endpoints. highly-reliable servers.
SYStem comPLEX (sysplex). A set of z/OS

systems communicating and cooperating with each
other, through certain multisystem hardware
components and software services, in order to
process customer workloads.
Glossary 225
Transient Data Queue (TDQ). A common
abbreviation for the CICS management module and
its associated functions for intra-partition and
extra-partition data queues.
Terminal-Owning Region (TOR). A CICS region

which owns most or all of the terminals defined
locally. See also application-owning region,
data-owning region.
Temporary Storage Queue (TSQ). A queue of data

items which can be read and reread, in any
sequence. The queue is created by a task, and
persists until the same task, or a another task
deletes it.
Virtual Storage Access Method (VSAM). An

access method for direct or sequential processing of
fixed-length and varying-length records on direct
access devices. The records in a VSAM data set or
file can be organized in logical sequence by a key
field (key sequence), in the physical sequence in
which they are written on the data set or file (entry
sequence), or by relative record number.
VSAM Record-Level Sharing (VSAM RLS). An

extension to the Virtual Storage Access Method
(VSAM) that provides direct record-level sharing of
VSAM data sets from multiple address spaces
across multiple systems. Record-level sharing uses
the z/OS Coupling Facility (CF) to provide
cross-system locking, local buffer invalidation, and
cross-system data caching.
Virtual Telecommunications Access Method

(VTAM). An IBM licensed program that controls
communication and the flow of data in an SNA
network. It provides single-domain, multiple-domain,
and interconnected network capability.
Workload Manager (WLM). A z/OS construct that

provides services to manage workload distribution,
balance workload, and distribute resources.

Related publications
The publications listed in this section are considered particularly suitable for a
more detailed discussion of the topics covered in this redbook.
IBM Redbooks
For information about ordering these publications, see “How to get IBM
Redbooks” on page 227. Note that some of the documents referenced here may
be available in softcopy only.
򐂰 IBM System z9 and zSeries Connectivity Handbook, SG24-5444
򐂰 GDPS Family - An Introduction to Concepts and Capabilities, SG24-6374
Other publications
These publications are also relevant as further information sources:
򐂰 DTR BASE24-es IBM CICSplex Installation Guide
򐂰 Setting up a Sysplex, SA22-7625
Online resources
These Web sites and URLs are also relevant as further information sources:
򐂰 z/OS UNIX ported tools
http://www.ibm.com/servers/eserver/zseries/zos/unix/bpxalty1.html
How to get IBM Redbooks

You can search for, view, or download Redbooks, Redpapers, Hints and Tips,
draft publications and Additional materials, as well as order hardcopy Redbooks
or CD-ROMs, at this Web site:
ibm.com/redbooks

Help from IBM
IBM Support and downloads
ibm.com/support
IBM Global Services

ibm.com/services

Index
A B
Banknet Interface 71
ACI application 17
BASE24-es
ACI Payments Framework 21
architecture 17
acquirer 53, 69–70, 77, 81, 85, 87–88, 139, 162,
flexible infrastructure 16
164–165, 195
introduction 14
acquirer route profile 69–70, 86
reporting 17
active AOR 165
software components 18
activity keypoint 176, 179
VTAM support 35
Adaptors 18
BASE24-es application 21, 45, 120, 124, 126–127,
ALLOWAUTOALT 177, 179, 181–182
150
AOR
alternate routing capabilities 202
architecture 56
maximum availability 124
AOR column 173–174
BASE24-es customer 119–120
AOR failure 106
BASE24-es Dynamic
AOR Message
Destination entry 141
Delivery program 56
Destination Map 139–140
Delivery transaction XDYR 54
Transaction 141, 146
AOR Scope 142, 171, 173
BASE24-es File
AORs 104–105, 107, 116, 118–119, 121, 124, 128,
Partitioning 150, 154
161, 165–166, 175, 193–194, 202, 210, 222
BASE24-es file 129, 146–151, 154
API 48, 142
BASE24-es Integrated Server 43, 47, 56–57,
API flag 142, 171, 173
59–62, 92, 94, 140–141, 185, 187, 194, 198
Appl name 171–173
multiple concurrent instances 43
application code 7, 19
process 140
same set 19
special instance 92
Application Key 141, 169, 171–173
task 222
Application Owning Regions (AOR) 104–107, 160,
transaction 140
162–163, 165–166, 168, 170–171, 173, 176, 184,
BASE24-es internal header 54, 59, 68
190, 192, 194, 198, 203, 222
only messages 60
associated TDQ
BASE24-es Metadata configuration
XMLS server 92
CONFCSV 151, 155
ATM processing 4
BASE24-es metadata configuration
authentication 15
file 185
authorization 15
file CONFCSV 147–148
level 15
BASE24-es product 13–14, 128
Authorization Script
BASE24-es solution 17, 101
OLTP Table 71
BASE24-es system 19, 55, 68, 101, 114, 189, 191,
Table 71
194
Automated Teller Machine (ATM) 1–6, 10, 89, 101
CICS transaction 55, 68
Automatic Restart Manager (ARM) 32, 103, 106,
financial integrity 194
128, 162–164, 166–167, 209
routable endpoint 55, 68
autonomic 9
BASE24-es Task
available AOR 48, 54, 162–163
Monitor 52, 54, 56–57, 106, 190, 192, 194

