EMC® Data Protection Advisor

Version 6.1 Service Pack 2

Installation and Administration Guide

P/N 302-001-483

REV 06
Copyright ©2005 - 2015 EMC Corporation. All rights reserved. Published in the USA.

Published August 2015

EMC believes the information in this publication is accurate as of its publication date. The information is subject to change without
notice.

The information in this publication is provided as is. EMC Corporation makes no representations or warranties of any kind with respect
to the information in this publication, and specifically disclaims implied warranties of merchantability or fitness for a particular
purpose. Use, copying, and distribution of any EMC software described in this publication requires an applicable software license.

EMC2, EMC, and the EMC logo are registered trademarks or trademarks of EMC Corporation in the United States and other countries.
All other trademarks used herein are the property of their respective owners.

For the most up-to-date support, product, and licensing information to EMC Online Support at http://support.EMC.com.

For technical support, go to EMC Online Support and select Support. There are several options, including one to create a service
request. To create a service request you must have a valid support agreement. Contact your EMC sales representative about obtaining
a valid support agreement or with questions about your account.

Your suggestions will help us to improve the accuracy, organization, and overall quality of user publications. To provide feedback,
please email [email protected].

2 DPA 6.1 SP2 Installation and Administration Guide

Revision history
The following table presents the revision history of this document:

Revision Date Description

01 November 7, First release of DPA 6.1 SP2 Installation and

2014 Administration Guide.

02 November 17, Reorganization of document.

2014 Modification to the following sections:
“System requirements” on page 10
“Permissions” on page 11
“Port settings” on page 11
“Installation and configuration overview” on page 15
“OS resource optimization” on page 18
“Command line installation” on page 23
“Upgrades” on page 32
“Non-anonymous bind” on page 40
“Customizing business information” on page 50
“Discovery overview” on page 74

03 November 24, Modifications to the following sections:

2014 “Loading historical backup job data” on page 71
“Monitoring Oracle RMAN” on page 88

04 February Modification to the following sections:

“Installation and configuration overview” on page 15
“Monitoring VMware environment” on page 104

05 April 30, 2015 Modifications to the following sections:

Table 17, “Server settings,” on page 47
“Datastore Replication administration” on page 53
“Monitoring Microsoft SQL Server” on page 94

06 August 10, 2015 Modification to the following sections:

“Upgrading by using the Patch Installer” on page 34
“Discovering EMC Symmetrix arrays” on page 80

3
4 DPA 6.1 SP2 Installation and Administration Guide
 Contents

Revision history ............................................................................................ 3

Chapter 1 Installation
Overview..................................................................................................... 10
System requirements .................................................................................. 10
Datastore storage................................................................................. 11
Permissions .......................................................................................... 11
Port settings.......................................................................................... 11
Ports required for installing any DPA Application service ....................... 13
Additional Ports required for installing any clustered DPA Application
service .................................................................................................. 14
Ports required for installing any DPA Datastore service.......................... 14
Ports required for installing any DPA Agent service ................................ 15
Installation and configuration overview ...................................................... 15
Configuring virtual infrastructure memory and CPU................................ 17
OS resource optimization ...................................................................... 18
DPA installation .......................................................................................... 18
Installing the Datastore service ................................................................... 19
Installing the Application service................................................................. 21
DPA Agent installation................................................................................. 22
Installing the DPA Agent ........................................................................ 23
Command line installation .......................................................................... 23
Datastore Replication installation................................................................ 28
Configuring Datastore Replication ......................................................... 28
Datastore Replication best practices ..................................................... 29
Application server clustering installation..................................................... 29
Adding an Application object to a cluster during installation ................. 30
Restrictions and recommendations for clustering .................................. 30
Migration from co-hosted to split-installation .............................................. 31
Before you begin: .................................................................................. 31
Procedure ............................................................................................. 31
Upgrades .................................................................................................... 32
Upgrade prerequisites........................................................................... 32
Upgrading by using the Patch Installer .................................................. 34
DPA postinstallation steps .......................................................................... 35
............................................................................................................. 36

Chapter 2

Chapter 3 Administering DPA

Managing licenses ...................................................................................... 38
Expired licenses .................................................................................... 38
Removing a license ............................................................................... 38
User and system settings ............................................................................ 38
User roles and privileges ....................................................................... 38
Viewing default user roles ..................................................................... 39
Creating and adding a new default user................................................. 39
Creating a custom role .......................................................................... 39
Editing and deleting default and custom user roles ............................... 39
LDAP authentication.............................................................................. 40
Digital certificate ................................................................................... 45
Time periods ......................................................................................... 45
Automatic report prioritization .............................................................. 45

DPA 6.1 Installation and Administration Guide 5

Contents

Schedules............................................................................................. 45
System settings..................................................................................... 46
Data Collection Agents .......................................................................... 46
Server ................................................................................................... 47
Replication analysis .............................................................................. 49
Agentless Discovery .............................................................................. 49
Manage Data Collection Defaults........................................................... 50
Customizing business information .............................................................. 50
Using the vtl templates ......................................................................... 50
Importing and exporting custom report templates and custom control
panels................................................................................................... 53
Datastore Replication administration .......................................................... 53
Configuring Datastore Replication after deployment .............................. 53
Carrying out Datastore server failover .................................................... 53
Integrating Slave Datastore after it has been offline .............................. 54
Reconfiguring Datastore servers ............................................................ 55
Stopping Datastore Replication ............................................................. 55
Clustering administration............................................................................ 56
Adding an Application object to a cluster after DPA deployment ............ 56
Removing an Application object from a cluster ...................................... 57
Windows considerations ....................................................................... 57
Backup of the Datastore server.................................................................... 57
Exporting the DPA Datastore server to a file ........................................... 58
Exporting the DPA Datastore server to Pipe............................................ 58
Importing the DPA Datastore server ....................................................... 58
DPA command line operations .................................................................... 59
Sourcing the DPA config file for UNIX users............................................ 59
DPA Server CLI commands..................................................................... 59
Command and option abbreviations ..................................................... 59
dpa application commands................................................................... 60
dpa datastore commands ..................................................................... 65
dpa service commands ......................................................................... 69
Loading historical backup job data ............................................................. 71
Job summary reports ............................................................................. 72
................................................................................................................... 72

Chapter 4 Environment discovery in DPA

Configuring the environment for discovery................................................... 74
Discovery overview................................................................................ 74
Defining objects to be monitored .......................................................... 74
Before you run the Discovery Wizard ..................................................... 75
Configuring application hosts for monitoring......................................... 78
Configuring storage arrays for replication analysis................................. 78
Discovering VNX/CLARiiON arrays ......................................................... 78
Discovering EMC Symmetrix arrays ........................................................ 80
Discovering EMC VPLEX arrays............................................................... 80
Performing hostless discovery on Symmetrix and VNX/CLARiiON........... 81
Configuring storage arrays that use EMC RecoverPoint to gather replication
data ...................................................................................................... 81
Monitoring data protection servers........................................................ 81
Monitoring backup applications configured as part of a cluster ............. 81
Monitoring CA BrightStor ARCserve........................................................ 82
Monitoring CommVault Simpana........................................................... 82
Monitoring EMC Avamar ........................................................................ 82

6 DPA 6.1 Installation and Administration Guide

 Contents

Monitoring EMC NetWorker ................................................................... 83

Monitoring HP Data Protector ................................................................ 83
Monitoring IBM Tivoli Storage Manager (TSM) ....................................... 86
Monitoring Oracle RMAN ....................................................................... 88
Manually configuring DPA Agent to monitor Oracle database and Oracle
RMAN.................................................................................................... 88
Before starting the Discovery Wizard ..................................................... 89
Monitoring Data Domain Backup Enterprise Applications ...................... 89
Monitoring Symantec Backup Exec........................................................ 89
Monitoring Symantec NetBackup .......................................................... 90
Monitoring Symantec PureDisk.............................................................. 91
Monitoring Microsoft Exchange Server................................................... 92
Monitoring Databases ........................................................................... 92
Monitoring Oracle ................................................................................. 92
Monitoring Microsoft SQL Server ........................................................... 94
Monitoring PostgreSQL.......................................................................... 95
Monitoring EMC RecoverPoint ............................................................... 96
Gathering data from operating systems ................................................. 97
Monitoring a Windows host remotely .................................................... 99
Monitoring a host for system data ....................................................... 100
Monitoring tape libraries ..................................................................... 100
Switches and I/O Devices.................................................................... 101
Monitoring Fibre Channel switches...................................................... 102
Monitoring IP switches ........................................................................ 102
Monitoring Xsigo I/O Director .............................................................. 102
File Servers ......................................................................................... 103
Monitoring EMC File Storage................................................................ 103
Backup appliances.............................................................................. 103
Monitoring EMC Data Domain.............................................................. 103
Monitoring Oracle ACSLS..................................................................... 104
Disk management servers ................................................................... 104
Monitoring HP Command View ............................................................ 104
Monitoring VMware environment......................................................... 104
Monitoring backup servers in a Symantec Cluster Server and Microsoft
Cluster Server environment ................................................................. 105
Monitored objects and groups .................................................................. 106
Objects ............................................................................................... 106
Data collection editing for objects ....................................................... 106
Groups ............................................................................................... 106
Configuration group ............................................................................ 107
Object attributes ................................................................................. 107
Smart groups ...................................................................................... 107
Report output objects.......................................................................... 108
Policies and rules...................................................................................... 108
Overview ............................................................................................. 108
Policies ............................................................................................... 108
Analysis policies ................................................................................. 109
Analysis policy rule categories ............................................................ 110
Protection policies .............................................................................. 122
Chargeback policies ........................................................................... 123
Generating events from policies .......................................................... 124
Parameters for generating alerts from scripts ...................................... 125
Rule Template ..................................................................................... 126
Applying a policy................................................................................. 127

DPA 6.1 Installation and Administration Guide 7

Contents

Chapter 5 Uninstallation
Uninstalling the software .......................................................................... 130
Silent uninstallations .......................................................................... 130
Uninstalling through user interface on Windows.................................. 130
Agent-only uninstallation .................................................................... 130

Chapter 6 Troubleshooting
Troubleshooting installation...................................................................... 132
DPA installation failure on Solaris ....................................................... 132
DPA Datastore server on Linux or Solaris failure to start after installation ...
132
DPA web console start failure on Windows Server 2012....................... 132
Postinstallation memory tuning........................................................... 133
Error messages during upgrades ......................................................... 133
Viewing log files........................................................................................ 133
Viewing install log file ......................................................................... 133
Viewing server log files........................................................................ 134
Viewing agent log files ........................................................................ 134
Managing log files............................................................................... 134
Enabling alternative log rotation on VMs running Windows ................. 134
Erroneous memory data in installer log file.......................................... 134
Running a DPA Agent request in debug mode using DPA web console . 134
Default modtest deletion schedule...................................................... 135
Troubleshooting data collection ................................................................ 135
Data collection troubleshooting first actions ....................................... 136
Data collection troubleshooting second actions .................................. 136
Preparing a log file for submission to EMC Support.............................. 136
Troubleshooting client discovery for replication analysis ........................... 137
Client discovery using remote execution.............................................. 137
Client discovery with agent.................................................................. 138
General client discovery ...................................................................... 139
Synchronizing Incorrect Recovery Point times...................................... 140
Troubleshooting report output failure ........................................................ 141
Troubleshooting report generation or publishing problems........................ 141
Synchronizing the system clock................................................................. 142

8 DPA 6.1 Installation and Administration Guide

 Installation

CHAPTER 1
Installation

This chapter includes the following sections:

◆ Overview................................................................................................................. 10
◆ System requirements .............................................................................................. 10
◆ Installation and configuration overview................................................................... 15
◆ DPA installation ...................................................................................................... 18
◆ Installing the Datastore service ............................................................................... 19
◆ Installing the Application service............................................................................. 21
◆ DPA Agent installation............................................................................................. 22
◆ DPA Agent installation............................................................................................. 22
◆ Command line installation ...................................................................................... 23
◆ Datastore Replication installation............................................................................ 28
◆ Application server clustering installation................................................................. 29
◆ Migration from co-hosted to split-installation .......................................................... 31
◆ Upgrades ................................................................................................................ 32
◆ DPA postinstallation steps ...................................................................................... 35

9
Installation

Overview
All of EMC® Data Protection Advisor (DPA) deployments include the following
installations:
◆ DPA Datastore server and a DPA agent on one host
◆ DPA Application server and a DPA agent on another host
When installing DPA the installation wizard takes you step by step through placement of
these components.
Installing the Application and Datastore servers on a single host is not supported. You can
connect multiple Application servers to the same Datastore server, where each additional
Application server is on its own host. You can install additional DPA Agents for system
monitoring and remote data collection. Clustering of the Application server is available to
provide load balancing, quality of service separation, and high availability.
"Migration from co-hosted to split-installation" on page 31 provides information on
migrating from a co-hosted deployment to a split-installation.

System requirements
DPA has the following basic minimum system requirements. The EMC Data Protection
Advisor Software Compatibility Guide provides a comprehensive list of system
requirements.
◆ Split-server installation only
Combined DPA Application and Datastore servers on one host is not supported
◆ 8GB RAM, 2 cores / CPUs per Application and Datastore server
◆ 18GB for the Application server
◆ 20GB for the Datastore Server
◆ The DPA Application server and DPA Datastore servers must not be used to run other
applications. The DPA Application server host and DPA Datastore server host
resources must be dedicated to DPA.
If running DPA in a virtualized environment the allocated CPU and memory must be
reserved for the DPA servers
The DPA installer has a soft threshold of 7892 MB and a hard threshold of 5844 MB.
The soft threshold allows the installation to continue, but the hard threshold does not.
Automatic sizing and tuning of internal DPA resource usage takes place during
installation.If resources (CPU, Memory) are taken away from the installation by other
applications performance of DPA could be adversely affected.
◆ Operating systems
• Support for 64-bit operating systems only
• Microsoft Windows Server 2008, 2008 R2, 2012, 2012 R2
• Red Hat Linux ES/AS 5, 6.0, 6.2, 6.4

10 DPA 6.1 SP2 Installation and Administration Guide

 Installation

• Run the Update Agent (up2date) to ensure that the latest operating system patches
are installed
• SUSE Linux 11 x86
Run the Update Agent (up2date) to ensure that the latest operating system patches
are installed
libaio is recommended to be installed on the system and available in the system
<filepath>LD_LIBRARY_PATH</filepath> for performance gains
The DPA Application server host and DPA Datastore host should not be used to run other
applications. That is, the DPA Application server host and DPA Datastore host resources
must be dedicated to DPA.

Datastore storage
For performance reasons, the installation of the DPA Datastore server on NAS-based file
systems, such as CIFS or NFS shares is not recommended because these file systems
might not have the bandwidth to manage the required I/O.
Although the standard datastore file system layout is adequate for most deployments, you
can distribute different file systems across different file systems to optimize performance
during installation under Advanced installation options.

Permissions
Ensure that you have the following permissions before you install the software to avoid
installation failure:
◆ Windows:
• Administrator privileges (domain or local with full access)
• If User Account Control (UAC) is enabled, use Run As Administrator
◆ UNIX / Linux:
• Root user
• If using security software to manage access to the root account, ensure the
permissions allow the creation of new users after you become root. This must
include the ability to create default home directories for the account to be created.

Port settings
To ensure communication between the DPA Server and DPA Agents, configure the firewalls
in the network to allow communication on these ports, as shown in Figure 1 on page 12.
Additional firewall configuration can be required for other ports depending on what you
plan to monitor. "Environment discovery in DPA" on page 73 provides more information.

System requirements 11
Installation

Figure 1 Ports and protocols in DPA 6.1 and later

12 DPA 6.1 SP2 Installation and Administration Guide

 Installation

Ports required for installing any DPA Application service

The tables below detail the additional ports required on deployment hosts to allow DPA to
function correctly. The ports listed must be able to accept connections and allow
responses back on any established connection. Some network vendors describe such
handshaking communication as Bi-Directional; and such network security devices should
reflect this accordingly.
The following table below details the ports that must be available and free to use on the
local host. These ports are user for intra-application communication and are used for
communication external to the host. These ports are referenced by DPA using a standard
IP address, called the bind address.
Table 1 Ports required for installing any DPA Application service on local host using bind address

Service Port Port Type Description

HTTP 9004 TCP inbound Used for UI, CLI and REST API communications

HTTPS 9002 TCP inbound Used for UI, CLI and REST API communications over SSL

DPA Agent 3741 TCP inbound Protocol port used by DPA to send control data and
manage the DPA Agent

REMOTING 4447 TCP inbound Used for intra-service communication

TXN_RECOVERY4712 TCP inbound Used for intra-service communication

TXN_STATUS 4713 TCP inbound Used for intra-service communication

MESSAGING 5445 TCP inbound Used for intra-service communication

MESSAGING_THROUGHPUT 5455 TCP inbound Used for intra-service communication

The table below details the ports that must be available and free to use on the local host.
These ports are user for intra-application communication and are not used for
communication external to the host. These ports are referenced by DPA using the standard
loopback address: 127.0.0.1

Table 2 Ports required for installing any DPA Application service on local host using loopback
address

Service Port Port Type Description

MANAGEMENT_NATIVE 9999 TCP inbound Used for intra-service communication

MANAGEMENT_HTTP 9005 TCP inbound Used for intra-service communication

OSGI 8090 TCP inbound Used for Service start-up control

System requirements 13
Installation

Additional Ports required for installing any clustered DPA Application service
The table below details the ports that must be available and free to use on the local host.
These ports are used for inter-application node communication and are used for
communication external to the host. These ports are referenced by DPA using a standard
IP address, called the bind address.
Note that UDP Multicast ports are associated with a UDP Multicast Address that is
specified during installation or via the DPA CLI.

Table 3 Additional ports for clustered DPA Application service on local host using bind address

Service Port Port Type Description

55200 UDP Multicast Used for communication between cluster nodes

JGROUPS-UDP

JGROUPS-UDP-LOCK5 5201 UDP Multicast Used for synchronization events between cluster
nodes

JGROUPS-UDP-FD 54200 UDP Multicast Used for resource handling between cluster nodes

JGROUPS-UDP-FD-LOCK5 4201 UDP Multicast Used for resource synchronization between cluster
nodes

Ports required for installing any DPA Datastore service

The table below details the ports that must be available and free to use on the local host.
These ports are user for inter-application communication and are used for communication
external to the host, specifically to the application nodes. These ports are referenced by
DPA using a standard IP address, called the bind address.

Table 4 Ports required for installing the DPA Datastore services on local host using bind address

Service Port Port Type Description

PQSQL 9003 TCP inbound Used for listening to connections to DPA Agents

DPA Agent 3741 TCP inbound Protocol port used by DPA to send control data and
manage the DPA Agent

14 DPA 6.1 SP2 Installation and Administration Guide

 Installation

Ports required for installing any DPA Agent service

The table below details the ports that must be available and free to use on the local host where a DPA Agent is to be
installed. These ports are user for inter-application communication. These ports are referenced by DPA using a standard
IP address, called the bind address.
Table 5 Ports required for installing the DPA Agent service on local host using bind address

Service Port Port Type Description

DPA Agent 3741 TCP inbound Protocol port used by DPA to send control data and
manage the DPA Agent

Installation and configuration overview

The Installation and configuration overview lists the tasks you need to perform for
installing DPA and configuring data monitoring.

Table 6 Installation and configuration overview (page 1 of 3)

Action Comments

1. Set up host computer

Provide at least two hosts for DPA must not be installed on servers already running other applications.
DPA server installation: For installation in a production environment, you need one host for the
One for the initial DPA Application service and a separate host for the Datastore service. EMC
Application server, and one recommends that you use a dedicated server with at least 2GB of
for the Datastore server. temporary space. The EMC Data Protection Advisor Software Compatibility
Guide provides more information.
Provide a host for DPA Agent If the DPA server resides on a Linux or Solaris host and you are performing
installation (optional). client discovery of Windows hosts, at least one DPA Agent must be
installed on a Windows proxy host. "Command line installation" on
page 23 provides more information.

Ensure that DPA and all its If you do not define exceptions for DPA components or associated files,
components are configured the antivirus software occasionally shuts down DPA or quarantines DPA
as exceptions in any antivirus files.
software.

Configure VLAN and shared • Allocate a dedicated VLAN for use by the DPA Application servers.
directory if installing multiple • Provision a load-balancing switch in front of that VLAN and configure an
clustered Application servers. IP address to be used for Application server communications
• Configure a shared directory that will be accessible by all Application
servers. DPA will use this shared directory for writing scheduled reports
and other temporary files that all Application servers need to access.

Check VMware or Hyper-V DPA has been certified to work on a Linux or Windows virtual machine in a
requirements. VMware or Hyper-V environment. The EMC Data Protection Advisor
Software Compatibility Guide provides more information.
Configure virtual "Configuring virtual infrastructure memory and CPU" on page 17 provides
infrastructure memory and more information.
CPU.

Installation and configuration overview 15

Installation

Table 6 Installation and configuration overview (page 2 of 3)

Action Comments

Install the host operating The EMC Data Protection Advisor Software Compatibility Guide lists the
system on the DPA Server and required architectures and patches.
Agent host and then install all
required patches.

Install all required software When monitoring applications or devices remotely, you might need to
on the Agent host after the install additional software on the Agent host. For example, the EMC
DPA 6.1 Application server is NetWorker client must be installed on the Agent host if the Agent will be
ready. used to monitor EMC NetWorker remotely. For more information see
Chapter 3, "Environment discovery in DPA."

If DNS is not enabled in the DPA and SharePoint integration requires the IP address and FQDN to
environment, add the IP enable you to publish reports to SharePoint and to configure the
address and FQDN of the SharePoint port. The SharePoint port is configurable. The default port, if
SharePoint server on the DPA no port is specified, is 80. You can set the port by using a standard URL in
Application server’s hosts the existing URL field in the SharePoint settings dialog. Table 20 on
file. page 49 provides information.

Obtain DPA Licenses

You must know the IP address of the Application server.

For more information on obtaining DPA licenses or types of DPA licenses
available, contact your EMC Account Representative.

• For new non-migrated A DPA license is required to log in to DPA after installation.
installations, obtain DPA For information on required DPA licenses or on purchasing licenses
licenses for all for your DPA installation, contact your EMC sales representative.
components that will be
monitored.
• For upgraded 6.x and
service pack and migrated
5.x installations, existing
licenses will be migrated.

Provide the Solutions Enabler • A minimum of one gatekeeper per HBA per Symmetrix is required.
(SE) licenses. • One SE host can discover all VNX/CLARiiON arrays through the IP
address. For VNX/CLARiiON discovery, EMC recommends installing
Solutions Enabler on the DPA server.
• The EMC Data Protection Advisor Software Compatibility Guide
describes the versions of Solutions Enabler required for storage array
discovery.