BASE24-es TCP/IP CICSplex System Manager (CPSM) 124, 134, 141,
client 51, 90 222
communications handler 89, 91 CMAS 160–161
Control File 138 Common Operating Environment (COE) 10
process 138–140 Communication Server 126
server 53, 89–90 Communications Server 29
task 139 Complementary Metal Oxide Semiconductor
BASE24-es transaction 48, 89 (CMOS) 28
BASE24-es workload management computing model 25
facility 222 centralized 25
file 55–56, 68 computing models 25
Business Application Services (BAS) 133 consumer authentication 14
business components 18 Controlled Access Protection Profile (CAPP) 10
business-critical application 31, 113 corresponding entry 139, 150
Coupling Facility (CF) 8, 96–97, 107–108, 124,
176–177, 179, 181, 183, 210–216, 218, 221
C allocated structures 217
Capacity Upgrade on Demand 10
logging information 210
card support 17
CP failure (CF) 206–207, 212, 215–216, 218
CEMT command 145–146
CRM 16
Central Processor (CP) 10, 108, 206
Cross Coupling Facility (XCF) 27
Central Processor Complex (CPC) 108–109, 115,
Cryptographic 63
117, 119, 124, 206, 209
cryptographic considerations 61
CF
Cryptographic Coprocessor 29
failure-isolated 109
Cryptography, provided by System z 28
CF failure 110
CSKL 52
As stated 108–112
customer relationship management (CRM) 16
failed CF support recovery 110
CF maintenance 110, 112
CFRM policy 112, 127, 177, 216–217 D
CICS Automatic 162–163 Data Access Layer (DAL) 19, 130
CICS facility 55, 68 data proximity 26, 38
CICS Link 63, 88, 141, 204 data set name (DSN) 145
CICS program 44, 46, 198 data source 71, 73–75, 77, 81, 83, 85, 87, 150, 153,
CICS region 45, 91, 100, 107, 126–127, 140, 184, 188–189, 206
167–168, 184–185, 209–210 default seq 171, 173
affinity 69 deployment
initialization 57, 69 single-region 46
server class 45 three-tier 48
start 140 two-tier 47
CICS statistic 96–97 DFSMS 30
CICS transaction 46, 60, 139–140 Diebold Number Table (DNT) 74
id 191, 193, 195 Direct Access Storage Device (DASD) 113, 150
XDYR 140 disaster recovery 32
CICS Transient Data Queues 43 Distributed Program Link (DPL) 64
CICSPlex 48–49, 94, 106–107 Domain Name Services (DNS) 29
CICSplex 126, 132–133 Dynamic Destination Map File (DDMF) 134
CICSplex System Management (CPSM) 132–134 dynamic routing
CICSplex System Manager 44 capability 124, 222

program 141 Audit_Store_and_Forward_Config 71
transaction XDYR 48 Authorization_Script 71
Dynamic Transaction 106, 141, 144, 146, 170, 172 Authorization_Script_OLTP 71
dynamic transaction 48 Banknet_Configuration 71
Banknet_Configuration_OLTP 72
Banknet_Group_Timers 72
E Banknet_Institution_Id 72
economies of scale 37
Banknet_Institution_Id_OLTP 72
Electronic Software Distribution (ESD) 129
Banknet_Merchant_Program 72
Encryption PIN Pad
Banknet_Merchant_Program_OLTP 72
public key information 73
Card 72
Encryption PIN Pad (EPP) 73
Card File 69
end-to-end protocol 42, 106, 160
Card_Account 72
end-to-end recovery 162, 164, 194
Card_Verification 72
e-payment system 99, 116
Card_Verification_OLTP 72
Error Table 188–189, 205
Chan_Profile_Instrm_Typ_Rel_OLTP 72
EuronetThe Exchange 4
Chan_Profile_Instrm_Typ_Relation 72
event message 75, 144, 185, 188
Channel_Instrm_Type_Relation 73
EXEC CICS
Channel_Public_Key_Security 73
Link 204
Channel_Security 73
READQ TS 185
Compiler_Message 73
external authorization 63
Context 73
data communication 65
Context File 69
distributed link 64
Context_Configuration 73
WebSphere MQ 67
Context_Configuration_OLTP 73
external authorization system 63–65, 67–68
Contingency_Interface 73
AOR 63
Contingency_Interface_OLTP 73
BASE24-es ix, 41–55, 57–65, 67–69, 87–93,
Currency_Conversion_Rate 74
95–96
Currency_Conversion_Rate_OLTP 74
excessive latency 66
definition 55
One option 63
Diebold_Number_Table 74
external authorization system (EAS) 63
DTR Control file 56, 69
external endpoint 46, 66
Dynamic Destination Map File 55, 68
Event Log 56, 68
F Socket Records file 56
fault tolerance 3, 7, 9, 160 Static Destination Map File 55, 68
files Synchronous Destination Map File 55, 68
Acquirer_Issuer_Relation 69 TCP/IP Configuration File 56
Acquirer_Issuer_Relation_OLTP 70 Terminal Configuration File 55
Acquirer_Route_Profile 70 financial institution 1, 3–4, 14
Acquirer_Txn_Allowed 70 financial transaction 3, 35, 141, 144, 149, 151–154,
Acquirer_Txn_Allowed_OLTP 70 171–172
Action_Code_Extrn_to_Intrn 70 dynamic routing 141
Action_Code_Extrn_to_Intrn_OLTP 70 foundation components 18
Action_Code_Intrn_to_Extrn 71
Action_Code_Intrn_to_Extrn_OLTP 71
Active_Script_Statistics 71 G
GDPS 33
Admin_Card 71
GDPS modes 33
Audit_Store_and_Forward 71
Index 231
GDPS/PPRC 33 J
GDPS/XRC 33 Journal file 79, 95, 147, 152–153
Geographically Dispersed Parallel Sysplex (GDPS) storage capacity 79
33–34 journal profile 78, 80, 153
H L
handshake trn 171, 173 Language Environment (LE) 126
Hardware Management logical file 147, 150, 152
Console 113 Long-running task 44, 56–58
Hardware Security Module (HSM) 61–63 LPAR 27, 105, 108, 113, 115–116, 118–119, 124,
hashing algorithm 147, 153 128, 221
Hierarchical File System (HFS) 129 LPARs 27
HiperSockets 33
host station 134, 141
host-acquired transaction 87, 89 M
mainframe
history 24
I member bank 7
I/O capacity 95, 97 Message Deliver Program 57
I/O device 37 message queuing 42
IGWLOCK00 127, 174, 176–177, 179, 181–182, message queuing (MQ) 43–44, 55, 59–60, 67–68
211, 213–214, 216–217 message type
IMT Interface 76 and/or transaction code 79–80
industry 2 combination 78
instrument type 72, 77, 81 MRO 44
Integrated Server 57 Multiple Image Facility (MIF) 109
Integrated Server (IS) 44, 55, 99–101, 103–113, multiple LPARs 109, 115, 118–120
115–120, 138–140, 146, 186–187, 204 Multiple Region Operations (MRO) 47–49, 94
Integrated Servers 45 multiple TCP/IP
Intelligent Resource Director (IRD) 31 Client 94
IP Client Server 94
task 189, 191, 193
IP client 51–53, 61–62, 87, 89, 94, 171
failure 186, 189–191, 193 N
name suffix 150–151, 154–155
process 163–164, 175
Network Control Program (NCP) 35
queue 191
Non-recoverable TDQs 45
recovery 191, 193
NOTRECOVABL 168
IP handler 44, 46, 52, 54, 66–67, 94, 105, 162–163,
167, 195
symbolic name 52, 54 O
IP process 167, 191, 193 offload processing 25
IP Server OLTP table 70–73, 76–78, 80, 82, 86
failure 160, 163, 186, 191 Online Transaction Processing
failure scenario 192 only table 70, 73, 76–78, 80–83, 86
process 54, 164, 167 Online Transaction Processing (OLTP) 44, 52,
task 191 69–73, 75, 77–79, 81–82, 84, 86, 147, 160, 184
IP tunnel 29 On-us transaction 79, 87, 89
issuer route profile 69–70, 78, 83 openness 36
Operating System 18, 24–25, 108, 129