2. Install DPA

Install the DPA software. Install the DPA server and Agent according to the installation instructions.
"Configuring virtual infrastructure memory and CPU" on page 17 provides
more information.

3. Configure EMC host array discovery and Solutions Enabler hosts

Configure Symmetrix and "Configuring storage arrays for replication analysis" on page 78 provides
VNX/CLARiiON array for more information. The steps in this section apply only if you are
discovery. monitoring a storage array, database, or Microsoft Exchange Server for
replication analysis.

Provide the SE host used to The EMC Data Protection Advisor Software Compatibility Guide describes
discover Symmetrix or the versions of SE that storage array discovery requires, and the software
VNX/CLARiiON storage arrays. that you must installation the SE host. The host must be able to connect to
the Symmetrix array by a SAN connection. The host must have the TCP port
443 or 2163 enabled for the VNX/CLARiiON connection.

4. Configure the environment for data protection monitoring

16 DPA 6.1 SP2 Installation and Administration Guide

 Installation

Table 6 Installation and configuration overview (page 3 of 3)

Action Comments

Ensure that the ports Chapter 3, "Environment discovery in DPA," lists the protocols and default
between the DPA Agent host DPA ports that communication between the Agent and the monitored
and the monitored server or device or server require.
devices are open and
communication is possible
over the protocol.

Ensure that the DPA Chapter 3, "Environment discovery in DPA," lists the default settings for
credential used to connect to the DPA credentials that are installed with DPA.
the monitored device or
server is sufficient, or have
the new credential details
ready.

Set up monitoring of EMC RecoverPoint agent host and Application host requirements are listed
RecoverPoint, if applicable. in Chapter 3, "Environment discovery in DPA."

Discover and configure • If a remote agent is used to import hosts, the DPA server must be able
Application Host import, if to resolve the agent host.
you are monitoring Microsoft • If application discovery is performed without an agent, "Configuring
Exchange or a database. application hosts for monitoring" on page 78 provides more
information.

5. Define the data protection policies

Prepare the details of the "User and system settings" on page 38 provides more information.
policies that DPA will monitor For replication analysis, the Data protection policy details consist of:
for compliance. • The type or replication, (SRDF/S, SRDF/A, MirrorView, RecoverPoint,
and so forth.).
• Whether the replication is Point-in-Time or continuous.
• The replication target destination.
For data protection reporting, the policies are:
• Chargeback Policies - For financial cost analysis of data protection
operations.
• Protection Policies - To analyze compliance with recovery time objective
(RTO) and recovery point objective (RPO) data protection targets.

Configuring virtual infrastructure memory and CPU

If you plan to deploy DPA in a virtualized infrastructure, perform the following steps:
◆ Ensure that the memory allocated is reserved exclusively for each VM.
◆ Place the DPA Application and Datastore VMs in a resource pool where the resource
allocation shares are set to High. Alternatively, select High Share Allocation for each
individual VM.
◆ Select Thick Provision Eager Zeroed for Datastore disks. Thick Provision Eager Zeroed
disk allocation causes all space to be allocated up-front, and the full disk file is
zeroed before the system is made available for usage.

Installation and configuration overview 17

Installation

OS resource optimization

General tuning
During installation, the installer tunes the DPA Datastore service for the host environment
on which it is being deployed. This tuning assumes that the host is dedicated to DPA and
takes into account resources such as Disk Space, Total Memory, and CPU cores. If other
applications are installed on this host and also using these resources DPA performance
will be degraded. If during the lifetime of the DPA Datastore service any of these physical
resources are increased or decreased, execute the dpa datastore tune command on the
Datastore server. "dpa datastore tune" on page 69 provides more information.

Hardware issues
For deployments where optimal performance is a concern, the type and quality of the
hardware you use for your Datastore server drastically impacts the performance of the
Datastore service. Following are some guidelines to consider when commissioning
hardware for your Datastore server:
◆ Disk types - Fast Ultra-320 SCSI disks are the best option. You can also use the
high-end Serial ATA (SATA) drives. With SATA, each disk is substantially cheaper and
with that you can afford more spindles than with SCSI on the same budget.
◆ RAM - RAM is directly proportionate to the disk cache. This greatly impacts
performance considering that memory I/O is faster than disk I/O.
◆ RAID Controller - Ensure that the RAID controller can keep up with eight fast HDDs.
◆ Disk configuration - The optimum configuration is RAID 1+0 with as many disks as
possible and with your transaction log on a separate disk or stripe. RAID 5 is not a very
good option for databases unless you have more than 6 disks in your volume.
A typical high performance configuration contains:
◆ Two disks in mirror for the OS and DPA Datastore binaries
◆ Two disks in mirror for the Datastore transaction log
◆ Four disks in RAID 10 for the Datastore data files

Note: Make sure that operating system partitions and RAID stripe boundaries are aligned
and undertaken when the partitions are created.

Usually, the performance is better when you have more RAM and disk spindles in your
system. This is because with the extra RAM you will access your disks less. And the extra
spindles help spread the reads and writes over multiple disks to increase throughput and
to reduce drive head congestion.
For production purposes the DPA Application service and the DPA Datastore service
should be placed onto different hardware. Not only does this provide more hardware
dedicated to the Datastore service, but the operating system's disk cache will contain
more Datastore data and not any other application or system data.

DPA installation
The DPA installation involves two stages:

18 DPA 6.1 SP2 Installation and Administration Guide

 Installation

1. "Installing the Datastore service"

and
2. "Installing the Application service"
Installation of the Application service before the Datastore service results in failure of
Application service installation. If you encounter issues during the installation,
"Troubleshooting installation" on page 132 provides information.
The procedures in this section are applicable to new installations. For upgrades from DPA
6.0 and service packs to DPA 6.1, and to install version 6.1 patches, see "" on page 24.
The following procedures explain performing a split installation in a Windows 64-bit
environment.

Installing the Datastore service

Prerequisites
◆ Ensure that you log in as a local administrator or a Domain administrator with full local
access.
◆ If UAC is enabled on a Windows host, start the installer by Run as Administrator.
◆ If installing on UNIX/Linux, ensure that you are logged in as root. Problems have been
experienced with the Datastore server if installation is done after becoming root
through certain SU-type security software; for example, using the sesu command.
◆ If installing Datastore Replication:
• Plan the final deployment topology before beginning installation. Additional
resources are available on the EMC Community Network (ECN) that provide
guidance and best practice for planning your deployment.
• Have all hosts and IP addresses predetermined and handy.
• Ensure that all Datastore server or cluster nodes are using the same IP type of IP
addressing, either IPv4 addresses or IPv6 addresses

Procedure
1. Double-click the DPA server binary to start the installation.
2. Click Next.
3. Read and accept End User License Agreement. Scroll to the end of the agreement to
activate the option to accept the terms of the License Agreement terms. Click Next.
4. In the Installation Options screen, select to install Datastore service, click Next.
5. If you do not perform an advanced installation, click Next and follow the installation
wizard.
To perform an advanced installation, select the Show Advanced Installation Options
checkbox in the Advanced Installation screen, click Next, and follow the installation
wizard.
The Advanced Options are:

Installing the Datastore service 19

Installation

• Do not register DPA services: Prevent the registration of the Datastore service with
the operating system manager. This will prevent the Datastore service from being
started after host reboot. You must use of the DPA Command Line Interface to
install the service with the operating system.
• Do not start DPA services: Prevent the starting of the Datastore services after
installation. Use of the DPA Command Line Interface will be required to start the
service.
• Install with advanced datastore layout: Configure the datastore service with the
required file systems distributed across different disks to optimize performance.
Selecting Advanced Installation Options also enables you to configure Datastore
Replication and select a replication role for this server later on in the installer.
6. When prompted, choose the installation folder. Choose the default location or browse
to another folder location.
7. Review the Pre-Installation Summary, the disk space information in particular, and
click Install.
The installation proceeds.
8. If there is not enough disk space, cancel the installation or choose a different drive to
install DPA on.
9. When prompted, select the IP addresses that the Datastore server should listen on for
connections from the DPA Application server.
10. When prompted, enter the IP address of the DPA Application server that will use the
Datastore server from step 6, and then click Add and Next.
If you are installing a single DPA Application server provide just the one address for
that server.
If you are installing multiple DPA Application servers in a clustered configuration,
enter IP addresses for each DPA Application server.
11. If you are configuring Datastore Replication, select Enable datastore replication and
select the replication role for this server (either MASTER or SLAVE). Click Next.
a. If you selected MASTER as the role for this server, then provide the IP address or
FQDN of the server that will be configured to be a Slave Datastore server.
b. If you selected SLAVE as the role for this server, then provide the IP address or
FQDN of the Master Datastore server.
c. When prompted, specify the name or IP address of the DPA Application server host
to enable communication between the DPA Agent and the Datastore server. By
default, the Agent communicates with the Application server specified earlier in
the wizard. If you are installing multiple clustered Application servers, provide the
IP address for the load-balancing switch that will be placed in front of the
Application server clusters. Click Next.
12. When the DPA Datastore server installation is complete, click Done.

20 DPA 6.1 SP2 Installation and Administration Guide

 Installation

Installing the Application service

Prerequisites
◆ Ensure that the Datastore service option is checked, and that the Datastore service is
running.
◆ If installing Application server clustering:
• Plan the final deployment topology before beginning installation. Additional
resources are available on the EMC Community Network (ECN) that provide
guidance and best practice for planning your deployment.
• Have all hosts and IP addresses predetermined and available, including the IP
address configured for the load-balancing switch that will be placed in front of the
Application servers.
• Ensure that all cluster nodes are using the same IP type of IP addressing, either
IPv4 addresses or IPv6 addresses.
• Specify a common directory that is shared across all nodes. This is the location of
the folder where the reports generated by the DPA Application node are stored.
◆ If installing Application server clustering on Windows, ensure that you:
• Specify the common shared directory as a UNC (Windows Universal Naming
Convention) path.
Note that the DPA installer does not validate the path specified. If you specify an
incorrect path, the installer proceeds, and the resulting path could be an incorrect
symbolic link to itself, or to a folder instead of a file.
• Configure and grant read and write access to a user account (username and
password) to the share specified above. This user account must have the Log on as
a service Windows permissions enabled.

Procedure
The Application service installation process is similar to installing the Datastore service.
1. Double-click the DPA server binary to start the installation.
2. Click Next.
3. Read and accept End User License Agreement. Scroll to the end of the agreement to
enable the option to accept the terms of the License Agreement terms. Click Next.
4. In the Installation Options screen, select to install Application service, click Next.
5. If you do not perform an advanced installation, click Next and follow the installation
wizard.
The Advanced Options are:
• Do not register DPA services: Prevents the registration of the service with the
operating system service manager. This option prevents the DPA services from
being started after a host reboot. You must use the DPA Command Line Interface to
install the service with the operating system.

Installing the Application service 21

Installation

• Do not start DPA services: Prevents the DPA services from being started after
installation. Use of the DPA command line interface is required to start the service.
• Install the DPA services as clusterable: Configures the DPA service to discover and
join any present DPA cluster. "Application server clustering installation" on
page 29 provides more information.
If you would like to add an Application Object to a cluster, select Install the DPA
services as clusterable and follow the steps in the wizard.
At the prompt for a common location for Application servers for reports, ensure
that you specify a common directory that is shared across all nodes. The Shared
Directory for reports is required when you run multiple Application nodes. If
installing on Windows, ensure that you configure the required common and shared
UNC folder and enter the Domain username and password with access to that
specified directory. "Prerequisites" on page 21 provides more information.
The rest of the installation is similar to the Datastore server installation.
6. When prompted, choose the installation folder. Choose the default location or browse
to another folder location.
7. Review the Pre-Installation Summary, the disk space information in particular, and
click Install.
The installation proceeds.
If there is not enough disk space, cancel the installation or choose a different drive to
install DPA on.

Note: A Datastore connection failure error might occur if the relevant firewalls required
to communicate between Application server and the Datastore server are not open.
"Port settings" on page 11 provides port information.

8. In the Connect to Remote DPA Datastore step, enter the IP address for the DPA
Datastore server previously installed.
The installation resumes.
9. When prompted, specify the name or IP address of the DPA Application server host
with which the DPA Agent will communicate. By default the Agent communicates with
the local Application server with IP address 127.0.0.1. In a clustered configuration
provide the IP address of the load balancing switch placed in front of the Application
servers. Click Next.
The DPA Application service installation is now complete.
10. Click Done.
After the installation is complete, start the DPA Server and license the Server. "DPA
postinstallation steps" on page 35 provides more information.

DPA Agent installation

This section describes how to install the DPA Agent using the Agent-only installation
package. It is applicable to new installations.

22 DPA 6.1 SP2 Installation and Administration Guide

 Installation

To upgrade from supported DPA versions, see "" on page 24.

Installing the DPA Agent

The following procedure describes how to install the DPA Agent in a Windows
environment.
1. Double-click the DPA agent binary to start the installation.
2. Click Next.
3. Read and accept End User License Agreement. Click Next.
4. Choose an installation folder and click Next.
5. Verify the Pre-Installation Summary and click Install.
6. Choose the Agent installation options:
• Do not start DPA Agent service - Prevents the DPA Agent service from being after
installation. If you select this option, you must manually start the DPA Agent from
the command line.
If you select Do not start DPA Agent service, click Next.
Type the fully qualified domain name or the IP address of the DPA Server that
communicates with the DPA Agent.
• Agent will be used to monitor Oracle Database: Select this option to monitor an
Oracle database with the DPA Agent.
If you select this option, browse to the directory where the Oracle Database device
driver files can be found by the DPA Agent.
7. Click Next.
8. Type the fully qualified domain name or the IP address of the DPA Server that
communicates with the DPA Agent. Click Next.
9. Click Done to complete the installation.

Command line installation

For a CLI installation use an appropriate command line:
◆ UNIX/Linux
./install.bin [option]

where option is one of the options listed for a silent or an interactive installation in
Table 7 on page 24.
◆ Windows
install.exe [option]

where option is one of the options listed for a silent or an interactive installation in
Table 7 on page 24.

Command line installation 23

Installation

Ensure that you perform the steps provided in "DPA postinstallation steps" on page 35.

Table 7 Installer command line options

Option Description

-? Displays help text

-i [swing | console | Specify the user interface mode for the installer:
silent] swing - Graphical interface
console - console only
silent - no user interaction

-D Shows the installer name-value pairs that might be set on the command line (using the -D option) to
<name>=”<value>" override default installer values, or placed in a response file and used with the -f option.

Quotes must be used around the value.

Example:
-D <variable name>="<value>"

Where:
<variable name> and <value> descriptions are included in the following tables.

Table 8 Datastore installer variables

Variable Name Description Possible Values Default Values

USER_INSTALL_DIR Installation location Valid Path Windows: C:\Program

Files\EMC\DPA
Linux: /opt/emc/dpa

CHOSEN_INSTALL_SET Installation set DS

VAR_INSTALL_SERVICE Advanced option to install TRUE/FALSE TRUE

the Datastore Service

VAR_START_SERVICE Advanced option to TRUE/FALSE TRUE

start/stop the Datastore
service after installation

VAR_DATASTORE_DATA_LOCATION Advanced Datastore Valid Path $USER_INSTALL_DIR$\services\dat

layout option to specify astore\
Datastore server data
directory for optimizing
performance

VAR_DATASTORE_XLOG_LOCATION Advanced Datastore Valid Path $USER_INSTALL_DIR$\services\dat

layout option to specify astore\data\
Datastore server Xlog
directory for optimizing
performance

24 DPA 6.1 SP2 Installation and Administration Guide

 Installation

Table 8 Datastore installer variables

Variable Name Description Possible Values Default Values

VAR_USERNAME (LINUX only) Advanced option to Existing username

specify an existing UNIX
user account to install the
Datastore service

VAR_DATASTORE_BIND_ADDRESSE IPAddress for Postgres to Valid IP Address

S listen on

VAR_DATASTORE_CLIENTS_ADDRES IPAddress of Valid IP Addresses

SES Application server(s) separated by ","
which will connect to the
Datastore service

Table 9 Datastore Advanced options Replication variables

Variable Name Description Possible Values Default Values

VAR_DATASTORE_REPLICATION_ROL Role for Datastore MASTER/SLAVE

E replication

VAR_DATASTORE_REPLICATION_VAL The IP Address of Master Valid IP Address of Master

UE or Slave. If or Slave
VAR_DATASTORE_REPLICA
TION_ROLE is set as
"MASTER", then the
Slave's IPAddress needs
to be entered and vice
versa when
VAR_DATASTORE_REPLICA
TION_ROLE is set as
"SLAVE"

Command line installation 25

Installation

Table 10 Datastore Agent variables

Variable Name Description Possible Values Default Values

VAR_AGENT_APPLICATION_ADDRES DPA Server FQDN or IP Valid IP Address or For multiple application servers and
S Address to manage the hostname for cases where the datastore
Datastore Agent service is communicatong with
In case of linux IPv6, linux IPv6 application server(s), this
<IPv6Address>%<Interface value will be empty. Otherwise the
_Id_Of_Datastore_Agent> default value is the same as
VAR_DATASTORE_CLIENTS_ADDRES
SES

VAR_AGENT_START_SERVICE Advanced option to TRUE/FALSE TRUE

start/stop Datastore
Agent after install

VAR_AGENT_ORACLE_DIRECTORY Advanced option used for Valid Path

monitoring Oracle by the
Datastore agent. Path
where the Oracle
Database device driver
files can be found

Table 11 Application installer variables

Variable Name Description Possible Values Default Values

USER_INSTALL_DIR Installation location Valid Path Windows: C:\Program

Files\EMC\DPA
Linux: /opt/emc/dpa

CHOSEN_INSTALL_SET Installation set APP

VAR_INSTALL_SERVICE Advanced option to Install TRUE/FALSE TRUE
the Application Service

VAR_START_SERVICE Advanced option to TRUE/FALSE TRUE

start/stop the Application
service after installation

VAR_APPLICATION_DATASTORE_AD IP Address of the Valid IP Address where

DRESS Datastore server Datastore service is
installed and running

VAR_ADMIN_PASSWORD DPA application's Set at installation or reset

administrator password using DPA CLI.

26 DPA 6.1 SP2 Installation and Administration Guide

 Installation

Table 12 Application server Agent variables

Variable Name Description Possible Values Default Values

VAR_AGENT_APPLICATION_ADDRES DPA Server FQDN or IP Valid IP Address or 127.0.0.1

S Address to manage the hostname
Application server's Agent

VAR_AGENT_START_SERVICE Advanced option to TRUE/FALSE TRUE

start/stop the Application
server's Agent after install

AVAR_AGENT_ORACLE_DIRECTORY Advanced option used for Valid Path

monitoring Oracle by the
Application server's
Agent.
Path where the Oracle
Database device driver
files can be found

Table 13 Application server Cluster Advanced option variables

Variable Name Description Possible Values Default Values

VAR_APPLICATION_ADDRESS The IP Address used by Valid IP Address

the Application server to
announce itself to other
DPA application nodes.

VAR_APPLICATION_CLUSTER_ROLE Role of the application MASTER/SLAVE

node in a cluster

VAR_APPLICATION_MASTER_ADDRE If Valid IP Address

SS VAR_APPLICATION_CLUST
ER_ROLE="SLAVE", this
value needs to be
entered.

VAR_APPLICATION_REPORT_DIRECT Path to the network Valid path

ORY shared report folder

VAR_APPLICATION_REPORT_USERN The user who will be Existing

AME owning the Application DOMAIN\\Username for
service and has windows, existing
permissions to the shared username for unix
report folder

VAR_APPLICATION_REPORT_PASSW The password of the

ORD (Windows only) above user

Command line installation 27

Installation

Table 14 Standalone Agent Installer variables

Variable Name Description Possible Values Default Values

USER_INSTALL_DIR Installation location Valid Path Windows: C:\Program

Files\EMC\DPA
Linux: /opt/emc/dpa

VAR_AGENT_APPLICATION_ADDRES Server FQDN or IP Address In case of linux IPv6,

SDPA to manage this AgentValid <IPv6Address>%<Interface
IP Address or hostname. _Id_Of_Agent>

VAR_AGENT_START_SERVICE Advanced Option to TRUE/FALSE TRUE

start/stop the Agent after
install

VAR_AGENT_ORACLE_DIRECTORY Advanced option used for Valid Path

monitoring Oracle. Path
where the Oracle
Database device driver
files can be found

Datastore Replication installation

DPA Datastore Replication enables continuous, safe, and reliable replication so that DPA
can maintain a replica copy, or Slave, of the primary Datastore, or Master, for resilience
against a single point of failure. In the event of failure of the Master Datastore, the Slave
can be updated to the Master role using the manual failover command, and the
Application servers are then configured to use this new Master. Reconfiguration should
normally take the same amount of time to take effect as the DPA Application and
Datastore services startup take. "Carrying out Datastore server failover" on page 53
provides more information.
There can be only one Master Datastore per deployment. All Datastores are Masters on
installation. Replication is enabled once a Slave Datastore can communicate with the
Master Datastore. Data starts being replicated when an Application server is started.
You can configure Datastore Replication during the Datastore installation or after
installation.
Ensure that all Datastore nodes are using the same IP type of IP addressing, either IPv4
addresses or IPv6 addresses.

Configuring Datastore Replication

To configure Datastore Replication:
1. Configure the Slave Datastore, either during or after installation.
2. Configure the Master Datastore, either during or after installation.
3. Install and start, if already installed, the Application server.

28 DPA 6.1 SP2 Installation and Administration Guide

 Installation

Configuring Datastore Replication during installation

The option to configure replication is provided at the end of the Datastore installation
process if you have chosen to install with Advanced Options. You can choose any of the
advanced options.
When installing the Slave Datastore the Application server chosen should be the same
one that the Master Datastore is using.
Select Do install with advanced installation options and follow the steps in the wizard.
Step 11 on page 20 provides information.

Datastore Replication best practices

Observe the following best practices for Datastore Replication:
◆ You must restart the Datastore service any time the role between Master Datastore and
Slave Datastore is changed.
◆ Use the replication configuration command dpa ds rep to check the status of
replication. Running the dpa ds rep command on the Master Datastore displays if
replication is on or off and what the Slave Datastore is. Running on the Slave
Datastore tells you what the Master Datastore is.
◆ Before exporting a Datastore, ensure that you create an empty directory on the
Datastore to which to export the Datastore file set. For example, /tmp/export.
◆ Master and Slave Datastores should have the same performance specifications and
be installed on the same operating system and version of DPA.

Application server clustering installation

DPA can be set up in a clustered configuration, with multiple DPA Application servers
working with a single DPA Datastore server. Clustering allows the ability for Application
objects to dynamically start, share workload with other Application objects, and be
stopped as demand decreases.
Clustered Application objects provide many benefits:
◆ Fault tolerance to hardware failure
◆ Load balancing of workload when placed behind a load-balancing switch that you
provide
◆ Workload partitioning
◆ Flexible, green resource management
◆ Reduction of single points of failure
Once multiple Application servers have been configured as a cluster you can start and
stop individual Application servers based on load, such as powering-on additional servers
for end-of-month reporting or other high-usage periods. You can add new objects to
running clusters to improve performance due to load.
Ensure that all cluster nodes are using the same IP type of IP addressing, either IPv4
addresses or IPv6 addresses.

Application server clustering installation 29

Installation

Adding an Application object to a cluster during installation

You can configure an Application object as a cluster during installation using the
Advanced Options.
The Advanced Options are:
◆ Do not register DPA services: suppresses the registration of the service with the OS
thus preventing them from starting after a host reboot
◆ Do not start DPA services: suppresses the automatic start of the services after
installation
◆ Install the DPA services as clusterable: Allows you to configure the DPA service to
discover and join any DPA cluster in your environment.
Select Install the DPA services as clusterable: and follow the steps in the wizard.
If you have already installed DPA and want to use the UI installer to add an Application
Cluster:
◆ uninstall DPA; "Uninstalling the software" on page 130 provides more information.
◆ Reinstall the DPA Application server; choose the Install the DPA services as clusterable
under Advanced Options.

Restrictions and recommendations for clustering

Observe the following restrictions and recommendations when configuring Clusters:
◆ Each cluster must be on its own LAN/VLAN
• Spanning LANs is not possible.
• Clustering is UDP-broadcast based.
◆ Clusters can communicate cross-LAN to Datastore.
◆ Although the Datastore server can be anywhere, there must be only a single cluster on
a LAN. This is because cluster objects discover each other via UDP but are configured
to use a Datastore server by IP.
◆ A physical load-balancing switch should be placed in front of the Application server
cluster to manage the load between DPA Application server objects. The use of
software load-balancing switches is not recommended.
◆ The Patch Installer is used for all upgrades and as well as for applying Generally
Available and non-Generally Available patches. The DPA 6.1 Release Notes includes a
table with supported upgrades and migration paths. "Upgrade prerequisites" on
page 32 provides information about installing patches in a clustered environment.
◆ Any configuration accessible via the DPA web console is stored in the Datastore server
and is accessible cluster-wide. Any configuration operation that requires the use of
the dpa executive utility, such as dpa application promote, is local to the object on
which it was executed. "Adding an Application object to a cluster after DPA
deployment" on page 56 and "dpa application commands" on page 60 provide
information on the dpa application promote command.

30 DPA 6.1 SP2 Installation and Administration Guide

 Installation

Migration from co-hosted to split-installation

Use this procedure to migrate from a co-hosted deployment to separate Application and
Datastore servers.
This procedure requires two new servers. The existing DPA server IP address and name is
reused for the new Application server.
Contact EMC Technical Support for alternative methods of migrating the co-hosted
deployment enable you to use the existing server as either the Application or Datastore
host.

Before you begin:

◆ Back up the existing DPA server, including taking an export of the DPA Datastore
server by running the dpa datastore export command. "Backup of the Datastore
server" on page 57 provides more information.
◆ Back up the export file and the DPA installation directory. "Exporting the DPA
Datastore server to a file" on page 58 provides more information.
◆ Take note of the DPA Server version. On the Application server, run the dpa app ver
command from the services/bin directory.
◆ Shut down the existing DPA server.

Procedure
1. Provision two new servers with a supported operating system. The DPA Software
Compatibility Guide provides information on supported operating systems.
Use the original DPA server hostname and IP address on the new server that you
provision as the Application server.
2. Install the DPA Datastore service on the new server that uses a new IP address and
hostname. "Installing the Datastore service" on page 19 provides information. Note
the following observations:
• Ensure that you install the same version that was on your original DPA server and
that you noted in "Before you begin:" on page 31.
• You might have to perform a full installation and then an upgrade to a patch by
using the Patch Installer, if the version previously installed was a patch version. If
you need upgrade to a patch, "Upgrading by using the Patch Installer" on
page 34provides information.
• If necessary, contact EMC Technical Support to access the same version that was
on your original DPA server.
3. Install the DPA Application service on the new server that has the original DPA server
IP address and hostname. "Installing the Application service" on page 21 provides
information.
• Ensure that you install the same version that was on your original DPA server and
that you noted in "Before you begin:" on page 31.

Migration from co-hosted to split-installation 31

Installation

• You might have to perform a full installation and then an upgrade to a patch by
using the Patch Installer, if the version previously installed was a patch version. If
you need upgrade to a patch, "Upgrading by using the Patch Installer" on
page 34provides information.
• If necessary, contact EMC Technical Support to access the same version that was
on your original DPA server.
4. Verify that the versions on the new Application and Datastore servers match the
versions on the previous cohosted server:
a. On the Application server, run the dpa app ver command from the services/bin
directory.
b. Check that the version output from step a on page 32 matches the version output
from "Before you begin:" on page 31.
5. Confirm that you can login to DPA on the new Application server. "DPA postinstallation
steps" on page 35 provides more information.
6. Import the DPA Datastore backup that you took in "Before you begin:" on page 31. Use
the dpa datastore import command to import the Datastore. "Importing the DPA
Datastore server" on page 58 provides more information.
7. Log in to the DPA web console and confirm that objects are restored. For example,
verify that:
• Objects have been restored by checking that the storage arrays in your original DPA
server appear in the new DPA server Object Library.
• Data collection is running successfully by checking that recent data is visible in
reports that you run since migrating to the new DPA server.

Upgrades
The Patch Installer is used for all upgrades and as well as for applying Generally Available
and non-Generally Available patches. The DPA 6.1 SP1 Release Notes includes supported
upgrades and migration paths.
Note the following:
◆ Upgrade from DPA 6.0 to DPA 6.1SP 1 is not supported.
1. Upgrade to 6.0 latest patch.
2. Upgrade to 6.0 SP1 latest patch.
3. Upgrade to 6.1 SP1 latest patch.
◆ If you have upgraded or migrated to 6.1, delete the browsing history/cache in your
browser before using 6.1.
◆ If you are running DPA 6.1 patch 117, you must upgrade to the latest 6.1 patch before
upgrading to 6.1 SP1

Upgrade prerequisites
◆ Backup the DPA Datasore by using the dpa ds export command."Datastore Replication
administration" on page 53 provides more information.

32 DPA 6.1 SP2 Installation and Administration Guide

 Installation

◆ Stop the DPA Datastore and Application servers. Good practice is to perform a
complete backup of the host running DPA Application and Datastore servers.
◆ Clear the browser cache.
◆ Ensure that you have admin/root privileges.
◆ Ensure that required 6.x package is installed on your system. The upgrade packages
supported on the Patch Installer are not full package releases.
◆ Take note of the DPA 6.x build installed on your system by running dpa app ver and
recording the output. This output is important when verifying package installation.
◆ Stop the DPA Application service. You must stop the Application service because when
the services are on separate machines, the installer cannot stop the services. Then
perform the Patch Installer installation on the Datastore server first, followed by the
Application server.
◆ When applying the Patch installer to clustered environments, stop the DPA Application
service on all objects and then apply the Patch installer first to the Datastore server
and then to one of the Application objects. You must stop the Application service
because when the services are on separate machines, the installer cannot stop the
services. Start the patched DPA Application. Confirm initialization completed and that
you can login to the DPA web console before applying patches to the remaining
clustered Application objects.
◆ If you are upgrading from a release prior to 6.1 SP1 ensure that you review and edit the
retention period on collection requests. Data collection requests contain a different
default retention period in 6.1 SP1 from previous DPA versions.
◆ If you are upgrading from a version prior to DPA 6.1 SP1 and if you are currently using
DPA for RMAN reporting through an existing DPA backup license, contact your EMC
Account Representative for the DPA for Enterprise Applications license. The DPA for
Enterprise applications license allows you to expand the number of RMAN servers
being reported in DPA when you upgrade to DPA 6.1 SP2.

Before upgrading DPA 6.x Datastore server with Advanced Layout on Windows
If you are upgrading a 6.x Datastore server on a Windows system using the Patch Installer
with Advanced Layout, you must run two extra commands manually to allow the local
service user access to the /data and /pg_xlog directories which are located outside of the
dpa installation root.
You must have administrative privileges to perform this procedure.
The 6.1 SP2 the Datastore service does not run under the local system user, but rather
under the local service user. The /data and /data/pg_xlog directories are actually
symlinks leading to other directories outside the dpa installation directory. This pertains
to problem numbers DPA-25875 and DPA-26033.
1. Open a command prompt, running with administrative privileges.
2. Run the commands:
icacls "<full path to data location>" /t /grant "LOCALSERVICE:(OI)(CI)F"
icacls "<full path to pg_xlog>" /t /grant "LOCALSERVICE:(OI)(CI)F"

Where:

Upgrades 33
Installation

• <full path to data location> is the path name to the /data directory, including the
drive letter
• <full path to pg_xlog location> is the path name to the /pg_xlog directory, including
the drive letter

Upgrading by using the Patch Installer

Ensure that you run that the Patch Installer as admin/root user.
1. Close the DPA 6.x web console.
2. On all clustered and non-clustered Application nodes:
a. Shut down the Resident Agent.
b. Stop the Application service.
3. On the Datastore node, shut down the Resident Agent and stop Datastore service.
4. On all Application node, back up the installation directory.
5. On all the Datastore node, back up the contents of the installation directory.
6. If you are using Advanced Datastore layout:
a. Ensure that you back up the directories that are symlinked to external locations.
b. Open a command prompt, running the account with administrative privileges.
c. Run the commands:
icacls "<full path to data location>" /t /grant "LOCALSERVICE:(OI)(CI)F"
icacls "<full path to pg_xlog>" /t /grant "LOCALSERVICE:(OI)(CI)F"

Where:
• <full path to data location> is the path name to the /data directory, including the
drive letter
• <full path to pg_xlog location> is the path name to the /pg_xlog directory, including
the drive letter

7. If you have a clustered environment and are using a shared report directory, back up
the shared report directory.
8. Patch the Datastore server. Follow the installation steps as directed in
InstallAnywhere. Ensure that the existing DPA installation directory is specified
correctly on the Patch Installer. You must install the DPA 6.1 Patch Installer package in
the same installation directory as your existing DPA package.
9. Patch a single Application object. Follow the installation steps as directed in Install
Anywhere. Ensure that the existing DPA installation directory is specified correctly on
the Patch Installer. You must install the DPA 6.1 upgrade package in the same
installation directory as your existing DPA package.
10. Wait for the files to be deployed under the <install_dir> services\applications folder.
"DPA postinstallation steps" on page 35 provides more information.

34 DPA 6.1 SP2 Installation and Administration Guide

 Installation

11. Confirm the upgrade package installation of the Application object. Compare the build
numbers shown when you run dpa app ver to the build numbers you noted before the
upgrade. "Upgrade prerequisites" on page 32 provides information.
If the output indicates any version other than DPA 6.1 SP2, contact EMC Technical
Support and reference DCE-1742.
12. Perform the steps provided in"DPA postinstallation steps" on page 35.
13. Patch the DPA Agent without uninstalling the existing Agent package. Follow the
installation steps as directed in Install Anywhere. Agent Patch Installer is displayed in
the banner. You must patch the DPA Agent on the DPA Datastore and DPA Application
servers.
Ensure that the existing DPA Agent installation directory is specified correctly on the
Patch Installer. You must install the DPA 6.1.2 Agent Patch Installer package in the
same installation directory as your existing DPA Agent package. Advanced options and
the IP/FDQN from the existing DPA Agent are carried over to the 6.1.2Agent.
When the Agent Patch Installer runs, it stops the Agent Service, overwrites all the
Agent files, except for the configuration files which remain unchanged, and then
restarts the DPA Agent service.
14. If working in a clustered environment, repeat step 10 through step 14 for remaining
Application objects.

DPA postinstallation steps

The first time you access the DPA web console, you might see a message indicating that
the DPA Server is starting. This message indicates initialization is in progress, which can
take approximately 10 minutes to complete.
During the initialization time, DPA is creating the database schemas, tables, views, and
the DPA Datastore. It also creates the various system reports and dashboards templates,
the default system users, Analysis Engine Rulesets, and various other default and initial
objects. Your network connection time affects the speed at which all these actions
complete.
Ensure that you perform the following steps after installing DPA.
1. If you have upgraded or migrated to 6.1, delete the browsing history/cache in your
browser before using 6.1.
2. (Optional) Verify DPA installation. Before the DPA console allows login, DPA must
complete various initialization tasks. Use the following steps to verify whether
initialization is still in progress or is complete:
a. Run dpa app status from a command prompt window.
Alternatively: On the DPA Application server, go to
<install_dir>\services\applications.
b. Check the *.rar ; *.ear, and *.war files for *.deployed, *.isdeploying, or .failed
extensions.
– If files have an extension of *.isdeploying, then server initialization is still in
progress.

DPA postinstallation steps 35

Installation

If the files from <install_dir>\services\applications show the *.isdeploying

extension for an unusually long time, for example, two hours, contact EMC
Technical Support.
– If files have an extension of *.deployed, then server initialization is complete
and you can login to the DPA web console.
– If files have an extension of *.failed, then server initialization failed; contact
EMC Technical Support.
3. Start the web console to verify successful DPA installation.
All DPA services must be running when you start the web console. The Adobe Flash
plugin in your web browser is required to start the web console.
a. Start a browser and connect to DPA Server over https on port 9002. Ensure that all
pop-up blockers are disabled. For example:
https://<server_name>:9002

where <server_name> is the name or IP address of the server or localhost.

b. Type the username and password. Use administrator for the username and
password for a new installation. Username and password fields are case-sensitive.
c. Click Login
4. Add licenses to the DPA server:
• If this is the first time you are running DPA, you are prompted to add a license.
Select Admin > System > Manage Licenses to add, view, and delete licenses.
• If you are migrating from DPA version 5.x to version 6.1SP1, the existing licenses
are migrated with your configuration and data.
For information on purchasing licenses for your DPA installation, contact your EMC
sales representative.
5. (Optional) Change the default password.

36 DPA 6.1 SP2 Installation and Administration Guide

CHAPTER 2

Administering DPA

This chapter includes the following sections:

◆ Managing licenses .................................................................................................. 38
◆ User and system settings ........................................................................................ 38
◆ Customizing business information .......................................................................... 50
◆ Datastore Replication administration ...................................................................... 53
◆ Clustering administration ........................................................................................ 56
◆ Backup of the Datastore server................................................................................ 57
◆ DPA command line operations ................................................................................ 59
◆ Loading historical backup job data.......................................................................... 71

37
Administering DPA

Managing licenses
Expired licenses
If a license expires, a license violation warning appears in the report title for reports run
from all objects enabled by the expired license. In addition, new objects cannot be added
in the web console for module components enabled by an expired license.

Removing a license
Removing a license causes a license violation warning to appear when running reports
against objects for that license. New objects of that type cannot be added in the web
console until a replacement license is supplied.
If you are using temporary licenses that have an expiration date, the License Expiration
dialog appears to notify you of the expiration of your temporary licenses. Permanent
licenses do not display.
To add new licenses, to go Admin > System and then click Manage Licenses.
To disable the automatic temporary license expiration dialog pop-up at login, go to User
Properties > Show License Expiration and uncheck the box.

User and system settings

User roles and privileges
Roles are used to handle the privileges allowed for users. Users gain their privileges by
being assigned to the appropriate role.
A user role is a way to limit the functionality and web console options granted to groups of
DPA users, by assigning only specific privileges. User role privileges, such as Create User
and Delete User, assign roles to a user and limit user roles.
Table 15 on page 38 explains the users and user roles created by default.

Table 15 Users and user roles

Users and user roles Default password Privileges

Administrator administrator Can perform all configuration and reporting

functions.

Application owner application owner Can perform all reporting functions and modify
credential settings.

Engineer engineer Can perform all reporting functions and most

configuration functions.
Engineers cannot create or modify users or user
roles, or modify system settings.

User user Can perform reporting functions only.

38 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

Four roles are supplied by default with DPA: Administrator, Application Owner, Engineer,
and User. These each have predefined permissions and settings which cannot be
changed. You can create users, edit profiles to customize preferences and web console
appearance, and delete users.

Viewing default user roles

1. Go to Admin > Users & Security > Manage Roles.
A list of the default users (administrator, application owner, engineer, user) is
displayed, together with any new users added since installation.

Creating and adding a new default user

1. Go to Admin > Users & Security > Manage Roles.
2. Click Create User.
Alternatively, select an existing user and click Save As.
3. In the Create User Properties tab, update the information in the respective tabs:
• In the User Properties tab, specify the name, logon name, role, authentication type
and password. Click OK after making changes between the respective tabs.
• In the Preferences and Appearance tabs, assign preferences and appearance
settings. Note that the role you assign to the user will determine which areas of
DPA they can access.
• If the user is to be authenticated via LDAP, choose the LDAP authentication type.
• Click OK after making changes between the respective tabs.
4. Click Save and Close.

Creating a custom role

You can create a new custom role with different permissions and settings.
1. Go to Admin > Users & Security > Manage Roles.
2. Either click Create Role, or choose an existing role and click Save As.
3. In the User Role Properties window,
• Type a name for the custom role in the Name field
• Type a description for the custom role in the Description field
• Set the Privileges, Accessible Groups, Dashboards, and Menus for the custom role.
Click OK after making changes between the respective tabs.
• Click OK to save the changes in the User Role Properties window.

Editing and deleting default and custom user roles

Observe the following considerations regarding editing and deleting user roles:
◆ You can edit and delete any of the users except for the Administrator.

User and system settings 39

Administering DPA

◆ You cannot delete default roles.

◆ Before deleting custom roles, ensure that any users allocated to that role have
alternative roles assigned to them; otherwise it is not possible to delete the role.
1. Go to Admin > Users & Security > Manage Roles.
2. Select the role and click Edit or Delete, as applicable.
The DPA online help system provides information on user properties and preferences.

LDAP authentication
DPA supports Microsoft Active Directory and OpenLDAP as LDAP servers. DPA allows the
integration of a Lightweight Directory Access Protocol (LDAP) server in the environment so
that user passwords do not need to be stored in the DPA database. When a user logs in,
DPA communicates with the LDAP server, verifies that the user exists, and authenticates
the password against that user stored in LDAP.
To enable LDAP authentication, select Admin > Users & Security > Manage External
Authentication.

Anonymous bind
When binding anonymously to the LDAP server, DPA connects to the LDAP server without
having to connect as a specific named user. This allows basic user authentication.
When authenticating a user using an anonymous bind, DPA must pass the Distinguished
Name (DN) of the user to the LDAP server to authenticate it. DPA must determine the full
DN of the user when the user logs in.
An example of a DN for a user is:
CN=Mark,CN=Users,DC=eng,DC=company,DC=com
You must specify the DN in the External Name field when creating an LDAP authenticated
user. When the user logs in to DPA, the username is authenticated against the LDAP
server.
The disadvantage of using the anonymous bind method of authentication is that every
user you want to log in must be predefined in DPA along with the DN of each user. It is
sometimes difficult to determine the DN of each user and it is possible to make typing
mistakes that might cause authentication to fail. The alternative is to use a
non-anonymous bind.

Non-anonymous bind
If you use a non-anonymous bind, DPA connects to the LDAP server as a named user
before attempting to authenticate users who log in to the application. By connecting as a
named user, DPA has access to search the directory for users. You do not need to specify
the DN of each user who you want to have access to the application when creating user
definitions. The username and password field in DPA LDAP configuration must be that of a
a user with read access to the base level DN.
You must specify a Base Name when configuring LDAP authentication. This is the point in
the directory where DPA starts searching for a user. When a user logs in, DPA searches the
Base Name point and below in the tree in an attempt to locate the user. This means that
you do not need to know the full path to each user inside the directory.

40 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

You must also specify an Identification Attribute in DPA LDAP configuration. This is the
attribute that DPA matches the provided login name against when searching the directory.
For example, with Active Directory integration this is often set to sAMAccountName, which
is the typical login name property for a user.
When DPA finds the correct user in the directory, DPA attempts to authenticate with the
password provided at login.

Auto-Login
Another advantage of using a non-anonymous bind to connect to an LDAP server is the
ability to use the Auto-Login feature. With Auto-Login, you do not need to define user
entries in DPA for each user requiring access to the application. Access is granted based
on the existence and group membership of a user in the LDAP directory.

Note: The Auto-Login feature is supported when connecting to a Microsoft Active Directory
server only. Do not attempt to use this feature with an LDAP server other than Microsoft
Active Directory.

If there are members of multiple groups in LDAP, and multiple groups have been defined in
the Group Mapping table, they are granted the role that is mapped to the first group in the
list of which they are a member. The groups that map to a role with greater permissions
should be highest in the list.
If using the Auto-Login feature, do not create a user definition in the application before
attempting to log in. The first time the user logs in, DPA authenticates the username and
password, and then attempts to determine the role that should be granted based on the
values specified in the Default User Role and Group Mapping fields. If a role cannot be
determined, the user is not granted access to the application. If the role can be
determined, DPA automatically creates a user definition and the user is permitted to log
in.
Users who are removed from LDAP are no longer granted access. If a user’s group
membership is changed in LDAP, the role is re-evaluated at next login and is updated.

Connecting to an LDAP server using SSL

DPA supports the ability to authenticate to an LDAP server using SSL in environments in
which the LDAP implementation uses Microsoft Active Directory. DPA supports connecting
to an LDAP Server using SSL Windows, Linux, and Solaris DPA servers.In this
configuration, Active Directory must be configured to accept connections using SSL.