out-of-balance condition 162–163, 165, 195 RLS 43
RLS access 107, 175–176
RLS mode 107, 212
P routable endpoint 68, 94
packet filtering 29
Routing Region
Parallel Sysplex 8, 10–11, 24, 27, 31–34, 39, 95,
IP client 171
107–109, 112, 124, 126–127, 212
routing region 47, 134, 142, 144, 146, 161–162,
early days 109
164, 167, 171
multiple CICS regions 126
CICS APPLIDs 142
Other CICS applications 107
host simulators 134
z/OS system images 107
RSTR transaction 130–131
Peripheral Component Interface Cryptographic Co-
processor (PCICC) 29
Personal Account Number (PAN) 151–152 S
Personal Identification Number (PIN) 16, 88 Scalability consideration 41, 94
platform-specific components 18 scale in 33
Point-of-Sale (POS) 2–5, 7, 53, 76, 89 scale out 33
POS device 5, 89 SDSF 30
POS network 220 secondary destination 101–102
Processing Code security 27
OLTP table 76 accountability 28
table 76 authentication 28
processing code 69–70, 76, 79, 85 Communications Server 29
first two characters 69–70 cryptographic 28
properly configured (PC) 128–129 Security, provided by z/OS 29
server class 45, 57
server resource 10
Q Service Name 134, 138–141
Quality of Service (QOS) 221
setxcf Start 110, 113
SHaring Control Data Sets (SHCDS) 107
R Single CPC 117, 119
real world Single LPAR 116
end-to-end recoverability 190, 193 single point 8–9, 16, 32, 107, 109, 111–113,
interface 162, 164–165, 195 116–117, 119–121, 221
Record Level Sharing (RLS) 43, 48, 56, 124, parallel processing environment 8
126–127, 222 single region
Redbooks Web site 227 CICS solution 127
Contact us xii CSD 133
Redundant I/O 7 installation 132
regional network 5 single-region deployment 46
relevant subsystem 159, 214 SMP/E 30
reliability 3 SMSVSAM 44, 222
requirements 2 SMSVSAM address space 107, 124
remote endpoint 45, 51–53, 56, 59, 94 SNA LU 54
Resource Access Control Facility (RACF) 28 sockets API 94
Resource Recovery Services 32 SPDH request 84
Resource Recovery Services (RRS) 32, 35 transaction code 85
Resource sharing 26–27 Store_and_Forward (SAF) 71, 85, 95
response time 156 surcharge datasource 85
Index 233
Synchronous Destination Map File (SYDMF) 134, 116, 118–119
202 Three-tier deployment 48
SYSPLEX COMMPLEX 211 Tivoli Enterprise Console 59
Sysplex Distributor 27, 31, 101, 124, 163 Tivoli Workload Scheduler (TWS) 30, 36
System Authorization Facility (SAF) 28 TOR
System Automation 30 architecture 50
System Initialization Table (SIT) 130 TOR architecture 50
System Integration Services (SIS) 129 TOR maintenance 160
System Interface Service (SIS) 51, 138, 200–201, TOR region 161, 163
204 TOR Scope 142, 171
System Interface Services 42 TORs 47–48, 50, 94, 101–104, 106, 116, 118–119,
System Management Facility (SMF) 28 121, 163, 165, 168, 194, 210, 215
System z ix, 7, 9, 23–24, 27–28, 30–35, 37, 39, 41, Total Cost
44 of Acquisition 37
BASE24-es physical architecture 44 of Ownership 37–38
hardware 32, 39 total cost 36
hardware platform 7 Total Cost of Acquisition (TCA) 37
machine 32 Total Cost of Ownership (TCO) 37
platform 37 transaction
processor 126 acquisition 7
server 7 authentication 7
value 7 categories 3
transaction code 69–70, 76, 79, 84
text description 85
T transaction manager 43
Table name 151–153, 155–156
transaction processing 15
target region 47, 141, 144
transactions
CICS APPLIDs 142
routing 14
TCP port 53
transactions per second (TPS) 97
TCP/IP Client 61
two-tier deployment 47
connection 103
TCP/IP client 51–52, 101, 139–140, 160–161
and/or server 139 U
process 139 Unified Modeling Language (UML) 21
TCP/IP communications handlers 45 Unplanned event 100
TCP/IP communications stack 43 Usage file 95, 147–148, 154–156
TCP/IP network 29 user interface 17, 19–20, 90, 92, 124, 126, 129,
TCP/IP protocol 34 131–132, 134–135
TCP/IP server 52–53, 140, 160, 163, 167
symbolic name 53
type 139
V
virtual IP 101, 103
TCP/IP service 91, 133
Virtual Private Networks (VPN) 29
TCP/IP stack 29
Virtual Telecommunications Access Method
TDQ 46, 48, 55, 57–60, 62, 66–67, 92–93, 106
(VTAM) 43
non recoverable 45
virtualization 30
Temporary Storage Queue (TSQ) 61–63
VISA DPS 124, 129
Terminal Owning Region 50
Visa PVV 87
Terminal Owning Region (TOR) 101, 124, 128, 222
VSAM data 107–108
that originated the request (TOR) 101, 103–106,
VSAM file 107, 120, 124, 127, 146–148, 151, 154,