LDAP Examples
The following examples show various ways to configure DPA LDAP integration with a
Windows Active Directory server. These examples assume the following AD configuration:
◆ LDAP Server: ad.lab.emc.com
◆ LDAP Domain Administrator Login:
CN=admin123,OU=MSP01,DC=lab,DC=emc,DC=com
◆ LDAP OU's where DPA users may exist:
◆ OU=Admin,OU=MSP01,DC=lab,DC=emc,DC=com
◆ OU=Users,OU=MSP01,DC=lab,DC=emc,DC=com

User and system settings 41

Administering DPA

◆ OU=Users,OU=MSG02,DC=lab,DC=emc,DC=com
All DPA Users are in organizational units below lab.emc.com
Example User:
◆ LDAP Object: CN=John Smith, OU=Admin,OU=MSP01,DC=lab,DC=emc,DC=com
◆ LDAP Login Name: smithj (the sAMAccountName property)
◆ LDAP Group Membership:
◆ CN=msp01admin, OU=MSP01,DC=lab,DC=emc,DC=com
• CN=exchadmin, OU=MSP01,DC=lab,DC=emc,DC=com
• CN=dpaadmin, OU=MSP01,DC=lab,DC=emc,DC=com
◆ DPA Role: Administrator

Example: Configuring DPA to authenticate with Anonymous Bind

1. Go to Admin > Users & Security > Manage Roles.
2. Click Create User.
3. Enter the following values in the User fields:
• Name: J Smith
• Logon: Namesmithj
• External Name: CN=John Smith, OU=Admin,OU=MSP01,DC=lab,DC=emc,DC=com
• Role: Administrator
• Authentication Type: LDAP
4. Click OK.
5. Go to Manage External Authentication > Configuration.
6. Enter the following values in the Manage External Authentication > LDAP/AD tab fields:
• Use LDAP Authentication: Selected
• Server: Serverad.lab.emc.com
• Use SSL: Selected
• Port: 636
• LDAP Version: 2
• Base Name: DC=lab,DC=emc,DC=com
• Identification Attribute: SAMAccountName
• Anonymous Bind: Selected
• User Properties: none selected- no username, no password
• Enable Auto Login: Not selected
7. Click Test User.
8. Enter the following in the Test User fields:

42 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

• Username:CN=John Smith, OU=Admin,OU=MSP01,DC=lab,DC=emc,DC=com

• Password: <smithj_AD_Password>

Example: Configuring DPA to authenticate without Anonymous Bind, without Auto-login

1. Go to Admin > Users & Security > Manage Roles.
2. Click Create User.
3. Enter the following values in the User fields:
• Name: J Smith
• Logon: Namesmithj
• External Name: smithj
• Role: Administrator
• Authentication Type: LDAP
4. Click OK.
5. Go to Manage External Authentication > Configuration.
6. Enter the following values in the Manage External Authentication > LDAP/AD tab fields:
• Use LDAP Authentication: selected
• Server: ad.lab.emc.com
• Use SSL: Selected
• Port: 636
• LDAP Version: 2
• Base Name: DC=lab,DC=emc,DC=com (DPA searches in all OUs below this level)
• Identification Attribute: AMAccountName
• Anonymous Bind: Not selected
• User Properties:
– Username: CN=admin123,OU=MSP01,DC=lab,DC=emc,DC=com
– Password: <admin123_domain_password>
• Enable Auto Login: Not selected
7. Click Test User.
8. Enter the following in the Test User fields:
• Username: smithj
• Password: <smithj_AD_Password>

Example: Configuring DPA To authenticate without Anonymous Bind, with Auto-login

1. Go to Manage External Authentication > Configuration.
2. Enter the following values in the Manage External Authentication > LDAP/AD tab fields:
• Server: ad.lab.emc.com

User and system settings 43

Administering DPA

• Use SSL: Selected

• Port: 636
• LDAP Version: 2
• Base Name: DC=lab,DC=emc,DC=com (DPA searchrd in all OUs below this level)
• Identification Attributes: AMAccountName
• Anonymous Bind: Not selected
• User Properties
– Username: CN=admin123,OU=MSP01,DC=lab,DC=emc,DC=com
– Password: <admin123_domain_password>
• Enable Auto Login: Selected
3. Click Test User.
4. Enter the following in the Test User fields:
• Username: smithj
• Password: <smithj_AD_Password>
5. Enter the following in the DPA Auto-Login Configuration fields:
• Default User Role
– If set to None , then only users in an LDAP group that is configured in DPA group
mapping can login.
– If set to a specific role, then users that are not in any DPA group mapping
groups will be allowed to login and receive this role.
• Enable Group Mapping: Selected
• Group Base: OU=MSP01,DC=lab,DC=emc,DC=com (DPA searches in all OUs below
this level)
• Group Attribute: CN
• Group Member Attribute: member
6. Enter the following in the Auto-Login Group Mapping fields:
• LDAP Group Name: dpaadmin
• User Role: Administrator
When configuring DPA LDAP with Auto-Login, do not create DPA users to match LDAP
users. Based on the above settings, when a DPA user first logs in to DPA using AD
credentials, DPA searches for the login name the user specifies and attempts to match
it to sAMAccountName (IDENTIFICATION ATTRIBUTE ) in DC=lab,DC=emc,DC=com
(BASE NAME) and in OUs below that level of the LDAP tree.
If the DPA finds the user, DPA checks the LDAP users group membership by searching
OU=MSP01,DC=lab,DC=emc,DC=com (the GROUP BASE) for a CN attribute (Group
Attribute) where member (the Group Member Attribute) matches dpaadmin.
If all the fields match, then DPA creates a new DPA user to match the LDAP user.

44 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

Digital certificate
DPA uses a self-signed digital certificate for identification and encryption. This certificate
expires in five years. To use a certificate that is signed by your own Certificate Authority,
you must:
1. Export a certificate request (CRT) file.
2. Ask your Certificate Authority administrator to create a signed certificate in Base-64
Encoded X.509 (CER) format.
3. Import the signed certificate into DPA.
The Online Help provides more information on adding digital certificate.

Time periods
When you run a report or create a scheduled report, you must decide the period of time
over which the report is run, for example right now or last week. Several predefined time
periods are provided by default and you can create custom time periods.
To create a custom time period, select Admin > System > Manage Time Periods.

Automatic report prioritization

The default number of reports to run concurrently per DPA Application server is 10. You
can configure the default settings. The maximum number of reports to run concurrently
per DPA Application server is 50; the minimum number is 2.
DPA automatically queues reports that are scheduled to run concurrently or that are
running concurrently, and automatically retries reports when the previously scheduled
reports have been run. Additionally, any reports that you can initiate from the web console
take precedence over automated scheduled reports running from the server, including
testing a scheduled alert.
In addition to giving priority to reports run from the web console, there is also a 30%
minimum fixed concurrent space reserved for these reports on the server. For example, if
the concurrency set is 10, three concurrent execution spaces on the server are reserved for
web console reports. Hence, there can be three or more out of a maximum of 10 web
console reports running at a particular instant. There can be only seven scheduled reports
which can run concurrently.
To configure concurrent report settings, select Admin > System > Configure Report Settings
> Concurrency.

Schedules
Schedules are used to define when to run a scheduled report or generate a dashboard
view block, or to define the backup window specified in the Protection Policy. Several
predefined schedules are provided by default and you can also create custom schedules.
A schedule is made up of components that define when each schedule produces certain
results or runs certain reports. The Schedule Editor provides two ways to create schedules:
◆ Basic editor - allows you to create schedules on a weekly basis only and edit the day
and time of the schedule.

User and system settings 45

Administering DPA

◆ Advanced editor - allows you to create more complex schedules by manually editing
the schedule parameters.
Schedules created in the basic editor can be edited using the advanced editor. However,
schedules created and saved in the advanced editor cannot be edited in the basic editor.
To create a schedule, select Admin > System > Manage Schedules.

System settings
This section describes how to modify the default system settings for DPA Agents, the
server, and the Datastore server. To view or edit system settings, select Admin > System >
Configure System Settings.

Data Collection Agents

Agent settings control the behavior of the Agent processes. Table 16 on page 46 describes
each Agent setting.

Table 16 Agent setting

Setting Description

Data Collection Agent Port Port on which the data collection agent listens for requests

Concurrency Maximum number of threads the data collection agent uses to

gather data. The default is five

Log Level Verbosity level when the data collection agent writes to the log
file. For example, selecting Fatal writes only critical errors to
the log file

Max Log File Size (MB) Maximum size to which a log file can grow before the creation
of a new log file (in MB). To set no limit for the size of the log
file, set this value to 0

Max Number of Log Files Maximum number of log files maintained on the system. If a
new file is created because the maximum file size of the
current log file is exceeded, the oldest log file is removed.

Max Forward Queue Length Maximum number of requests stored by the agent locally if the
Server is offline.

Max Forward Queue Size (MB) Maximum total size of all requests stored by the DPA data
collection agent locally if the Server is offline (in MB).

Enabled Enables data collection agent on the host.

Reload Data Collection Agent Allows you to manually reload the data collection agent. This is
done automatically when configuration changes are made in
the DPA web console that affect a data collection agent.

Remove Data Collection Agent Removes the selected data collection agent.

Make Agent Default Makes the selected data collection agent the default host.

46 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

Additional server settings configure global agent and email settings. The server settings
are described in Table 17 on page 47.

Table 17 Server settings

Setting Description

Binary Multiplier Switching this global setting on, defaults all Agents to use the
binary multiplier. Binary multiplier converts all incoming data
as 1024 KB= 1MB. Applies to NetWorker agents only where
the incoming data from Backup server is converted as 1000 KB
= 1MB. Binary Multiplier is ignored when monitoring other
applications.

Timeout(s) Time out setting that the server uses when talking to the
agent. The default is 120 seconds.

Mail Server Hostname Mail server to which email messages are forwarded when sent
from DPA.

Mail From Address E-mail address assigned to email messages sent from DPA.

Mail Server Port Mail server port number.

Global Logging Settings Global logging settings for the Analysis Engine, Configuration,
Listener, Publisher, Recoverability Analysis, Reporter, and
REST API. Settings can be INFO, DEBUG, WARN, ERROR, and
FATAL.

Global SharePoint Server SharePoint Server settings required to publish the report or
Settings control panel to Microsoft SharePoint as a CSV, image (png),
PDF, HTML, or XML file.
Ensure that you create the "Shared Documents" path and
folder for publication. For example:
http://sharepoint-2013/sites/demo/Shared Documents

Data Deletion Schedule to delete data gathered from your environment. The
default is 9 a.m. to 5 p.m. every day.

Server

Data Deletion Schedule

DPA 6.1 implements a default data deletion schedule for collected data and
system-generated data. Collected data is the data gathered by the configured requests
within Manage Data Collection Defaults. System-generated data is the data generated by
the system processes, such as log messages, histories of reports, and alerts.
When data exceeds the retention period then the data is eligible for deletion. This data is
then purged based on the data deletion schedule. Any unprocessed items remain in the
queue until the next scheduled start time, at which point deletion of data continues.
You cannot delete a schedule that is currently used for scheduling a collected data
deletion job. An error message is displayed if you attempt to do so.
Collected and system-generated data that is deleted is tracked in the server.log. For
example:
Deleted 10 rows from table host_config

Deleted 10 rows from Request History

Deleted 10 rows from reportlogentry

User and system settings 47

Administering DPA

Deleted 10 rows from dpa_request_statistics

Deleted 10 rows from reporterjob

The default data deletion schedule is from 9:00 a.m. to 5:00 p.m daily.

Configuring Data Deletion Schedule

You can configure and specify a new schedule for use in Schedule Properties.
To configure data deletion, select Admin > System > Configure System Settings > Server >
Data Deletion. The DPA Online Help provides more information.

Default retention periods

Table 18 on page 48 provides information on default collected data retention periods.

Table 18 Default collected data retention periods

System information Default retention period

Configuration data 365 days

Status data 90 days

Performance data 30 days

Job data forever

Occupancy data 365 days

Default collected data retention periods are user-configurable within the Manage Data
Collection Defaults.
Table 19 on page 48 provides information on default system-generated data retention
periods. Default system-generated data retention periods are not user-configurable.

Table 19 Default system-generated data retention periods

Policy Default retention period

alerts (analysisalert table) 365 days

report history (reporterjob table) 365 days

agent error log entries (reportlogentry table) 14 days

request statistics (dpa_request_statistics table) 28 days

48 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

Replication analysis
The replication analysis settings configure client-server time difference. The settings are
described in Table 20 on page 49.

Table 20 Replication analysis settings

Setting Description

Client-server Time Difference The time offset you calculated Client-server Time Difference (in
seconds or minutes).

Display dirty recovery points in The recovery points that you do to want DPA to use for RPO
Replication Status Diagram calculations like replicas that are used for reporting or test
and Topology Reports systems.
You can define the dirty recovery points in the data protection
policy.

Symmetrix and CLARiiON Log Log level settings for Symmetrix and CLARiiON. Settings are
Level INFO and DEBUG.

Support Symmetrix Masking Option to render Out of Memory exceptions obsolete and
Reports prevent memory issues when dealing with the Masking Config
and LUN Mapping reports data.

Support Application Discovery Option to enable administrator login as administrator when

Impersonation retrieving the application discovery data from the discovered
host.Enabled by default.

Days to keep Temporary Files Days the system maintains the temporary files. The files are
stored under installationDir\services\agent\tmp\output. The
default is 14. Older files are deleted at a nightly process.

Agentless Discovery
The Agentless Discovery settings are described in Table 21 on page 49.

Table 21 Agentless Discovery settings

Setting Description

Agent Response Timeout The time that DPA waits for response from the agent before
timeout.

Telnet/SSH Login Prompt The time that DPA waits for Telnet/SSH session to be created
Timeout before timeout.

Telnet/SSH Handshake The time that DPA waits for Telnet/SSH handshake before
Timeout timeout.

Delete files created on the Defines if temporary files will be deleted from the analyzed
client during agentless object at the end of the discovery.
discovery The default is that the files will be deleted.

Sudo Program Path The sudo program path for Agentless discovery settings. The
default path is /usr/local/bin/sudo. The sudo command can
also be located in either /sbin or /usr/sbin.

User and system settings 49

Administering DPA

Manage Data Collection Defaults

A DPA request contains data on how and when to gather data from an object. Data
collection defaults are the template used by the Discovery Wizard to assign requests to
objects. You can set the global default settings in Admin > System > Manage Data
Collection Defaults
All requests have a default data gathering frequency and a set of options associated with
them. You can edit global data collection default values to be picked up by the Discovery
Wizard for certain objects. The DPA online help provides information on editing requests.
You can gather certain types of data with DPA without deploying an agent on the
monitored device. To do this, an agent on another computer (such as the DPA Server)
gathers the data remotely. When gathering data remotely, the agent’s host is referred to as
a proxy server. The agent uses a protocol to gather data from the remote computer and
forwards it back to the DPA server. The protocol used depends on the type of data being
collected.
For certain devices types, such as IP switches and Fibre Channel switches, data must
always be gathered remotely as it is impossible to install an agent directly on a switch.
To configure remote data collection within DPA, configure the details when assigning
requests. If the Discovery Wizard created the objects, this configuration is already created.
However, if proxy or credential details have changed, modify the details as required.
Retention Periods on Requests are set on individual request using the Edit Request dialog
box. Table 18 on page 48 provides information on default retention periods for Data
Collection policies.

Customizing business information

This section provides information on the types of DPA service customization which can
only be done by an administrator. You must have physical access to the host on which DPA
is running.
The DPA Product Guide provides information on customizing viewlets, dashboards, and
reports. These customizations can be carried out by a user.

Using the vtl templates

When the Publisher process creates reports when publishing to HTML, it uses vtl(Velocity
Template Language) templates located in the <DPAHOME>/services/shared/vtltemplates
directory on the DPA Server to determine the report’s default layout and style. By default,
the DPA Server process uses the following template files: reportcard.vtl, chart.vtl, and
table.vtl however, you can use another template file. You can create template files to
change the appearance of reports that are published by the DPA Server process.
The template types are:
◆ Default uses the default VTL for the renderer.
◆ pivot is for generating pivot tables.
◆ pivot.css is for generating pivot tables using CSS.

50 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

◆ controlpanel.css is for generating control panels using CSS. Does not contain the date
and version at the bottom.
◆ pivot.controlpanel.css is for generating pivot tables in control panels using CSS.
◆ CSS is for generating Dashboards and control panels that maintain their layout when
published.
Table 22 on page 51 lists the VTL templates.

Table 22 VTL templates

VTL template Description Template type

chart.controlpanel.css.vtl Same as chart.vtl except this uses CSS. N/A

chart.css.vtl Same as chart.vtl except this uses CSS css

chart.vtl Used by chart renderers that produces an image for the HTML output Default
such as Area, Bubble, Column, Line, Pie, Topology, and 3D

controlpanel.css.vtl Using CSS the Dashboard reports are placed using absolute controlpanel.css
positioning so that it resembles the layout in the UI. Does not work in
some older mail clients

email.attach.vtl Used when sending the report as an attachment to the email N/A

email.image.embed.vtl Used for embedding the report inside of the email N/A

email.notification.vtl Used for creating the notification that can be sent out after a report N/A
was published

healthstatus.controlpanel.c Same as healthstatus.vtl except this uses CSS. Also does not contain N/A
ss.vtl the date and version at the bottom

healthstatus.css.vtl Same as healthstatus.vtl except this uses CSS css

healthstatus.vtl Used for Health Status Default

reportcard.controlpanel.css. Same as reportcard.vtl except this uses CSS. Also does not contain the N/A
vtl date and version at the bottom

reportcard.css.vtl Same as reportcard.vtl except this uses CSS css

reportcard.vtl Used for ReportCard Default

table.controlpanel.css.vtl Same as table.vtl except this uses CSS. Also does not contain the date N/A
and version at the bottom

table.css.vtl Same as table.vtl except this uses CSS css

table.pivot.controlpanel.css Same as table.pivot.vtl except this uses CSS. Also does not contain pivot.controlpanel.css
.vtl the date and version at the bottom

table.pivot.css.vtl Same as table.pivot.vtl except this uses CSS pivot.css

table.pivot.vtl Used for Pivot Table pivot

table.vtl Used for Table Default

timeline.controlpanel.css.vt Same as timeline.vtl except this uses CSS. Also does not contain the N/A
l date and version at the bottom

timeline.css.vtl Same as timeline.vtl except this uses CSS css

timeline.vtl Used for timeline charts. HTML gets embedded in the VTL Default

Customizing business information 51

Administering DPA

Example - Part 1: Add a message and your company details to the table VTL template
If you are required to send daily or weekly reports in HTML format to your customers, and
you accomplish this with scheduled reports, then you can add custom text (such as a
message or company contact information) to the scheduled report by creating a custom
VTL template. The custom text displays for all HTML reports using this template.
1. In the styles or vtltemplates directory on the DPA Server, copy the table template
(table.vtl) and rename it. For example, if you are creating a VTL template for table
reports for the company EMC, use the naming standard of table.<companyName>.vtl
then rename the table template to table.emc.vtl
2. Open the VTL in your text editor.
3. Using HTML tags, add text similar to the following within the body.
<body bgcolor="$background"><font face="Arial, Verdana, Helvetica, Sans-serif"
color="$foreground">
<body>
Dear customer,
<p>
Your daily system status report is below.
<p>
Thank you,<br>
EMC Corporation
<p>
US Phone:1-800-555-5555<br>
Email:[email protected]<br>
Website: www.EMC.com
<p>
<table>
...
</table>
</body>

4. Save the VTL.

Example - Part 2: Use your custom VTL template in a scheduled report

Now that you have your custom VTL template, select this VTL in the Scheduled Report
Wizard.
1. In the DPA web console, create a new or update an existing scheduled report.
2. In Publish Settings, select the Web Page (.html) report format and complete the
remaining fields.
3. In Advanced, select your EMC template and then click OK. The template named Default
is the unedited table.vtl.
4. Click the test icon to send the scheduled report to the Publisher. If you publish to file,
proceed to the default directory to view the report and then make any necessary
updates to the VTL template. The default directory of the report is
<install-dir>\services\shared\report-results\scheduled.
5. If no further updates need to be made to the VTL template, save and close the
Scheduled Report Editor.

52 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

Importing and exporting custom report templates and custom control panels
You can import and export custom report templates and custom control panels from DPA
5.5.1 and later into DPA from a WDS file through the Custom Templates section. Importing
and exporting to XML is not supported. You cannot import or export system templates. The
imported reports must be supported on DPA 6.1.
You can import and export custom report templates and custom control panels to fulfill
the following needs:
◆ Import custom reports from DPA 5.x.
◆ Import custom reports that were created by another DPA system.
◆ Export custom reports as an alternate backup.
The EMC Data Protection Advisor online help system provides more information on how to
import and export custom report templates.

Datastore Replication administration

Note the following limitations for Datastore Replication:
◆ In busy environments, best practice is to stop the Application servers for a Datastore
Replication export so that the export can complete and be imported to the Slave
Datastore, and resync with the Master Datastore.
◆ DPA supports Datastore Replication exports from the Master Datastore only. DPA does
not support Datastore Replication exports run from the Slave Datastore.

Configuring Datastore Replication after deployment

Note that the CLI commands in this section are formatted for Linux RHEL.
To configure Datastore replication on a system that is already installed and operational:
1. Confirm that the Datastore server is installed as a Slave. If it is not, configure the
Datastore server as a Slave Datastore:
• Install using Advanced Options and select SLAVE when configuring Replication.
• Alternatively, if the Datastore server is already installed as a Master, then run the
dpa.sh ds rep --role SLAVE <Ip of master> command to make the Datastore server a
Slave.
2. Follow the procedure “Integrating Slave Datastore after it has been offline” on
page 54.

Carrying out Datastore server failover

When the Master Datastore fails, perform a failover to the Slave Datastore.
1. On the Slave Datastore, type:
dpa.sh ds rep --failover

2. Stop the Application server. Type:

dpa.sh app stop

Datastore Replication administration 53

Administering DPA

3. Reconfigure the Application server to point to the new Master Datastore. Type:
dpa.sh app con -m <hostname/IP of new MASTER>

4. Start the Application server. Type:

dpa.sh app start

5. Verify that the Datastore server is running. Type:

dpa.sh ds status

Output is INSTALLED, STOPPED, or RUNNING.

6. If it is not running, start it. Type:
<DPA_HOME>/emc/dpa/services/bin/dpa.sh ds start

Integrating Slave Datastore after it has been offline

This procedure is applicable if Datastore Replication was previously configured and the
Slave Datastore goes down. This procedure is also applicable if you are introducing
Datastore Replication into an already operational deployment. You then reintegrate a
Slave Datastore.
The Datastore is configured to allow approximately 6 hours on downtime before it needs
reinitialization. However, this value is approximate and a heavily loaded server may
require reinitialization if down for less time. We recommend that you carry out testing to
determine the threshold for your deployment.
This procedure is also applicable to resynchronizing a standalone Slave Datastore after
isolation. Examples of isolation could be a network outage or break down in
communications between the Master and Slave Datastores.
1. Create an empty directory on the Master Datastore to which to export the Master
Datastore file set. For example, /tmp/export
2. Export the Master Datastore file set from the running Master Datastore. Type:
dpa.sh ds rep --export /tmp/export

3. Create an empty directory on the Slave Datastore into which to copy the Master
Datastore file set.
4. Use the appropriate platform command to copy the files to the empty directory on the
Slave Datastore.
5. Import the Slave Datastore. Type:
<DPA_HOME>/emc/dpa/services/bin/dpa.sh ds rep --import /tmp/import

where <DPA_Home> is the location of the DPA installation.

6. Start the Slave Datastore server. Type:
<DPA_HOME>/emc/dpa/services/bin/dpa.sh ds start

where <DPA_Home> is the location of the DPA installation. The status of the Slave
Datastore at this point is STARTED.

7. Verify that replication is functioning. On the Master Datastore, type:

54 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

bin/dpa.sh ds rep

Output such as the following on the Slave Datastore appears:

EMC Data Protection Advisor

[INFO] Replication State : SLAVE (for 10.11.111.112)

Command completed successfully.

Output such as the following on the Master Datastore appears:

EMC Data Protection Advisor

[INFO] Replication State : MASTER

[INFO] Defined Slaves
: 10.11.111.111/12

[INFO] Replication Status : ON

Command completed successfully.

Reconfiguring Datastore servers

Use this procedure if you failed over to your Slave Datastore and want to reconfigure the
former Master Datastore as a Slave Datastore.
1. On the new Master Datastore, use the addSlave command with the IP of the new
Master Datastore. Type:
dpa.sh ds rep --addSlave <ip_of_master>

2. Restart the new Master Datastore. Type:

dpa.sh ds restart

3. Export the new Master Datastore. Type:

dpa.sh ds rep --export /export

4. Configure the new Slave Datastore as SLAVE. Type:

dpa.sh ds rep --role SLAVE <ip of MASTER>

5. Stop the Slave Datastore. Type:

dpa.sh ds stop

6. Import the Master Datastore to the Slave Datastore. Type:

dpa.sh ds rep --import /import

7. Start the Slave Datastore server. Type:

dpa.sh ds start

Stopping Datastore Replication

To stop Datastore Replication, stop the Slave Datastore. On the Slave Datastore, type
dpa.sh ds stop.

Datastore Replication administration 55

Administering DPA

Clustering administration
Adding an Application object to a cluster after DPA deployment
You can modify a DPA Application object that was installed as a stand-alone object (the
default) to be part of a cluster after DPA is deployed and operational using the DPA CLI.
1. On the Application server, type dpa application stop to stop the Application
service. The Application service must be stopped before adding to a cluster.
2. On the Application server, type dpa application promote --bind
<ip_address> to promote the Application to a running cluster, where <ip_address>
is the IP address of the Application server. If this server has multiple IP addresses, you
should specify the IP address that you want to use for communication with the other
clustered Application servers.
If running on Windows, type:
dpa application promote --bind <ip_address> --user (DOMAIN\username) --path <path>

where:
• <ip_address> is the IP address of the Application server. If this server has multiple
IP addresses, you should specify the IP address that you want to use for
communication with the other clustered Application servers.
• (DOMAIN\username) is the user account that has read and write access to the
shared folder. If omitted the local system user is used. This user account must
have the Log on as a service Windows permissions enabled.
• <path> is the path that is shared among the clusters.
3. Type the following command on the Datastore server to set the number of connections
per Application server:
dpa ds tun -c <value> <memory>MB|GB

where <value> is 150 per Application server, and <memory> is the amount memory in
MB or GB that is available on the Application server.
4. If you are not running Windows, proceed to step 5 on page 57 . If running on Windows,
uninstall and install the Application service to modify the Windows service to run the
DPA Application node using the domain user. Run the following commands:
dpa app uninstall
dpa app install --user (DOMAIN\username) --password (password)

where:
• (DOMAIN\username) is the user account with which to run the Application service.
The Log on as a service Windows permissions must also be enabled.
• <password> is the password for the user specified.

56 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

5. On the Datastore server, type dpa datastore configure --add

<ip_address> to add the Application object to the Datastore server, where the
<ip_address> is the address of the Application server. If the Application server has
multiple IP addresses you should specify the IP address used to communicate with
the Datastore server.
6. On the Application server, type dpa application configure to verify that you
are clustered.
7. On the Application server, type dpa application start to start the Application
service and restore the Application server functionality.

Removing an Application object from a cluster

You remove an Application object from a cluster using the DPA CLI. This makes the
Application object a stand-alone object.
1. On the Application server, type dpa application stop to stop the application
service. The Application service must be stopped before removing from a cluster.
2. On the Application server, type dpa application demote to demote the
Application from a running cluster.
3. On the Application server, type dpa application configure to verify that the
Application is removed from the cluster; it will show as type STANDALONE.
4. On the Application server, type dpa application start to start the Application
service and restore the Application Server functionality.
“DPA Server CLI commands” on page 59 provides more information on DPA Clustering CLI
commands.

Windows considerations
◆ If the password for the Domain user is changed, you must uninstall and reinstall the
DPA Application node. Run the following commands:
dpa app uninstall
dpa app install --user (DOMAIN\username) --password (password)

where:
• (DOMAIN\username) is the user account with which to run the Application service.
The Log on as a service Windows permissions must also be enabled.
• <password> is the password for the user specified.

Backup of the Datastore server

It is best practice to back up the DPA Datastore server regularly and particularly prior to
making any major change to DPA such as upgrading to a newer version or migrating to new
hardware. An export of the Datastore server contents is part of an overall backup of the
DPA instance.
Exporting and importing a DPA Datastore server is supported only on the same version of
the DPA Datastore server.

Backup of the Datastore server 57

Administering DPA

Exporting the DPA Datastore server to a file

With this export command, a complete and consistent copy of the Datastore server is
exported to the local file system, in a location that can optionally be specified.
The default file name of the export is: datastore-<date-and-time>.
Type the following command from a command line prompt.
dpa datastore export [options]

The exported Datastore file is saved to the same directory where the export command was
run.
To save the exported Datastore file to a specific directory, specify the location at the end of
the command line. For example, the following command line exports the file to C:\
because that is the location specified:
C:\Program Files\EMC\DPA\services\bin>dpa datastore export C:\

Exporting the DPA Datastore server to Pipe

With this export format, a complete and consistent copy of the Datastore server is
streamed to a named pipe from where a Backup Manager can read the contents.
Type the following command from a command line prompt.
dpa datastore export --pipeline

For example dpa datastore export --pipeline /mydir/mypipe

DPA supports backup up to Avamar using ds export and pipe it directly to Avamar. For
more information, see the Avamar documentation on how to pipe a backup into Avamar
using "named pipes."

Importing the DPA Datastore server

The DPA Datastore import command line option is used to import the contents of a
Datastore file to the DPA Datastore server.
1. Stop the DPA Application service.
2. Import the Datastore server.
3. Start the DPA Application service.
From a command line prompt, type the following:
dpa app stop
dpa datastore import [options] <file name>
dpa app start

Where <file name> is the previously exported Datastore file.

The import command will replace the existing Datastore server contents with the contents
contained in the Datastore export file.
For a complete list of DPA commands, type dpa --help from a command line prompt. “DPA
command line operations” on page 59 provides more information.

58 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

DPA command line operations

Sourcing the DPA config file for UNIX users
UNIX users must source the DPA config file before running any of the command line
operations.
To source the config file, navigate to the etc folder of the DPA installation directory and run
the command as follows:
cd <DPA install dir>/agent/etc

. ./dpa.config

DPA Server CLI commands

In a default DPA installation, the dpa CLI command can be found in
<install_dir>/services/bin on UNIX and Linux and in <install_dir>\services\bin on Windows.
Use the following syntax:
Windows:
dpa <service_part> <command> [options]

UNIX/Linux:
dpa.sh <service_part> <command> [options]

Where <service_part> is Application, Datastore, or service. The service component

includes both the Application and Datastore services.
dpa application <command> [options]
dpa datastore <command> [options]
dpa service <command> [options]

The dpa server start/stop/restart command apply to whichever services are installed on
the current host only. For example, running dpa server stop on the DPA Datastore does not
stop services running on the DPA Application server.

Command and option abbreviations

The dpa command supports the abbreviations shown in Table 23 on page 59. Refer to the
specific dpa command for available options for that command.

Table 23 Command and option abbreviations (page 1 of 2)

Command and option Abbreviation

--add -a

--bind -b

--cluster -c

--delete -d

--help -h

--master -m

DPA command line operations 59

Administering DPA

Table 23 Command and option abbreviations (page 2 of 2)

Command and option Abbreviation

ping -p

--pipeline -p

--platform -p

tune -t

dpa application dpa app

dpa datastore dpa ds

dpa service dpa svc

dpa application commands

Use the dpa application commands to manage the DPA Application service.
dpa application [options]
dpa application configure [options]
dpa application demote [options]
dpa application install [options]
dpa application ping [options]
dpa application promote [options] [<Application Server_IP_Address>]
dpa application restart [options]
dpa application start [options]
dpa application status [options]
dpa application stop [options]
dpa application tune <value>MB|GB [options]
dpa application uninstall [options]
dpa application version [options]

After you start, stop, or restart a service, it might take a number of minutes to complete
and might not result in an immediate state change.

dpa application configure

Configures the Application service, including specifying the Datastore serverserver and
cluster to communicate with.
dpa application configure [options]
dpa app con [options]

Command options
--master <IP_address> — Identifies the Datastore serverserver with which to communicate.
The Application service must be stopped for this command to operate.
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages
--bind (-b) — Specifies the IP address to announce the Application service on

Examples
dpa app -b 192.168.1.10

60 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

# dpa application ping

EMC Data Protection Advisor
[INFO] Datastore Service : 127.0.0.1
[INFO] Operation Mode : STANDALONE

dpa application demote

Demotes the Application service from a cluster environment. The Application service will
operate as a stand-alone object instance. The Application service must be installed and
stopped for this command to operate.
dpa application demote [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

Examples
dpa application demote
dpa app demote

dpa application install

Installs the Application service. The Application service will operate as a system managed
service, manageable through normal operating system service commands. Management
of the lifecycle of the service can also be managed through this command line tool. This
command will install the service, but will not start it automatically. If the Application
service is already installed this command will fail.
dpa application install [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

--user (-u) (DOMAIN\username) — Applicable only to clusters on Windows; User account

with which to run the Application service. The Log on as a service Windows
permissions must also be enabled.

--password (-pass) <password> — Applicable only to clusters on Windows; Password for

the user specified. This user account must have the Log on as a service Windows
permissions enabled

Examples

dpa app install -u DOMAIN\pault -pass <mypassword>

DPA command line operations 61

Administering DPA

dpa application ping

Tests the connection between the Application object from which it is sent and the defined
Master Datastore service.
dpa application ping [options]
dpa app pin [options]

Command Options
--help (-h) — Displays help screen
--version — Displays tool version information
--quiet — Displays warnings and errors only

dpa application promote

Promotes the Application service to a cluster environment. The Application service will
operate as a object within a cluster of objects. Management of the lifecycle of the service
can also be managed through this command line tool. The Application service must be
installed and stopped for this command to operate.
Note that for Windows you must to modify the windows service to run the DPA Application
node using the domain user, you will need to uninstall and install the Application service.
“dpa application install” on page 61 and “dpa application uninstall” on page 64 provide
information.
dpa application promote [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages
--bind (-b) — Specifies the IP address to announce the Application service on
--user (-u) (DOMAIN\username) — Windows only; points to the user account that has read
and write access to the shared folder. If omitted the local system user is used.
--path (-p) <path> — Windows only; path that is shared among the clusters.

Examples
dpa application promote -b 192.168.1.10
dpa app promote -b 192.168.1.10 -u DOMAIN\pault -p \\mymachine.emc.com\shared

dpa application restart

Restarts the Application service. This command first stops the Application service and
then starts the service. The Application service must be running for this command to
operate.
dpa application restart [options]

62 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa application start

Starts the Application service. The Application service must be installed and stopped for
this command to operate.
dpa application start [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

Delays when starting and stopping DPA services

You might experience delays in starting the web console when starting the DPA services. If
the DPA services have just been installed, there is a delay of up to 10 minutes in starting
the web console. Similarly, if the DPA services are restarted, there might be a delay of
about 3 minutes in starting the web console.

Note: The DPA services must be running if you want to start the DPA web console.

dpa application status

Displays the status of Application service. For example, RUNNING (STARTING...), RUNNING,
STOPPED
dpa application status [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

Examples
# dpa application status
EMC Data Protection Advisor
The status of the Application Service is RUNNING

dpa application stop

Stops the Application service. The Application service must be installed and running for
this command to operate.
dpa application stop [options]

DPA command line operations 63

Administering DPA

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa application tune

Configures tunable parameters of the Application service for the available host memory
resources.
dpa application --tune <size> MB|GB
dpa app tune <size> MB|GB

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

Note: Restart the Datastore server for the new parameters to be applied. Subsequently,
restart the connecting Application service.

dpa application uninstall

Uninstalls the Application service.
dpa application uninstall [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa application version

Displays the version information for the various functional libraries that make up the
Application service. The functional libraries include Apollo, Controller, DPA, RemoteX, and
UI.
dpa application version [options]

Command options
--platform — Include platform version info
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

64 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

Examples
# dpa application version
[INFO] Version for Apollo EAR is 1.0.0.3304
[INFO] Version for Controller RAR is 6.0.0.69338
[INFO] Version for DPA EAR is 6.0.0.69338
[INFO] Version for Remotex EAR is 1.0.0.3304
[INFO] Version for UI WAR is 6.0.0.local

dpa datastore commands

Use the dpa datastore commands to manage the DPA Datastore service.
dpa datastore [options]
dpa datastore configure [options]
dpa datastore export [options]
dpa datastore import [options] <import_file name>
dpa datastore install [options]
dpa datastore reindex [options]
dpa datastore replicate [options]
dpa datastore restart [options]
dpa datastore start [options]
dpa datastore status [options]
dpa datastore stop [options]
dpa datastore tune <size>MB|GB [options]
dpa datastore uninstall [options]

After you start, stop, or restart a service, it might take a number of minutes to complete
and might not result in an immediate state change.

dpa datastore configure

Configures the Datastore service, including adding or removing an Application service to
the list of allowed connections to the Datastore service.
dpa datastore configure [options]
dpa ds configure [options]

Command options
--bind <IP_address> — Set the bind address for the Datastore service. The default is
127.0.0.1



--bind cannot be specified with --add or --delete.

--add <IP_address> — Add an Application service node as a valid Datastore client

--delete <IP_address> — Remove an Application service node as a valid Datastore client
ping — Pings the Datastore service that has been defined. The Operation Mode in the
output identifies whether the Application is within a cluster or stand-alone.
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

DPA command line operations 65

Administering DPA

Examples
dpa datastore --add 192.168.1.10
dpa ds con --bind 192.168.1.1

dpa datastore export

Exports the contents of the Datastore to the file name or pipeline specified.
◆ When exporting to a file, a complete and consistent copy of the Datastore server will
be exported to the local file system in a location that can be optionally specified. The
default file name of the export is: datastore-<date_and_time>.
◆ The exported Datastore file is saved to the given directory. If not directory was
specified on the command line the export file will be created where the export
command was run.
◆ When exporting to pipe, a complete and consistent copy of the Datastore server will
be streamed to a named pipe from where a Backup Manager can read the contents.
The Datastore service must be installed and running for this command to operate. Any
existing file name present will be overwritten.
dpa datastore export [options] <directory>

Command options
--pipeline — Export to pipe
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

Examples
C:\Program Files\EMC\DPA\services\bin>dpa datastore export C:\

dpa datastore import

Imports the contents of the Datastore export file to the Datastore. The import files must be
available on the local file system. The Datastore service must be running for the import
command to execute. Any existing file name present will be overwritten.
dpa datastore import [options] <file name>

Where <file name> is a previously exported Datastore file. The import command replaces
the existing Datastore server contents with the contents contained in the Datastore export
file.

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages
<import_file name> — File name of the exported file to import

66 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

Examples
# dpa datastore import datastore-2013-02-20-1205
EMC Data Protection Advisor
Datatstore imported from file : datastore-2013-02-20-1205
Imported to the datastore successfully

dpa datastore install

Installs the Datastore service. The Datastore service will operate as a system managed
service, manageable through normal operating system service commands. Management
of the lifecycle of the service can also be managed through this command line tool. This
command will install the service, but will not start it automatically. If the Datastore service
is already installed this command will fail.
dpa datastore install [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa datastore reindex

Reindexes the Datastore service to ensure optimal performance. By default, the Datastore
server automatically undertakes activities to ensure optimal performance. Reindex is a
manual option to begin these activities outside of the automatic cycle. For example, if a
Datastore server suffers from many writes and many deletes in a short time span, then run
dpa datastore reindex to ensure that you have optimal performance.
dpa datastore reindex [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa datastore replicate

Configures the Datastore service to replicate to another instance.
dpa datastore replicate [options]
dpa ds rep [options]

Command options
--addSlave (-a) <hostname/IP of SLAVE> — Adds a Slave Datastore to a Master Datastore
--deleteSlave (-d) <hostname/IP of SLAVE> — Deletes a Slave Datastore from a Master
Datastore
--role (-r) MASTER — Redefines the role of Slave Datastore to Master Datastore
--role (-r) SLAVE <IP of MASTER> — Redefines the role of Master Datastore to Slave
Datastore

DPA command line operations 67

Administering DPA

--failover — Initiates failover between Slave Datastore and Master Datastore

--import (-i) <import> — Initializes a SLAVE Datastore with replica located in specified
file system
--export (-e) <export> — Produces a clone of the MASTER Datastore to specified file system
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa datastore restart

Restarts the Datastore service. This command first stops the Datastore service and then
starts the service. The Datastore service must be running for this command to operate.
dpa datastore restart [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa datastore start

Starts the Datastore service. The Datastore service must be installed and stopped for this
command to operate.
dpa datastore start [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa datastore status

Displays the status of Datastore service. For example, RUNNING (STARTING...), RUNNING,
STOPPED
dpa datastore status [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

Examples
# dpa datastore status
EMC Data Protection Advisor

68 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

The status of the Datastore Service is RUNNING

dpa datastore stop

Stops the Datastore service. The Datastore service must be installed and running for this
command to operate.
dpa datastore stop [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa datastore tune

Configures tunable parameters of the Datastore service for the available host memory
resources.
dpa datastore tune <size>MB|GB [options]
dpa ds tune <size>MB|GB [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

Note: Restart the Datastore for the new parameters to be applied. Subsequently, restart
the connecting Application service.

dpa datastore uninstall

Uninstalls the Datastore service.
dpa datastore uninstall [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa service commands

Use the dpa service commands to manage the DPA Application and the DPA Datastore
services.
dpa service restart [options]
dpa service start [options]
dpa service status [options]
dpa service stop [options]
dpa service uninstall [options]

DPA command line operations 69

Administering DPA

dpa service restart

Restarts whichever services are installed on this host. For example, if the DPA Application
service has been installed on this host,then this command restarts the instance of the
Application server.
dpa service restart [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa service start

Starts the DPA service installed on this host.
dpa service start [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa service status

Displays the status of DPA services installed on this host. For example, RUNNING
(STARTING...), RUNNING, STOPPED
dpa service status [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

Examples
# dpa service status
EMC Data Protection Advisor
The status of the Datastore Service is RUNNING
The status of the Application Service is RUNNING (STARTING ...)

dpa service stop

Stops the DPA services installed on this host. The services must be installed and running
for this command to operate.
dpa service stop [options]

70 DPA 6.1 SP2 Installation and Administration Guide

 Administering DPA

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

dpa service uninstall

Remove from GA doc since cohosted deployment not supported? or leave with note?
Uninstalls the application service and then the Datastore service.
dpa service uninstall [options]

Command options
--help — Displays the help screen
--version — Displays the tool version information
--quiet — Suppresses all output except for warning and error messages

Loading historical backup job data

After a backup application object is created and requests are assigned, the agent
immediately begins gathering data on backup jobs to store in the Datastore. However, the
agent also can gather data on backup jobs that were run prior to object creation in DPA.

IMPORTANT
To commit the data to the DPA server, the installed agent must have previously been
started and successfully registered with the DPA Server. However, it need not be currently
running in order to load the historical data.

Do not run this command twice within the same date span because doing so introduces
duplicate data into the database.
Each backup module has an equivalent executable in the installed Agent's bin directory,
<DPA_HOME>/agent/bin directory, where <DPA_Home> is the location of the DPA
installation.
The following example collects backup job data run on a NetWorker server:
Example <install_dir>/agent/bin/dpaagent_modnetworker -c -f jobmonitor -t
NetWorkerServer_IP -B "01/01/2012 00:00:00" –E "01/01/2012 00:00:00"

Running the executable with the -? parameter shows the valid command line options.
Module options applicable to the request (eg. timeformat) might also need to be specified
explicitly on the command line in order to ensure consistent behaviour with "normal" data
collection. Specifically, in the case of the DataProtector jobmonitor request, the
occupancy option must be specified explicitly if you want historic data to be included in
occupancy calculations. The DPA Data Collection Reference Guide provides more
information on options. The “Job Monitor” section provides more information on the
occupancy option.

Loading historical backup job data 71

Administering DPA

To load historical backup data, run the agent binary from the command line with the
following parameters: You should specifically use:
◆ -f <function name> — Name of data gathering function to execute. Always jobmonitor.
Mandatory.
◆ -t <target host> — Host address of backup Application server. The default is localhost.
◆ -B <start time> — Start time from which to gather backup jobs. The format is
dd/mm/yyyy hh:mm:dd.
◆ -E <end time> — End time from which to gather backup jobs. The format is dd/mm/yyyy
hh:mm:dd.
The start and end times can also be in UNIX epoch time format.
If <start time> is specified and <end time> is not, <end time> is set to the current time.
This includes all the backup jobs that ended after <start time>.
If <end time> is specified and <start time> is not, <start time> is set to 0. This includes all
the backup jobs that end before <end time>.
◆ -i — TSM instance name (TSM only).
◆ -l <log file name> <log file path and name>- Name of the log file only or name and path
of log file to generate when running the command to load historical data.
The default log file location is the location from which the command is run.
◆ -U — Username to connect to the backup application (TSM and Avamar only).
◆ -P — Password to connect to the backup application (TSM and Avamar only).
◆ -c- Commit — Instructs the module to send the data to the DPA Server. Running the
command without the -c flag can be used for troubleshooting DPA data collection.
The following example collects backup job data run on an Avamar server:
Example:
dpaagent_modavamar.exe -f jobmonitor -t De-dup-muc.corp.emc.com -U viewuser -P
viewuser1 -c -B "01/01/2012 00:00:00" -l /tmp/mod_avamar.log

Job summary reports

The job summary reports provide overviews of the totals of backup and maintenance jobs
(such as all jobs, successful jobs, failed jobs) that have occurred on backup servers. The
summary reports rely on the most up-to-date data in the Datastore server to produce
accurate summary results.
While historical backup job data is loading using the agent command line options,
summary reports might display inaccurate totals. It is best to wait until all historical job
data is loaded before running summary reports for the loaded historical periods.

72 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

CHAPTER 3
Environment discovery in DPA

This chapter includes the following sections:

◆ Configuring the environment for discovery............................................................... 74
◆ Monitored objects and groups .............................................................................. 106
◆ Policies and rules.................................................................................................. 108

73
Environment discovery in DPA

Configuring the environment for discovery

Discovery overview
The diagram below shows the relationship between the DPA Application object and the
DPA Agents deployed to monitor your data protection infrastructure. Some types of
devices need to be monitored by using a DPA Agent deployed as a proxy. A proxy is used
typically where the object being monitored is hardware and access for agent installation is
not possible. Most types of backup managers can be monitored by an agent directly
installed on the same host as the backup manager, or remotely by using proxy agent if the
backup manager is resource constrained.