222
RLS access 107
VSAM with Record Level Sharing 43
VTAM Reader/Writer 54–55, 89
VTAM reader/writer
call 55
CICS transaction 55
program 54, 88
task 89
transaction 54
W
WebSphere MQ 43, 55, 59–60, 67–68
workload balancing 11
workload driver 146, 148, 162, 164, 166–167, 171,
175, 184–187, 190, 192–194, 203–204, 210
transactions flow 186, 204
Workload Management 48
Workload Manager (WLM) 31
X
XCF GRPNAME 183
XDYR transaction 66
XMLS TDQ 93
Z
z/OS operating system
IBM System z hardware 44
z/OS strengths 24
z/OS value 23, 38
zAAP 38
zIIP 38
zSeries Application Assist Processor (ZAAP) 10
Index 235
A Guide to Using ACI Worldwide’s BASE24-es on z/OS
(0.2”spine)
0.17”<->0.473”
90<->249 pages
Back cover ®
A Guide to Using
ACI Worldwide’s
BASE24-es on z/OS
Set up and use In this IBM Redbook we explain how to use the ACI
BASE24-es on z/OS BASE24-es product on z/OS. BASE24-es is a payment engine INTERNATIONAL
utilized by the financial payments industry. Failure in a TECHNICAL
High availability financial payments environment is a high-visibility customer SUPPORT
scenarios service issue, and outages at any level have debilitating ORGANIZATION
effects on customer loyalty. The entire payments cycle must
be conducted in near real-time. In such an environment, high
Configure for highest
availability is a mission-critical requirement. We demonstrate
availability how you can achieve a high availability configuration for BUILDING TECHNICAL
BASE24-es on z/OS. We begin by outlining the requirements INFORMATION BASED ON
PRACTICAL EXPERIENCE
of a payments system, and then introduce the structure and
functionality offered by the BASE24-es product. We describe
the strengths and abilities of System z and z/OS, and explain IBM Redbooks are developed by
the technical and physical architecture of BASE24-es on the IBM International Technical
z/OS. We guide you in designing a system layout and in Support Organization. Experts
installing, tailoring, and configuring your workload on z/OS. from IBM, Customers and
Partners from around the world
Finally, we detail the numerous failure scenarios that we create timely technical
tested in order to verify the robustness of the solution. These information based on realistic
scenarios were carefully selected in areas such as data scenarios. Specific
environment, CICS/CICSplex, the BASE24-es application, and recommendations are provided
hardware. Communication was handled by ATM and POS to help you implement IT
solutions more effectively in
device simulators and a Visa network simulator. Note that the your environment.
information in this redbook is specific to BASE24-es release
06.2 and is subject to change in subsequent releases of the
product.
For more information:
ibm.com/redbooks
SG24-7268-00 ISBN 073849593X