Figure 2 Relationship between DPA Application nodes and DPA Agents monitoring applications

Defining objects to be monitored

To define objects to be monitored in DPA, follow the steps in Table 24 on page 75.

74 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Table 24 Data monitoring setup summary

Step Description

Check licenses Check that the licenses to monitor your device, host, or environment have
been purchased and installed. See step 1 on page 75 .

Install the agent If you are monitoring the object from a host other than the DPA server
host, you need to install a DPA Agent. See "DPA Agent installation"on
page 22.

Install third-party This step is required for remote or agentless (proxy) data collection.
binaries or define the You might need to install binaries on the DPA host or the remote agent
object for monitoring host to connect to the monitored object. You also might need to define an
account or connection on the monitored object.
The following sections describes the prerequisite configuration for all
objects:
• "Configuring application hosts for monitoring"on page 78
• "Configuring storage arrays for replication analysis"on page 78
• "Monitoring data protection servers"on page 81
• "Monitoring Databases"on page 92
• "Monitoring EMC RecoverPoint"on page 96
• "Gathering data from operating systems"on page 97
• "Monitoring tape libraries"on page 100
• "Switches and I/O Devices"on page 101
• "File Servers"on page 103
• "Backup appliances"on page 103
• "Monitoring Oracle ACSLS"on page 104
• "Disk management servers"on page 104
• "Monitoring VMware environment"on page 104

Create or modify the A credential stores the information used to connect to the monitored
DPA credential object. You might need to modify the default credential or create a new
one with the account details from the previous step.

Run the Discovery Use the Discovery Wizard to define objects to be monitored. Select
Wizard Inventory > Object Library > Run Discovery Wizard.

Modify the data Because database maintenance plans are not migrated, these must be
collection default reviewed to set the retention times of the data per request.
settings Data collection requests are assigned to the object created by the
Discovery Wizard. If you want to modify the default data collection, select
Admin > Systems > Manage Data Collection Defaults.

Test data collection After at least 10 minutes of letting the request run, run a report from the
object that should include data (for example, Backup Job Summary or a
configuration report).

Before you run the Discovery Wizard

1. Check the installed licenses. In the DPA web console, go to Admin > System > Manage
Licenses.
The options that are available for configuration in the Discovery Wizard depend on the
types of licenses that you have installed with DPA. If you do not have the correct
license installed, the option to create that device or host is disabled in the wizard.

Configuring the environment for discovery 75

Environment discovery in DPA

2. If you are performing discovery on a Linux host, ensure that the libstdc++.so.6 library
is installed on the host.
3. Ensure that you take note of the connectivity details outlined in Table 25 on page 76.
It is not necessary to run the Discovery Wizard for objects that have been migrated or if
you have migrated DPA 5.x connection settings.

Table 25 Connectivity details for configuring data collection through the Discovery Wizard (page 1
of 3)

Item Value

Network configuration information for DPA server or agent if agent is remote to DPA server

Hostname

IP Address

Network mask

Primary DNS server address

Secondary DNS server address

Gateway Address

Time zone

Credential information needed for discovery of virtual disks through SSH

IP Address of ESX Server

ESX Server Root Credential

Information needed for ldap user authentication— optional

LDAP Server Name/IP

Use SSL?

LDAP Server Port

LDAP Version

Distinguished Name of Base Directory

Identification Attribute

Credential information needed for discovery of servers and arrays

Server Name/IP

SSH Credentials

RPC Credentials

WMI Credentials

Solutions Enabler Host Credentials

Requires root/administrator credentials

RPA Credentials

Credential information needed for monitoring of oracle databases

Oracle username and password required

76 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Table 25 Connectivity details for configuring data collection through the Discovery Wizard (page 2
of 3)

Item Value

Oracle Service Name and Port,

specifically the Oracle SID and TNS port

Oracle Monitor RMAN

An oracle user with catalog access to the
RMAN schema and the username and
password is required

Oracle Host Name

Oracle Monitor Schema

If multiple RMAN schemas are present on
one Oracle SID, then each RMAN schema
owner and username and password are
required.

Credential information needed for SQL Server databases

SQL Database User Account

SQL Server Instance

SQL Database Name

PostgreSQL Credentials

PostgreSQL User Account (must be a

super user)

Credential information for backup servers, tape libraries, I/O devices

CommVault User Account

EMC Avamar User Account

Note: This is not required unless you have

changed the username and password
defined with Avamar for DPA's use.

HP Data Protector User Account

IBM TSM host, TSM Instance Name, TSM

port and TSM username and password for
each TSM instance is required

Symantec Backup Exec User Account

Symantec PureDisk User Account

SNMP community string for EMC Data

Domain
SSH username and password for Data
Domain, preferably a separate username
and password than the Data Domain’s
system administrator default credentials.
Both are required because data is
collected using both of the mechanisms

SNMP Community String for EMC Disk

Library

Configuring the environment for discovery 77

Environment discovery in DPA

Table 25 Connectivity details for configuring data collection through the Discovery Wizard (page 3
of 3)

Item Value

SNMP String for Fibre Channel Switch

SNMP Community String for Tape

Libraries

SNMP Community String for IP Switch

4. Ensure that communication between the monitored host and the recoverability
process is enabled:
• For monitoring Windows servers remotely, RPC services must be enabled and
accessible to the recoverability agent.
• For UNIX / Linux remote application monitoring, SSHD must be enabled and
accessible to the recoverability agent.
• For UNIX / Linux remote application monitoring, FTP/Telnet services must be
enabled and accessible to the recoverability agent.

Configuring application hosts for monitoring

Use the Discovery Wizard to define application hosts for monitoring in DPA. DPA can
monitor database hosts and Microsoft Exchange Server for replication analysis.

Configuring storage arrays for replication analysis

DPA monitors EMC VNX/CLARiiON, Symmetrix, and EMC VPLEX storage arrays. If these
storage arrays are replicated with EMC RecoverPoint, additional configuration is required
to enable complete replication analysis.

Discovering VNX/CLARiiON arrays

Port
DPA connects to the VNX/CLARiiON on TCP port 443. However, if the VNX/CLARiiON is
configured to use port 2163, use port 2163.

Discovering
EMC VNX/CLARiiON storage arrays must be monitored remotely from an agent that runs on
a different host, such as the DPA server. This is also known as the SE host or Connector.
The SE host can be used for discovery through a DPA Agent installed on it or through an
agentless mechanism that requires a privileged user's credentials.
DPA discovers all of the storage arrays that are being managed and creates objects in the
object library inventory.
You need to supply the name of the host on which EMC Solutions Enabler (SE) is installed.

78 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Setting up EMC Solutions Enabler

Discovering VNX/CLARiiON hosts from DPA requires SE to be installed. The DPA Software
Compatibility Matrix provides information on minimum versions required.
1. Download EMC Solutions Enabler minimum version required from EMC Online Support
at http://support.emc.com.
2. Install SE on the DPA server or any supported host that can connect to the
VNX/CLARiiON array through HTTPS.

Note: ASE license is not required to discover VNX/CLARiiON.

3. Create a text file with the following CLARiiON information by specifying one line per
VNX/CLARiiON:
<SPA IP> <SPB IP> <Username> <Password>

where
• <SPA IP> is the IP address of the first controller (SP-A).
• <SPB IP> is the IP address of the second controller (SP-B).
• <username> and <password> are the account name and password of a
VNX/CLARiiON user with view permissions.
The first field must be the first VNX/CLARiiON controller, followed by the second
controller.
Give the complete path, including the file name, to the file as a parameter in the
<filename> option. The best location is one that is near the SE installation, so that
you can use and monitor it for future needs.
4. If the Base license of SE exists, run the following command on the Solutions Enabler
host to register the VNX/CLARiiON:
symcfg disco -clar -file <file name>

5. If the Base license of SE does not exist, copy the ClarApiDiscovery executable file from
the following directory to the SE host:
Windows
C:\Program Files\EMC\DPA\services\agent\win-x86\policyimport-clar\

Linux/UNIX
/opt/emc/dpa/services/agent/<OS - e.g. linux>/policyimport-clar/

to the SE host.
6. Run the following command:
Windows
ClarApiDiscovery.exe REGISTER -file=<file name>

UNIX
./ClarApiDiscovery REGISTER -file=<file name>

Configuring the environment for discovery 79

Environment discovery in DPA

The available platforms are:

• AIX
• HP
• Linux
• Solaris
• Win32
7. Run the following command to confirm that the VNX/CLARiiON has been registered:
symcfg list -clar

8. If the VNX/CLARiiON appears in the list, run the Discovery Wizard in DPA and configure
the VNX/CLARiiON.

Discovering EMC Symmetrix arrays

Discovering
EMC Symmetrix storage arrays must be monitored remotely from an agent that runs on a
different host (such as the DPA server).
To configure multiple hosts and multiple storage arrays, use the Discovery Wizard. DPA
discovers all of the storage arrays that are being managed and creates objects in the
object library inventory.
You must supply the name of the host on which EMC Solutions Enabler is installed.
In order for Solution Enabler to see Device Groups which are stored locally by default on
Solution Enabler, you must open Global Name Services in Solution Enabler options file, as
follows:
1. Open option file under /./emc/API/symapi/config/
2. Find the line #SYMAPI_USE_GNS = ENABLE and unremark it so that it is:
SYMAPI_USE_GNS = ENABLE

3. Save the file.

4. Verify that the GNS service is running by running the stordaemon list command.
5. Run the symcfg disco command.

Discovering EMC VPLEX arrays

Port
DPA connects to the VPLEX on TCP port 443.

Discovering
EMC VPLEX storage arrays can be monitored from the DPA Server or remotely from any host
that has DPA Agent installed.
DPA discovers all of the storage arrays that are being managed and creates objects in the
object library inventory.

80 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Performing hostless discovery on Symmetrix and VNX/CLARiiON

Host discovery with replication monitoring requires either the installation of a local agent
on the host or the deployment of a remote agent with credentials for host access. Either
method might be prevented by customer security policies.
To use the agentless option, the Solutions Enabler host credentials should be provided.
The prerequisites for hostless discovery are the same as those described in "Discovering
EMC Symmetrix arrays"on page 80.

Configuring storage arrays that use EMC RecoverPoint to gather replication data
If your VNX/CLARiiON or Symmetrix storage arrays are replicated with EMC RecoverPoint,
DPA provides replication analysis for RecoverPoint replication operations.
To perform replication analysis for RecoverPoint, you have to configure the VNX/CLARiiON
or Symmetrix storage arrays and the RecoverPoint host in DPA in the correct order.
To gather RecoverPoint replication data:
1. Use the Discovery Wizard to create the host object for the Solutions Enabler host that
is connected to the storage array replicated with RecoverPoint.
2. Discover the arrays attached to the host.
3. Configure the Symmetrix or VNX/CLARiiON arrays by using the Discovery Wizard.
4. Import replication policy data from the storage arrays.
5. Configure the EMC RecoverPoint appliances’ data monitoring, as described in
"Monitoring EMC RecoverPoint"on page 96.
6. Ensure that the RecoverPoint Configuration request is assigned to the RecoverPoint
appliance object that handles replication for the storage array. Run this request.
7. After the RecoverPoint Configuration request is run and sufficient time has passed,
DPA should have begun gathering replication analysis data for RecoverPoint. Reports
can be run from the storage array objects and the Replication Analysis area will show
the mapping of storage and recovery points.

Monitoring data protection servers

This section describes how to monitor various data protection servers.

Monitoring backup applications configured as part of a cluster

You can monitor your backup applications that are configured as part of a cluster in a
couple of ways.
To monitor to a backup application in a cluster environment:
1. Install a remote Agent on a system outside of the cluster. Ensure that:
• the Agent can access the virtual server of the cluster using the required ports.
• the Agent has any required backup application binaries installed.
2. Discover the virtual server of the cluster by using the DPA Discovery Wizard.

Configuring the environment for discovery 81

Environment discovery in DPA

3. Collect data by using the remote Agent.

In this configuration if the server fails over, the cluster name always resolves and
provides the backup data.
To monitor a backup application in a cluster environment as well as monitor the local host
resources:
1. Install a local agent on each host in the cluster for host monitoring only.
2. Select one of the agents on the physical servers to monitor the virtual server.

Monitoring CA BrightStor ARCserve

CA BrightStor ARCserve servers are monitored from an agent running on the CA BrightStor
ARCserve server or from an agent running on any other Windows computer in the
environment.

Before starting the Discovery Wizard

Install the ARCserve Manager on the computer on which the agent is running. The agent
credentials must match the existing ARCserve account.
You will need to know the resolvable hostname or IP address of the ARCserve server. When
running ARCserve 11.x, the hostname must be the host short name (aliases cannot be
used).

Monitoring CommVault Simpana

Monitor CommVault Simpana servers from an agent running on the CommVault Simpana
database or from an agent running on any other Windows computer in the environment.

Before starting the Discovery Wizard

The DPA Agent service must run with a named account if Windows authentication is being
used by the CommVault SQL Server. The named account chosen for the DPA Agent service
must have permission for read access to the CommVault SQLServer Database.
Alternatively, if SQL authentication is used, you must define DPA credentials for the
CommVault requests; for example, username: cvadmin; password: password of cvadmin
user.
You will need to know:
◆ The resolvable hostname or IP address of the CommVault server.
◆ The database hostname and instance name if the CommVault database is remote to
the server.

Monitoring EMC Avamar

Monitor EMC Avamar® servers using a DPA agent installed on any remote computer in the
environment, including the DPA Server. Do not install a DPA agent on the EMC Avamar
server or storage object.

To enable the Clone Operations report to display data when the source grid is selected as
the scope for the report, you must monitor the source Avamar grid using the Job Monitor

82 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

request from an Avamar replication setup.

Before starting the Discovery Wizard

No additional software is required to monitor an EMC Avamar server remotely.
To gather data from EMC Avamar, DPA connects directly to the EMC Avamar database. It
connects to the mcdb database on the default port for EMC Avamar, which is 5555. If
these parameters were modified, edit the Avamar Configuration, Avamar Job Monitor and
Avamar Status request options to specify the database name and port in use. In the DPA
web console, go to Inventory > Object Library > [select object ] > Data Collection.
When DPA connects to the database, it uses the viewuser account to log in to the
database. If the EMC Avamar installation was modified so that this user does not have
permission to log in to the database, or the password for this user has been modified, edit
the Default Avamar Credentials in the DPA web console from Admin > System > Manage
Credentials to reflect the relevant username / password.

Note: The agent must be installed on a host that is in the same time zone as the Avamar
server.

Before you start the Discovery Wizard, you will need to know the resolvable hostname or IP
address of the Avamar server.

Monitoring EMC NetWorker

Monitor NetWorker either from an agent running on the backup server or remotely using an
agent running on the DPA Server or any other remote computer in the environment.

Before starting the Discovery Wizard

If monitoring NetWorker remotely, the NetWorker client package must be installed on the
agent’s host. The NetWorker module uses commands such as jobquery and nsradmin to
communicate with the NetWorker server and requires access to the binaries within the
NetWorker client package.
If monitoring NetWorker 7.6 or later remotely, the DPA user and the proxy host must be
added to the Users list of the NetWorker Administrators User Group. For example, if you
are monitoring NetWorker remotely from the host DPAAgentHost and the agent is running
as the Windows user DPAAgent, the following line must be added to the Users list of the
properties for Administrators:
user=DPAAgent,host=DPAAgentHost

Before you start the Discovery Wizard, you will need to know the resolvable hostname or IP
address of the NetWorker server.

Monitoring HP Data Protector

An agent can monitor HP Data Protector servers running on the HP Data Protector Cell
Manager or remotely from another computer.

Configuring the environment for discovery 83

Environment discovery in DPA

Before starting the Discovery Wizard

If monitoring a Cell Manager remotely, follow the same instructions as documented in
"Monitoring HP Data Protector remotely"on page 85.

Note: The status request cannot be assigned when monitoring the HP Data Protector
server remotely, as it relies on a command (omnisv) that is only available on the Data
Protector server.

If you are monitoring a Data Protector environment that uses the Manager of Managers
option, you must configure DPA as if monitoring a remote Data Protector server.
To monitor HP Data Protector remotely, the HP Data Protector client software must be
installed on the agent’s host and the client must be configured on the Data Protector Cell
Manager so that it has permission to run reports. "Monitoring HP Data Protector
remotely"on page 85 provides information on testing connectivity from the agent host.

Gathering occupancy data

Gathering occupancy data is not enabled by default for HP Data Protector. To enable
occupancy data gathering, the occupancy option for the DataProtector Jobmonitor request
must be enabled and the DataProtector Client Occupancy request must be assigned to the
Data Protector client in the Edit Request dialog.

Note: Gathering occupancy information for HP DataProtector can have a significant

performance impact on the Data Protector server.

You can use the DP_OCCUPANCY_DB_PATH environment variable for the DPA Agent to
control where the occupancy data is stored when you run the jobmonitor request. If you
do not use the DP_OCCUPANCY_DB_PATH environment variable, then the system stores
the occupancy data in the temporary directory.

Changing the location of Occupancy database on Linux

1. Stop the DPA Agent.
2. Use the cd command to access the /opt/emc/dpa/agent/etc directory.
3. Edit the dpa.custom file. Add the following to the end of the file:
COLLECTOR_DP_OCCUPANCY_DB_PATH=/your/absolute/path/
export COLLECTOR_DP_OCCUPANCY_DB_PATH

Ensure that you include the trailing backward slash (/) character in the above path.
4. Restart the DPA Agent.

Changing the location of Occupancy database on Windows

1. Stop the DPA Agent.
2. Run the regedit.exe as the administrator user.
3. Expand the HKEY_LOCAL_MACHINE registry key.
4. Expand the SOFTWARE registry key.
5. Create an EMC registry key if one does not already exist.

84 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

6. Create a DPA registry key if one does not already exist.

7. Ceate an Agent registry key if one does not already exist.
8. Create a new String registry value with name DP_OCCUPANCY_DB_PATH and set the
value to the desired directory path. For example:
C:\DPA\OccupancyData\

Ensure that you include the trailing backward slash (/) character in the above path.
9. Restart the DPA Agent.

omnirpt patch
HP has released a patch for Data Protector 6.1 that must be installed on a Data Protector
6.1 installation before it can be supported by DPA.
Table 26 on page 85 lists the required patch ID by platform.

Table 26 HP Data Protector 6.1 patch IDs

Platform Patch ID

Windows DPWIN_00417

HPUX PA-Risc PHSS_39512

HPUX IA64 PHSS_39513

Linux DPLNX_00077

Solaris DPSOL_00371

The patch is available for General Release from HP from www.hp.com. Type the patch ID
into the Search field of the HP home page. You are directed to the patch download page.

Monitoring HP Data Protector remotely

To install the client software on the computer that monitors the Cell Manager:
1. Start the Data Protector Manager administration GUI to add a client.
2. When selecting the software components to install on the client, ensure that the User
Interface option is selected.
The DPA Data Protector module requires access to commands such as omnirpt and
omnicellinfo to gather data from the Cell Manager. These components are only
installed when the user interface component is installed, so it is essential to select
this option.
3. Configure the client to have permissions to run reports on the Cell Manager. First
determine the user for which the Agent process will be running:
• On UNIX systems, the Agent always runs as the root user.
• On Windows systems, the Agent runs as the DPA Agent service user. To verify the
user for the service on a Windows system, start the Windows service control
manager and view the details of the DPA Agentservice.
4. Create a user on the Cell Manager that matches the Agent’s username. Type the name
of the host in the user definition field.

Configuring the environment for discovery 85

Environment discovery in DPA

5. Add the user to a Data Protector User Group that has Reporting and Notifications and
See Private Objects permissions.
Typically, this means adding the user to the admin group. However, to restrict a user
from inheriting other administrator privileges, create a new group with Reporting and
Notification and See Private Objects permissions and add the user to that group.
6. Verify that remote authentication privileges are set up correctly by running the
following command from the Agent’s host:
omnirpt -tab -report list_sessions -timeframe 06/01/01 12:00
06/01/30 12:00

If successful, this command returns a list of all the sessions that have run on the Data
Protector server during the time period specified. If an error indicating insufficient
permission to run reports appears, review the configuration settings on the Data
Protector server.

Monitoring IBM Tivoli Storage Manager (TSM)

Monitor a TSM server from an agent running on the TSM Server or remotely from an agent
running on a different host (such as the DPA server). If you are monitoring TSM remotely,
follow the instructions in "Monitoring TSM remotely"on page 87 before configuring the
server in DPA.

Before starting the Discovery Wizard

The TSM Credential must use the name and password of a TSM Administrator. The
Administrative user does not need full system privileges: Analyst or Operator privileges
are sufficient.
Select Admin > System > Manage Credentials to modify the TSM Credentials that are
created after you have used the Discovery Wizard to create a TSM object.
If the Server being monitored is a shared Library Client, the agent also must query the
Server’s Library Manager to gather certain data. By default, the agent uses the same
credentials used to query the Library Client to query the Library Manager.
If different credentials are required to access the Library Manager, they can be set using
the following DPA environment variables (UNIX) or registry settings (Windows):
◆ AGENT_TSM_LIBMGRUSERNAME
◆ AGENT_TSM_LIBMGRPASSWORD

Gresham Clareti EDT

In Tivoli Storage Manager environments that use Gresham Clareti EDT for device control,
DPA communicates with EDT to gather device configuration information by reading
information from two files:
◆ elm.conf
◆ rc.edt
DPA reads from elm.conf at the following location:
◆ On Windows, an environment variable called EDT_DIR is set by EDT. DPA looks up the
location specified in EDT_DIR.

86 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

◆ On UNIX, DPA looks first in /opt/GESedt-acsls/bin for elm.conf. If not found, on AIX
DPA looks in /usr/lpp/dtelm/bin. On other flavours of UNIX/Linux, DPA looks in
/opt/OMIdtelm/bin.
If the elm.conf file is not present in these directories, the registry variable (Windows) or
environment variable (UNIX) AGENT_TSM_ELMCONF_FILE NAME can be set to the location
of elm.conf if required.
DPA reads from the rc.edt file at the following location:
◆ On Windows, DPA looks up the location specified in the environment variable
EDT_DIR.
◆ On UNIX, DPA looks first in /opt/GESedt-acsls/SSI for rc.edt. If not found, on AIX DPA
looks in /usr/lpp/dtelm/bin. On other flavours of UNIX/Linux, DPA looks in
/opt/OMIdtelm/bin.
If the rc.edt file is not present in these directories, the registry variable (Windows) or
environment variable (UNIX) AGENT_TSM_RCEDT_FILE NAME can be set to the location of
rc.edt if required.

Note: Because a TSM environment using EDT requires the agent to read from these files to
collect configuration data, the agent must be on the same server as the TSM server.

Monitoring TSM remotely

When monitoring a TSM instance remotely, the TSM client software must be installed on
the host that will monitor the TSM instance. The TSM module uses the dsmadmc
command included with the TSM client software to connect to the TSM instance and
gather data.
In a default TSM Client installation on a Windows computer, the administrative
components required by DPA are not installed. To install the administrative components:
1. Click Custom when prompted during the TSM client installation.
2. Select Administrative Client Command Line Files and click Next. The TSM client
installation continues.
3. After the TSM client installation is complete, initialize the client for the first time by
starting the TSM Backup-Archive GUI from the Start menu. Use the wizard to configure
the client.
4. To configure the client, accept the default Help me configure the TSM Backup Archive
Client value and click Next. Either import an existing options file or create a new one
when prompted.
5. Accept the default value Create a new options file. A blank options file (dsm.opt) will
have to be created in the baclient directory under the install directory for TSM (default
C:\Program Files\Tivoli\TSM).
6. Continue to progress through the wizard. Complete all of the windows in the wizard
until a new options file is created.

Configuring the environment for discovery 87

Environment discovery in DPA

Monitoring Oracle RMAN

DPA no longer ships Oracle client (OCI) libraries with the DPA Agent. Therefore, in order for
the DPA Agent to collect data from an Oracle database (or Oracle RMAN), DPA requires the
following libraries for Oracle:
◆ libociei.so
◆ libclntsh.so linked to libclntsh.so.11.1
◆ libocci.so linked to libocci.so.11.1
The above libraries are included in the Oracle Instant Client, which can be downloaded
from the Oracle website. If you use the full Oracle Database Client, the libociei.so library is
not included. So you must and must manually copy it into AGENT_ORACLE_CLIENT_PATH
in order to work with the DPA Agent.
On Windows this is OCI.DLL and on UNIX, it is libclntsh.so.

Note: The library must be for the same platform as the DPA Agent. Example, if a 64- bit
Windows DPA agent is installed, then you must use the 64-bit Windows Oracle library.

You can download the Oracle Database Instant Client for free at
http://www.oracle.com/technetwork/database/features/instant-client/index.html
While installing the DPA Agent, you are prompted to specify if you want to utilize the Agent
to monitor Oracle and if so, provide the location of the Oracle client libraries. This action
sets a registry setting (on Windows) or modifies an environment variable in the dpa.config
file (on UNIX). If you change the location of the libraries after the install process is
completed, then you need to perform these steps manually.

Manually configuring DPA Agent to monitor Oracle database and Oracle RMAN
◆ To manually configure the DPA Agent to monitor Oracle RMAN:
On Windows, set the HKLM/Software/EMC/DPA/Agent registry of value type REG_SZ
as follows:
Value name: ORACLE_CLIENT_PATH
Value data: <directory containing the Oracle client libraries - oci.dll>

Note: The registry key is created if you have selected the Oracle database to be monitored
option while installing the DPA Agent. If the registry key is not created, you must create it
manually.

◆ On UNIX, modify the dpa.config file

The dpa.config file is available in <installdir>/agent/etc/dpa.config. Search for line,
AGENT_ORACLE_CLIENT_PATH=
and set the variable to the directory containing the Oracle client libraries -
libclntsh.so.
Restart the Agent service if you have changed the dpa.config file to include the Oracle
client path.

88 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Before starting the Discovery Wizard

To monitor Oracle RMAN remotely, the Oracle client must be installed on the agent's host
so that the agent process can communicate with the database.
The version of the client installed on the computer must be compatible with the target
database to which the agent connects.

Monitoring Data Domain Backup Enterprise Applications

DPA supports Data Domain Backup Enterprise Applications (DDBEA) for backing up
databases without the use of another backup application, such as backing up Oracle
RMAN without the use of NetWorker. The EMC Data Protection Advisor Software
Compatibility Guide provides information on supported databases.
• If monitoring the Enterprise App for backing up Oracle RMAN, follow the procedure provided in
"Monitoring Oracle RMAN"on page 88.

Monitoring Symantec Backup Exec

Monitor Symantec Backup Exec servers from an agent running on the Backup Exec server
or from an agent running on any other Windows computer in the environment. Also, the
DPA Agent service needs to run with a named account that can authenticate with the
BackupExec server.

Before starting the Discovery Wizard

To monitor a Symantec Backup Exec backup server remotely, the agent must run as a
named user account rather than the Local System account. When installing the agent, you
are prompted to specify whether the agent runs using the Local System account or as a
named user.
The Backup Exec Credentials must use the username and password of a Windows
administrator account on the Backup Exec server.
Select Admin > System > Manage Credentials to modify the Backup Exec Credentials that
are created after you have used the Discovery Wizard to create a Backup Exec object.

Monitoring Backup Exec Remotely

To verify that the agent is running, start the Windows Service Control Manager (Start >
Settings > Control Panel > Administrative Tools > Services). Right-click on the DPA agent
service and select Properties:
1. Select the Log On tab of the Service Properties panel.
2. Select This Account.
3. Type the username and password of the local administrator account to run the service.
4. Modify the service account details and click OK.
5. Restart the service to activate the changes.

Configuring the environment for discovery 89

Environment discovery in DPA

Monitoring Symantec NetBackup

Configure a Symantec NetBackup server to be monitored from an agent running on the
NetBackup Master Server or from an agent running on a different host such as the DPA
server.

Before starting the Discovery Wizard

Media Server Status data can only be collected if an agent is installed on the Media Server
itself. It cannot be collected through proxy.

Configuring NetBackup authentication for remote data collection

To gather data remotely, the following must be configured:
◆ The NetBackup Remote Administration Console, a component of the NetBackup Server
software, must be installed on the agent's host.
◆ The agent’s host must be able to successfully resolve the NetBackup Master Server.
◆ The NetBackup Master Server must be able to successfully resolve the agent’s host.
To resolve the agent host from the NetBackup Master Server:

Unix
If the NetBackup Master Server is running on a UNIX computer, the name of the host on
which the agent is running must be added to the bp.conf file on the NetBackup Master
Server. To add the host:
1. Open /usr/openv/netbackup/bp.conf for editing and add the following line:
SERVER = Agenthost

where Agenthost is the agent’s hostname. The agent’s hostname must be resolvable
by the Master Server.
2. Restart NetBackup on the Master Server for the changes take effect.

Windows
If the NetBackup Master Server is running on a Windows computer, add the name of the
agent host through the NetBackup Administration Console:
1. On the NetBackup Server, start the NetBackup Administration Console and open the
Master Server Properties dialog box:
• Select Netbackup Management > Host Properties> Master Servers.
2. Double-click Host in the right-hand panel.
3. In Master Servers Properties, Servers field, type the name of the agent host to the list
of additional servers that are allowed to access the Master Server.
4. Click OK.
5. Restart the NetBackup services. Alternatively, reboot the machine to activate the
changes.

90 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Monitoring Symantec PureDisk

Configure a Symantec PureDisk server to be monitored from an agent running on the
PureDisk Server or from an agent running on a different host. Symantec PureDisk can only
be monitored on SUSE Linux 10. The root user cannot be used to gather data from
PureDisk.

Before starting the Discovery Wizard

PureDisk servers implement a firewall that might prevent DPA from gathering data from
PureDisk or from communicating with an agent installed on the PureDisk server. To ensure
successful data gathering and communications, the following sections describe how to
configure the PureDisk server before configuring the server in DPA.
The configuration process depends on the version of PureDisk being monitored.

Manually configuring the firewall (versions of PureDisk earlier than 6.5)

To manually configure the firewall for data collection:
1. Log on to the PureDisk server as the root user.
2. Stop the PureDisk firewall by running the following command:
/etc/init.d/pdiptables stop

3. Edit the file /etc/puredisk/iptables-rules by inserting one of the following lines

directly after this line in the file:
-A INPUT –p icmp –j ACCEPT

Note: It is important that the line is inserted at the correct location in the file,
otherwise it might not take effect.

• If you are monitoring PureDisk with an agent installed on the PureDisk server, add
the following line:
-A INPUT –p tcp –m tcp –-dport 3741 –j ACCEPT

• If you are monitoring PureDisk from an agent running on a different host, add the
following line:
-A INPUT –p tcp –m tcp –-dport 10085 –j ACCEPT

4. Restart the PureDisk firewall by running the following command:

/etc/init.d/pdiptables start

Updating the IP tables rules (PureDisk version 6.5)

Manually configuring the firewall will not work for PureDisk version 6.5. To update the
PureDisk IP table:
1. Open the following file in a text editor:
/etc/puredisk/custom_iptables_rules

2. If the DPA agent is installed on the PureDisk server, add the following line to the rules
file (three columns separated by a tab):

Configuring the environment for discovery 91

Environment discovery in DPA

tcp {controller_host_ip} 3741

This allows connections from the controller host to the DPA agent on port 3741 on the
PureDisk server.
3. If the DPA agent is installed on a remote host, add the following line to the rules file
(three columns separated by a tab):
tcp {agent_host_ip} 10085

This allows connections from the agent host to the postgres database on port 10085
on the PureDisk server.
You can specify a single host or an entire subnet (by including a /mask), as in the
following example:
tcp10.64.205.0/24 10085

The /etc/puredisk/custom_iptables_rules file provides additional information on

configuring this file.

Monitoring Microsoft Exchange Server

An Exchange Server can be monitored for recoverability from an agent installed on the
same host as the Exchange Server or an agent installed remotely.

Note: Microsoft Exchange can only be monitored for replication analysis, and for system
information from the Exchange server host.

Before starting the Discovery Wizard

The account used to connect DPA to the Exchange server must be a domain user with
Exchange read-only administrator rights and local administrator rights. DPA does not
support replication analysis for two Exchange information stores on a cluster. To connect
to the exchange application you must have Exchange read-only administrator rights. To
retrieve the disks information from Windows you must be an operating system user with
local administrator rights.

Monitoring Databases
This section describes how to monitor Oracle, SQL Server, and PostgreSQL databases.

Monitoring Oracle
An Oracle database can be monitored from an agent running on the same host as the
Oracle server, or from an agent running on a different host, such as the DPA server.

Oracle client libraries

The DPA Agent requires that the Oracle client libraries be installed on the same machine
as the Agent in order to gather data from the Oracle database. Oracle client libraries are
not included in the DPA installer package.
For UNIX platforms, you must create or ensure that the symbolic links for the following two
files exists:

92 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

◆ libclntsh.so
◆ libocci.so
For example, if you install version 11.1 of Oracle, ensure the libclntsh.so is a symbolic link
to the libclntsh.so.11.1 library file.

Before starting the Discovery Wizard

To monitor an Oracle database for data protection data, the Agent must connect to the
database as an Oracle user. To gather data successfully, this user must be able to perform
selects on the following tables and views:
◆ V_$INSTANCE
◆ V_$PROCESS
◆ V_$DATABASE
◆ V_$PARAMETER
◆ DBA_DATA_FILES
◆ V_$SYSTEM_PARAMETER
◆ V_$DATAFILE
◆ V_$SESS_IO
◆ V_$SESSION
◆ DBA_FREE_SPACE
◆ V_$SESSMETRIC (Oracle 10 only)
◆ V_$BACKUP_DATAFILE
◆ V_$BACKUP_PIECE
◆ V_$RMAN_STATUS
Any user with the SYSDBA role will have these privileges by default, so it is recommended
that you specify a user that has the SYSDBA role when configuring the database for
monitoring.
If you do not want to use a user with the SYSDBA role to connect, then a separate user can
be created and explicitly granted permissions on those tables, as the following example
shows:
CREATE USER limited_user IDENTIFIED BY password;
GRANT CREATE SESSION TO limited_user;
GRANT SELECT ON V_$INSTANCE TO limited_user;
GRANT SELECT ON V_$PROCESS TO limited_user;
GRANT SELECT ON V_$DATABASE TO limited_user;
GRANT SELECT ON V_$PARAMETER TO limited_user;
GRANT SELECT ON DBA_DATA_FILES TO limited_user;
GRANT SELECT ON V_$SYSTEM_PARAMETER TO limited_user;
GRANT SELECT ON V_$DATAFILE TO limited_user;
GRANT SELECT ON V_$SESS_IO TO limited_user;
GRANT SELECT ON V_$SESSION TO limited_user;
GRANT SELECT ON DBA_FREE_SPACE TO limited_user;
GRANT SELECT ON DBA_TABLESPACES TO limited_user;
GRANT SELECT ON DBA_EXTENTS TO limited_user;

On version 10 of Oracle, add the following line:

Configuring the environment for discovery 93

Environment discovery in DPA

GRANT SELECT ON V_$SESSMETRIC TO limited_user;

Replication analysis
To monitor an Oracle database for replication analysis, the Agent must connect to the
database as an Oracle user able to perform selects on the following tables and views:
◆ DBA_DATA_FILES
◆ DBA_TEMP_FILES
◆ DBA_TABLESPACES
◆ V_$DATAFILE
◆ V_$LOGFILE
◆ V_$CONTROLFILE
◆ V_$LOG_HISTORY
◆ V_$ARCHIVED_LOG
◆ V_$INSTANCE
◆ V_$DATABASE
◆ V_$PARAMETER
◆ DICT
◆ DBA_TAB_COLUMNS
When monitoring Oracle on a Windows platform, the operating system user specified in
the Credential must belong to the group ORA_DBA. On UNIX, if UNIX authentication is
used, no further definition is required in the database.

Updating Oracle statistics

To gather accurate figures on the number of rows and size of tables and indexes, it is
important that Oracle statistics are updated on a regular basis. The Oracle documentation
contains more details on how to set up a job to update Oracle statistics.
One method to update Oracle statistics on a Schema is to run the following command:
exec dbms_stats.gather_schema_stats(ownname => '***SCHEMANAME***',
estimate_percent => 5, cascade => true, options => 'GATHER');

Monitoring Microsoft SQL Server

Monitor Microsft SQL Servers from an agent running on the SQL Server database, or from
an agent running on any other Windows computer in the environment. The DPA Agent
service needs to run with a named account that can authenticate with Microsft SQL
Servers.

Before starting the Discovery Wizard

To connect to SQL Server using Windows Authentication, the DPA Agent must run as a
named user with MS-SQL access and not as the Local System Account. Verify that the
service is running as the correct user before proceeding with the configuration of the
database.

94 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Agent requirements
The Agent needs to be able to connect to the SQL Server master database in order to
gather the data required. The Agent can either:
◆ Use SQL Server Authentication using the credentials of the request (if set).
◆ Use SQL Server Authentication using the credentials against an explicit master
database in the list of databases to be monitored (if set)
◆ If these are not set, the Agent uses Windows Authentication using the logon ID of the
agent process.
If none of these are sufficient to connect to the master database, the request will not
gather data.

User account requirements

To gather data successfully, the user account used to connect to the SQL Server database
must be granted specific privileges. Any SQL Server user with dbo access will have the
correct privileges by default.
If you do not want to connect with a user with dbo access, configure a user with the
following:
◆ The user must be mapped to the database with the public role.
◆ The user must be explicitly granted the VIEW SERVER STATE and VIEW DEFINITION
privileges (SQL Server 2005 only).
The VIEW SERVER STATE privilege is granted at the server level. The VIEW DEFINITION
privilege might be granted at the server level (under the name VIEW ANY DEFINITION)
or at the database, schema, or individual object level.

SQL Server 2005 and 2008

To grant server-wide privileges to the SQL Server login used by the Agent, including VIEW
DEFINITION privileges for all database tables, connect to the SQL Server as an
administrator and run:
GRANT VIEW SERVER STATE TO <login\domain>
GRANT VIEW ANY DEFINITION TO <login\domain>

However, to grant VIEW DEFINITION privileges for only the specific databases that you want
to monitor, connect to the SQL Server as an administrator and run:
GRANT VIEW SERVER STATE TO [login\domain]
GRANT VIEW DEFINITION ON DATABASE :: <dbname> TO <username>

Replication analysis
The DPA server must connect as a database user with connect privileges for all of the
databases and write privilege for the TEMPDB database. For Windows authentication, the
user must be able to connect to all SQL Server databases and should have write privilege
for the TEMPDB database.

Monitoring PostgreSQL
A PostgreSQL database can be monitored from an Agent running on the same host as the
PostgreSQL database or from an Agent running on a different host, such as the DPA server.

Configuring the environment for discovery 95

Environment discovery in DPA

Before starting the Discovery Wizard

To monitor a PostgreSQL database, the agent must connect to the database as a
PostgreSQL super user. A super user will have the correct privileges by default. EMC
recommends that you specify a super user when configuring the database for monitoring.
To create a super user, the PostgreSQL administrator must be a super user, and create the
account as in the following example:
CREATE ROLE xxxxx WITH login superuser password yyyyyy ;
where xxxxx is the new username and yyyyyy the new user's password.
The following parameters will not be populated in the database server parameters table
unless you are connecting to the database as a super user:
◆ config_file
◆ data_directory
◆ dynamic_library_path
◆ external_pid_file
◆ hba_file
◆ ident_file
◆ krb_server_keyfile
◆ log_directory
◆ log_file name
◆ preload_libraries
◆ unix_socket_directory
The following items are also unavailable unless you are connecting as a super user:
◆ In the datafile configuration table, the full path to the datafiles cannot be shown, as
the path of the file is found in the data_directory parameter. The string (postgres data
directory) is shown instead.
◆ In the connection status table, the f_command and f_status fields will not be
populated with the right information. These fields will be set to <insufficient
privileges>.
Connecting to the database as a super user will populate all fields.

Monitoring EMC RecoverPoint

EMC RecoverPoint must be monitored from an Agent installed remotely (the DPA server, for
example).

Before starting the Discovery Wizard

DPA needs to be able to connect to the RecoverPoint environment Command Line Interface
(CLI) through a secure SSH connection on port 22. DPA connects to the RecoverPoint
appliance using the default CLI user admin, but any defined user with sufficient privileges
to run a CLI command remotely using SSH is possible; the monitor account is sufficient.

96 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

However, DPA must not connect with the RecoverPoint user boxmgmt because user
boxmgmt is reserved for starting the RecoverPoint installation manager automatically.

Gathering data from operating systems

DPA can gather data from operating systems that enable users to report on operating
system configuration, status, and performance. There are several DPA modules that gather
different types of information, as described in Table 27 on page 97.

Table 27 System monitoring modules

Module Description

Host Gathers basic information about the operating system type.

Disk Gathers configuration, status, and performance information on the

disks attached to the host.

Fibre Channel Gathers configuration, status, and performance information on Fibre

HBA Channel HBAs configured on the computer.

File system Gathers configuration, status, and performance information on the

file systems mounted to the host.

Memory Gathers configuration, status, and performance information on

memory in the host.

NetInt Gathers configuration, status, and performance information on

network interface cards in the host.

Process Gathers information on any processes running on the host.

Processor Gathers configuration, status, and performance information on all

CPUs on the host.

UNIX
To perform system monitoring on UNIX computers, install an Agent on the host that is to
be monitored. It is not possible to gather system information remotely from UNIX
computers.

Discovering agent hosts for UNIX

UNIX hosts are discovered using SSH or telnet/ftp with root access.
If security requirements do not allow for root credentials to be supplied to DPA, sudo is a
workaround that can temporarily elevate a user's credentials to root for specific
commands configured in the sudoers file.

Modifying sudoers file for DPA storage discovery

A user can log in to a UNIX host as a non-root user, and use sudo to run SCSI commands
successfully to discover storage related information for the host. The following is an
example of what needs to be added to the sudoers file.
# sudoers file.
#
# This file MUST be edited with the 'visudo' command as root.
#
# See the sudoers man page for the details on how to write a sudoers
file.
#

Configuring the environment for discovery 97

Environment discovery in DPA

# Host alias specification

# User alias specification
# Cmnd alias specification
# Defaults specification
# User privilege specification
root ALL=(ALL) ALL
# Uncomment to allow people in group wheel to run all commands
# %wheel ALL=(ALL) ALL
# Same thing without a password
# %wheel ALL=(ALL) NOPASSWD: ALL
# Samples
# %users ALL=/sbin/mount /cdrom,/sbin/umount /cdrom
# %users localhost=/sbin/shutdown -h now
user_alias ALL = (ALL) PASSWD: /var/tmp/IllumAgent/apolloreagent

# Defaults specification
# User privilege specification
root ALL=(ALL) ALL
CMGU ALL=NOPASSWD:CMGEMC
# Uncomment to allow people in group wheel to run all commands
# %wheel ALL=(ALL) ALL
# Same thing without a password
# %wheel ALL=(ALL) NOPASSWD: ALL
# Samples
# %users ALL=/sbin/mount /cdrom,/sbin/umount /cdrom
# %users localhost=/sbin/shutdown -h now
#cmguser ALL=(ALL) NOPASSWD: ALL

Windows
To gather performance data from a Windows host, Windows Management Infrastructure
(WMI) must be installed on the Windows host you are monitoring.
It is possible to gather all system monitoring information remotely from Windows
computers, with the exception of Fibre Channel HBA information. To gather Fibre Channel
HBA information, the agent must be installed on the computer. "Monitoring a Windows
host remotely"on page 99 provides more details on the steps required to monitor a
Windows host remotely.
To set up system monitoring for a system on which an agent is installed, assign the system
monitoring requests to the host or group to monitor.

Discovering agent hosts for Windows

If application discovery is being performed without an agent, Windows host discovery
uses Remote Procedure Calls (RPC) for replication analysis and WWI for System
information.

Checking RPC Communication

1. Open the Run dialog box from the Windows Start menu.
2. Type:
net use \\<servername>\admin$ /user:<username>

3. Click Enter. Type the password.

4. A successful connection should return the following message: "The command
completed successfully."
5. Delete the network map. Type:

98 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

net use \\servername\admin$ /delete

Checking WMI Communication

1. Open the Run dialog box from the Windows Start menu.
2. Type WBEMtest and click Connect in the Windows Management Instrumentation Tester
dialog box.
3. In the Connect field, type \\<servername>\root\cimv2.
4. In the Credentials fields, type the username and password used to connect to the
application host you are monitoring.
5. Click Connect to return to the Windows Management Instrumentation Tester dialog
box. Click Query.
6. In the Enter Query field, type:
select * from win32_processor

7. Click Apply. If WMI can connect, data from the application host is displayed.

Monitoring a Windows host remotely

All system information can be gathered remotely from a Windows computer with the
exception of Fibre Channel HBA information. To monitor a Windows computer remotely, an
agent must be installed on another Windows computer. It is not possible to remotely
monitor a Windows computer from an agent running on a UNIX computer.
To monitor a Windows host from another Windows computer, the DPA Agent service must
run as administrator on the computer performing the monitoring. "Modifying the login
parameters of the agent service"on page 99 provides more information.