ACI Guide

Uploaded by

Copyright:

Available Formats

ACI Guide

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ACI Guide

Uploaded by

Copyright:

Available Formats

Front cover

High availability scenarios

Configure for highest

Alex Louwe Gordon Fleming

A Guide to Using ACI Worldwide’s BASE24-es on

First Edition (August 2006)

© Copyright International Business Machines Corporation 2006. All rights reserved.

Chapter 1. Running a payments system . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Chapter 2. Introducing BASE24-es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

© Copyright IBM Corp. 2006. All rights reserved. iii

Chapter 3. System z and z/OS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Chapter 4. BASE24-es technical architecture on z/OS . . . . . . . . . . . . . . . 41

Chapter 5. Designing the system layout . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

iv A Guide to Using ACI Worldwide’s BASE24-es on z/OS

Chapter 6. Installing BASE24-es on z/OS . . . . . . . . . . . . . . . . . . . . . . . . . 123

Chapter 7. Setting up and running the initial workload . . . . . . . . . . . . . . 137

Chapter 8. Tuning BASE24-es on z/OS . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

Chapter 9. Achieving high availability on z/OS . . . . . . . . . . . . . . . . . . . . 159

Chapter 10. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219

Related publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227

vi A Guide to Using ACI Worldwide’s BASE24-es on z/OS

© Copyright IBM Corp. 2006. All rights reserved. vii

Eserver Parallel Sysplex™ RMF™

The following terms are trademarks of other companies:

viii A Guide to Using ACI Worldwide’s BASE24-es on z/OS

Failure in a financial payments environment is a high-visibility customer service

We begin by outlining the requirements of a payments system, and then

We guide you in designing a system layout and in installing, tailoring, and

This redbook is focused on business verification and can be used by system

BASE24-es is an evolving product. The information provided in this redbook is

© Copyright IBM Corp. 2006. All rights reserved. ix

Alex Louwe Kooijmans is a Project Leader with the International Technical

Edward Addison is a Software Engineer working in Raleigh, NC as the

Claus Koefoed is a Program Manager/zSeries from CompeteCenter in IBM

x A Guide to Using ACI Worldwide’s BASE24-es on z/OS

Bob Spanke is a Master Engineer in ACI Worldwide's Technical Steering group.

Mingming Tao is a z/OS Systems Programming leader with the China

Dore Teichman is an IBM Certified IT Architect and Open Group Certified

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

Paola Bari, Robert Haimowitz

We want our Redbooks™ to be as helpful as possible. Send us your comments

xii A Guide to Using ACI Worldwide’s BASE24-es on z/OS

Chapter 1. Running a payments system

In this chapter we provide an overview of bank payment systems and

© Copyright IBM Corp. 2006. All rights reserved. 1

1.1.1 The requirement for reliability

1.1.2 System availability

2 A Guide to Using ACI Worldwide’s BASE24-es on z/OS

In the event that a financial institution is unable to respond to a transaction within

A financial institution or card issuer with unsatisfactory or poor reliability or poor

System reliability and data integrity

1.1.3 Basic banking transaction categories

An example of the cooperative networks is STAR, owned by First Data

Chapter 1. Running a payments system 3

Other industry competitors in this marketplace worldwide are Plus, Cirrus,

Shared ATM networks provide added availability to financial institution

1.2 How ATM and POS processing is accomplished

4 A Guide to Using ACI Worldwide’s BASE24-es on z/OS

On Us The transaction arrives via a bank-owned ATM or POS