Modifying the login parameters of the agent service

Checking if this is required. To modify the login parameters of the agent service:
1. Start the Windows Services control manager (Start > Settings > Control Panel >
Administrative Tools > Services).
2. Select the DPA Agent service.
3. Right-click and select Properties from the menu.
4. Select the Log On tab in the Properties dialog box.
5. Select This Account.
6. Type the username and password of the administrator that the service to run as.
7. Click OK and restart the service.

Monitoring activity on a remote computer

To monitor activity on a remote computer:
1. Create a host object for the computer to monitor in the web console. The name of the
object is the hostname of the remote host. The hostname must be resolvable from the
computer on which the Agent that will be monitoring the object is running.
2. Assign requests to that object to specify the data to gather.

Configuring the environment for discovery 99

Environment discovery in DPA

3. Mark each request as a proxy request and complete the details.

4. To complete the proxy details, type the name of the host for the Agent in the Proxy
Host field.
5. Create a Windows credential for the Administrator account on the computer being
monitored. This account can be the name of a Local Administrator or that of a Domain
Administrator.
6. Notify the Agent that will monitor the server of the changes by reloading the Agent.

Monitoring a host for system data

Monitor an application host for system data from an Agent running on the host or another
host in the environment.

Before starting the Discovery Wizard

System data can only be gathered from UNIX systems by an Agent local to the UNIX host.

Monitoring tape libraries

DPA can gather information about tape libraries and the drives within those tape libraries.
When specifying a hostname, it is important that the name of the tape library is resolvable
from the host that is monitoring the tape library.

Before starting the Discovery Wizard

The tape library credentials must contain the read-only community string for the tape
library in the Password field of the Credential Properties dialog box. Unless the community
string was modified on the tape library, set the community string to Public.
Select Admin > System > Manage Credentials to modify the tape library credentials that are
created after using the Discovery Wizard to create a tape library object.

Monitoring the IBM System Storage TS 3500 tape library

Use the Tape Library Specialist web interface to enable Simple Network Management
Protocol (SNMP) requests for the IBM System Storage TS 3500 Tape Library. To enable
SNMP requests:
1. Type the Ethernet IP address on the URL line of the browser.
2. Select Manage Access > SNMP Settings. In the SNMP Trap Setting field, view the
current setting then click to enable SNMP requests.
3. Ensure that the SNMP Requests Setting field is set to Enabled.

Monitoring the IBM TotalStorage 3583 tape library

Configure the Remote Management Unit (RMU) to enable SNMP for the IBM TotalStorage
3583 Tape Library. To enable SNMP:
1. In the RMU, click Configuration.
2. In the SNMP Configuration region, perform the following:
• To enable the feature, select ON in the SNMP Enabled field.

100 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

• To enable or disable SNMP alerts, select ON or OFF in the Alerts Enabled field.
• In the Manager field, type the SNMP server address.
• In the Public Name field, type the name of the read-only SNMP community.
• In the Private Name field, type the name of the read/write SNMP community.
3. Click Submit and review the changes.
4. Type the password and click Confirm. Redirect the browser if required.
5. Click Done to reboot.

Monitoring the IBM TotalStorage 3584 tape library

To enable SNMP from the web interface of the IBM TotalStorage 3584 tape library:
1. From the Welcome screen of the Tape Library Specialist Web Interface, select Manage
Access > SMNP Settings.
2. In the SNMP Trap Setting field, view the current setting, and select the button to
enable or disable SNMP requests.
Alternately, to enable SNMP requests from the operator panel:
1. From the Activity screen of the tape library operator panel, select MENU > Settings >
Network > SNMP > Enable/Disable SNMP Requests > ENTER. The screen displays the
current status of SNMP requests.
2. Press UP or DOWN to specify ENABLED or DISABLED for SNMP messaging, and click
ENTER. The Enable/Disable SNMP Requests screen re-displays the new setting. To
accept the new setting and return to the previous screen, click BACK.

Monitoring the Oracle SL24 Tape Autoloader and SL48 tape library
Configure the Remote Management Interface (RMI) to enable SNMP for the Oracle
StorageTek SL24 Tape Autoloader or SL48 Tape Library. To enable SNMP:
1. In the RMI, navigate to Configuration > Network.
2. Ensure the SNMP Enabled checkbox is enabled.
3. The Community Name string must be contained in the credentials used to connect to
this Tape Library in DPA.
4. Click Submit and review the changes.

Monitoring the HP StorageWorks tape library

Configure the NeoCenter utility to enable SNMP for the tape library. To enable SNMP:
1. Start the NeoCenter utility from the host.
2. Select Configure from the Main screen menu. The Configure dialog box appears.
3. Select the SNMP Traps tab.
4. In one of the available Trap Address fields, type the IP address of the DPA server.

Switches and I/O Devices

DPA monitors Fibre Channel switches, IP switches, and Xsigo I/O Director devices.

Configuring the environment for discovery 101

Environment discovery in DPA

Monitoring Fibre Channel switches

DPA gathers information about ports on Fibre Channel switches, including configuration,
connectivity status, and throughput.
When specifying a hostname, it is important that the name of the switch is resolvable on
the Agent’s host.

Before starting the Discovery Wizard

To ensure that Brocade switches return all data, verify that the Fibre Channel Alliance MIB
is loaded and enabled on the switch. This MIB might not be installed on the switch by
default. To enable FA-MIB support on Brocade switches, log in as an administrator and run
the snmpmibcapset command. Change the FA-MIB parameter to Yes. Click Enter to accept
the default for the other settings.
For example:
telnet <switch>
> snmpmibcapset
The SNMP Mib/Trap Capability has been set to support
FE-MIB SW-MIB FA-MIB SW-TRAP FA-TRAP
FA-MIB (yes, y, no, n): [yes]
SW-TRAP (yes, y, no, n): [enter]
FA-TRAP (yes, y, no, n): [enter]
SW-EXTTRAP (yes, y, no, n): [enter]
>

Monitoring IP switches
When you are specifying a hostname, ensure the name of the switch is resolvable on the
Agent’s host.

Before starting the Discovery Wizard

The IP Switch Credentials must contain the SNMP community string for the IP switch in the
Password field of the Credential Properties dialog box. Unless the community string was
modified on the IP switch, set the community string to public.
Select Admin > System > Manage Credentials to modify the IP Switch Credentials that are
created after you have used the Discovery wizard to create an IP switch object.

Monitoring Xsigo I/O Director

When you are specifying a hostname for the Xsigo I/O Director, ensure the hostname or IP
address of the Director is resolvable on the Agent’s host.

Before starting the Discovery Wizard

The Xsigo Director SNMP credentials must contain the SNMP community string for the
Director in the Password field of the Credential. Unless the community string was modified
on the Director, set the community string to public.
Select Admin > System > Manage Credentials to modify the default Xsigo Director SNMP
Credentials if required, or to create a new credential.

102 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

File Servers
DPA supports monitoring EMC File Storage.

Note: EMC File Storage is interchangeably referred to as Celerra File Storage.

Monitoring EMC File Storage

EMC File Storage must be monitored from an Agent running on a remote computer, for
example, the DPA server.

Before starting the Discovery Wizard

The EMC File Storage module gathers information from EMC File Storage through an XML
API and directly from the EMC File Storage Control Station. An administrator with specific
privileges must be created on the EMC File Storage:
1. Log in to the EMC File Storage Manager web browser interface as an administrator.
(You can also use the command line interface to create a DPA administrator.)
2. Navigate to Security > Administrators.
3. Create a new administrator, with a username of DPA, for example.
4. Select Local Only Account and type and confirm a password for the administrator.
5. Select a Primary Group of at least opadmin level of privilege. DPA does not need
greater privileges than those assigned by opadmin.
6. Enable the following client access options:
• XML API v2 allowed
• Control Station shell allowed
7. Click OK.
The DPA Credential used to connect to the EMC File Storage must contain the username
and password of the EMC File Storage administrator you created.

Backup appliances
DPA supports monitoring EMC Data Domain.

Monitoring EMC Data Domain

DPA monitors EMC Data Domain backup appliances. For EMC DDOS 4.8, only Tape Drive
and Tape Library Status and Configuration information is returned.

Before starting the Discovery Wizard

SNMP on port 161 and SSH on port 22 must be enabled on the EMC Data Domain backup
appliance. You also need to set the SNMP community string. This can be done from the
command line.
Ensure that you have user role rights to run SSH requests on the Data Domain system.
To enable SNMP on the EMC Data Domain appliance:

Configuring the environment for discovery 103

Environment discovery in DPA

1. Log on to the EMC Data Domain appliance console using the sysadmin account.
2. Type the following command to check the existing configuration:
snmp show ro-communities
snmp add ro-community <string> hosts <host IP address>

where <string> is the selected community string (for example, public) and <host IP
address> is the IP address of the DPA Agent that you are using to monitor the Data
Domain. You will have to disable and re-enable SNMP for the new string to take effect.
snmp disable
snmp enable

If you are not using a community string of public, you must change the community
string used in the EMC Data Domain Credential.
3. Edit the EMC DPA Data Domain SSH Credential to specify an SSH username and
password configured on the Data Domain device. Go to Admin > System > Manage
Credentials in the DPA web console.
SNMP settings can also be set through the System Settings tab of the EMC Data Domain
Enterprise Manager interface.

Monitoring Oracle ACSLS

ACSLS cannot be monitored remotely. A DPA Agent must be installed on the ACSLS Solaris
or AIX host.

Before starting the Discovery Wizard

The Agent must be installed and running on the ACSLS server that you want to monitor.
After installing the Agent, verify that the ACS_HOME value in the DPA.config file matches
the location in which ACSLS is installed. Verify that the ACSDBDIR value in the DPA.config
file matches the path to the ACSLS DB folder (the default is export/home/ACSDB 1.0).

Disk management servers

DPA monitors HP Command View.

Monitoring HP Command View

Monitor a HP EVA Disk Array through HP Command View from an Agent running on the
Command View host, or remotely from an Agent running on a different host (such as the
DPA server).
The username and password used to gather data must match a valid username and
password defined in the CommandView CIM server (configurable from the CommandView
management interface).
DPA gathers data from HP Command View using SMI-Son the default secure port of 5989.

Monitoring VMware environment

Monitor your VMware environment from an Agent running on the VirtualCenter Server or
remotely from an Agent running on a different host (such as the DPA server).

104 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

◆ The Discovery Wizard can be used to add a vCenter server to DPA. Go to Admin >
System > Discovery Wizard > Virtualization Management.
◆ To add a vCenter server, you must provide the vCenter hostname and credentials for a
vCenter user with administrative privileges.
◆ You can select whether to monitor the vCenter host only or to also monitor the virtual
machines connected to the vCenter host.
• If you select to monitor virtual machines, DPA queries the vCenter Server and
displays a list of virtual machines. Note that the discovery process can take a while
if there are a large number of virtual machines configured on the vCenter server.
• For each virtual machine you can select whether you wish to discover the host in
DPA. Discovering the host adds the host to the DPA inventory.
• For each virtual machine selected for discovery, you can select whether to enable
Host System Monitoring., which gathers configuration, performance and analysis
data; and Replication Monitoring, which enables replication analysis.
• For each virtual machine selected for Host System Monitoring, you can specify
which DPA Agent should be used to monitor the virtual machine. You can change
the DPA Agent for multiple machines simultaneously by using CNTRL-Click or
SHIFT-Click to select multiple systems.
– Windows virtual machines can have Host System Monitoring performed using a
remote DPA Agent such as the DPA Agent installed on the DPA Server; or a local
agent, such as DPA Agent installed on each Windows virtual machine.
– UNIX / Linux virtual machines must have a DPA Agent installed on the virtual
machine for Host System Monitoring, on a local agent.
• You must provide Windows credentials for each Windows Virtual Machine being
monitored with a remote agent. The credentials can either be a local administrator
or a domain administrator. You can change the credential for multiple machines
simultaneously by using CNTRL-Click or SHIFT-Click to select multiple systems.
• Discovered virtual machines are displayed under the vCenter object in DPA and by
default will also be added to Configuration / Servers / Application Servers group.
You can change and add groups for the virtual machines to appear. Go to Admin >
System > Discovery Wizard > Destination Group.
◆ The final screen of the vCenter Discovery Wizard displays a summary of options
selected. Clicking Finish adds the objects to DPA and enables monitoring options
selected.

Monitoring backup servers in a Symantec Cluster Server and Microsoft Cluster Server
environment
To monitor backup servers in Symantec Cluster Server and Microsoft Cluster Server
(MSCS) environments, an Agent must be installed within the clustered environment so
that only one version of the Agent is running on a physical host. This Agent should be
configured within the DPA Server as the name of the Cluster Alias itself rather than that of
the physical host that it is running on.

Configuring the environment for discovery 105

Environment discovery in DPA

Supported platforms
◆ Symantec Cluster Server is supported on Linux and Solaris
◆ MSCS is supported on Windows
The EMC Data Protection Advisor Software Compatibility Guide provides more information
on supported platform versions.

Monitored objects and groups

Objects
DPA discovers the applications and devices in your data protection environment and
stores these logical and physical entities as objects in the object library. Discovered
objects are grouped into the following categories in the object library:
◆ Applications
◆ Hosts
◆ Storage
◆ Switches
The following rules apply to objects:
◆ No two objects can share the same name
◆ No object can share its name with an alias of another object
The object library enables you to view objects and their attributes. To view the object
library, select Inventory > Object Library.

Data collection editing for objects

As part of the discovery process, the DPA Discovery Wizard assigns data collection
requests directly to an object during object creation. To edit the default data collection
requests for a specific object:
1. Select Inventory > Object Library.
2. Select a host and then click the Data collection tab.
3. Click Properties.
4. Select a request and then click Edit.
"Manage Data Collection Defaults"on page 83 provides information on default data
collection requests. The DPA online help set provides procedures to add, edit and view
data collection requests.

Groups
A group is a collection of objects. For example, you can create a group of objects that are
used by an application. This way, when you apply a policy to the group, the policy is
applied to all of the objects within the group.

106 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Note: An object can however exist in more than one group.

Configuration group
The Configuration group is created by default. The Configuration group is created with an
initial structure that groups the data protection environment into Servers, Switches, and
Storage. All data protection hosts, devices, and applications discovered by the Discovery
Wizard are first added to the Configuration group. Objects that are removed from the
Configuration group are not deleted. Objects removed from Configuration group appear
under Objects Not In Groups.

Object attributes
Object attributes extend the information that DPA holds about an object. After a custom
attribute is created, the attribute can be enabled for any valid objects as per custom
attribute settings and a value can be assigned.
When creating or editing an object, attributes are filtered to be associated with one or
more specific types of objects, and only to objects with an existing attribute that matches
a given value.
For example, an Asset Tag attribute might be created to represent an asset identifier for
the physical components of an operating environment (such as hosts, storage arrays, and
switches). The Asset Tag attribute need not be assignable to logical components like
database instances or processes.
In the attribute definition, the Asset Tag is configured to be associated with a subset of
physical object types. You can further configure this attribute to only be associated with
physical object types that have an attribute of Business Unit, for example.

Smart groups
Smart groups allow users with administrative privileges to create groups that are
populated dynamically with information from the results of DPA reports. A smart group
runs a custom report and then creates objects based on the results of the report.
The main benefit of smart groups is that they provide high levels of flexibility.
Administrators can set up smart groups to dynamically create lists of objects that match
specific business and technical criteria.
Children of Smart Groups are not filtered by role permissions. Even if you are not allowed
to see some hosts in the inventory, you can still see them in the Smart Group. Hosts that
are filtered by role permission remain not viewable from internal Smart Groups, such as All
Hosts.
To create a smart group, select Inventory > Group Management. Click Create Group and
then Create Smart Group.
Observe the following consideration when creating Smart Groups: Avoid creating Smart
Groups that contain themselves, either directly or through other Smart Groups, when
expanded. The DPA web console hangs if Smart Groups are created under these
conditions. This is pertaining to problem DPA-26241.

Monitored objects and groups 107

Environment discovery in DPA

Report output objects

Report output creates objects that return a single set of objects from a report, contained in
one level of hierarchy, depending on the report and objects selected.
You can assign the same items that you can assign to regular objects, including analysis
and scheduled reports. DPA can then generate alerts and reports for a smart group
outputting objects.
For example, a financial firm might have a convention where the first two characters of
each backup client indicate the business unit to which the client is assigned. If the first
two characters are a and m, then the backup client belongs to the asset management
group. Due to the nature of the business, a large number of clients are created, renamed,
or removed daily. Rather than spend a lot of time updating the group configuration each
day, the DPA administrator can create a smart group that uses the existing Backup Client
Configuration report to list each backup client. In the smart group, the administrator can
filter the results to only contain clients that start with a and m.
As DPA automatically updates the client configuration list every time it obtains data from
the backup server, this list is kept up-to-date with whatever changes are made within the
backup environment.
Other examples include:
◆ All backup clients containing exch.
◆ All hosts with an E: drive.
◆ All objects with severity 1 alerts in the last day.
To keep the Smart Group children list updated, the user with Administrator privileges can
select a generation frequency to once a day, schedule, or update on-demand.

Policies and rules

Overview
DPA contains customizable policies and rules that control how DPA generates alerts,
measures backup and replication performance, and determines values for chargeback
reporting.

Policies
Data Protection Advisor policies are a collection of user data about how backup and
replication should operate in the environment (recoverability and data protection policies)
or about the cost of storage and data protection operations (chargeback policies).
Recoverability, backup, and service level management reports then show how the
operations in the environment compare to the policy settings, for example, gaps in the
recoverability chain for a storage array, or if a backup server is not meeting a Recovery
Point Objective.
DPA provides the following policy types:

108 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

◆ Analysis policies - are a collection of one or more rules that are used primarily for
generating alerts. Alerts are displayed by default in the Advisor section. You can edit
the policy to send events to emails, scripts, SNMP traps, or Windows Event Logs.
"Generating events from policies"on page 124 provides more information.
◆ Protection policies - are a collection of user data about how backup and replication
should operate in the environment. These policies consist of recoverability and
protection rules. These are used primarily for generating alerts. Alerts are displayed
by default in Alerts and in the Replication Analysis of the Advisor section.
◆ Chargeback policies - are used to determine the cost of storage and data protection
operations for chargeback reports.
By default, analysis, protection, and chargeback policies are off for all objects and groups.

Analysis policies
An analysis policy is a collection of one or more rules that is assigned to an object or
group. Rules contain the logic for when to issue an alert. The analysis engine compares
monitored data to the conditions in a rule, and triggers alerts when a rule is matched.
Event-based rules trigger an alert in response to data that is streaming into the DPA server.
Schedule-based rules periodically compare data in the DPA Datastore server against rules
to detect a match. Alerts can contain dynamic textual information and might include
populated links to reports. Only analysis policies can generate alerts.

Analysis rule template

An analysis rule template is a set of instructions that defines the rules logic. When a rule
template is added to an analysis policy, the Analysis Engine carries out certain operations
and then displays the resulting events in the Advisor section of the web console.
A rule template consists of the name of the rule along with details that specify how that
rule is run.
For example, a rule template can be created to monitor whether a file system is likely to
exceed 90% utilization in the next hour.
An Analysis Policy contains multiple rules that apply to different object types. The Analysis
Engine only runs the rules that are applicable to a given object. For example, if the object
is a switch, then the Analysis Engine will only run the rules in the policy that apply to
switches.

Creating an analysis rule

Use the DPA rule editor to create an analysis rule template. The following is a high-level
overview of the process. The online help available in the DPA web console provides
detailed instructions.
1. In the DPA web console, navigate to Policies > Analysis Policies > Rules Templates.
2. Click Create Rule Template. This open the rules editor.
3. Provide a name and description for the alert that is triggered by this rule.
4. Select a category associated with the rule.
The DPA online help provides information on rule categories and descriptions.

Policies and rules 109

Environment discovery in DPA

5. Specify whether the rule is event based or a scheduled rule.

An event-based rule triggers an alert in response to data that is streaming into the DPA
server. A Schedule-based rule runs periodically to check whether to issue an alert.
6. Select the appropriate object types:
• by hierarchy
• by function
7. Define when and how the alert must be triggered.
Note that DPA does not support the option to test the Lack of event trigger for Number
of samples, even though the option still appears as valid in the DPA web console. DPA
still supports this option for Time window.

Adding an analysis rule to an Analysis Policy

After a rule template is added to an Analysis Policy, the Analysis Engine carries out certain
operations and then displays the resulting events in the Advisor section of the web
console.
The Analysis Policies can contain multiple analysis rules that apply to different types of
objects. DPA automatically applies the appropriate rules from the applied Analysis Policy
to an object. For example, DPA applies rules for switches to switches only, not to backup
servers.

Analysis policy rule categories

Capacity planning
Capacity planning analysis policies create alerts about events that indicate that resources
might soon run out. Table 28 on page 111 describes these jobs.

Assigning alerts for pools and storage array analysis policies

When assigning the following analysis policies to objects, the recommended severity
levels are:
◆ Pool is Filling Up - Severity 3
◆ Pool is Filled Up - Severity 2
◆ Storage Array is Filling Up - Severity 1

110 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Table 28 Capacity planning

Name Description Rule Parameters

File system May Be Generates alerts if a file File system Max Predicted
Full In 2 Weeks system utilization will filling up Utilization - 100%
exceed 90% in the next 2 Number of hours
weeks. to forecast 336

Less Than 25 Generates alerts if the Running out of Maximum client

Available Backup license only permits you to backup client licenses - 25
Client Licenses monitor less than an licenses
additional 25 computers.

Pool is Filling Up Alerts when according to Pool is Filling Minimum Free

the growing trend there will Up Space Allowed - 0
not be space left on the Days to Forecast -
pool for the selected time 90
period.

Pool is Filled Up Alerts when there is no Pool is Filled Initial Consumed

space on the pool to Up Capacity - 3
physically allocate a new
LUN.

Storage Array is Alerts when there is no Storage Array Initial Consumed

Filling Up space left to allocate a new is Filling Up Capacity - 2
LUN on the pool and there
are no free disks available
on the storage array.

Tape Pool May Be Generates alerts if there will Running out of Maximum
Out of Empty Tapes be no empty tapes empty tapes Predicted Count -
within 6 weeks available in a tape pool 0
within 6 weeks. Number of hours
to forecast - 1008

TSM Database May Generates an alert if the TSM Database Number of Hours
be Full in Two TSM Database is predicted Filling Up to Forecast - 336
Weeks to reach 100% usage within Maximum
2 weeks. Predicted
Utilization - 100

TSM Recovery Log Generates an alert if the TSM Recovery Number of Hours
May be Full in Two TSM Recovery log is Log Filling Up to Forecast - 336
Weeks predicted to reach 100% Maximum
usage within 2 weeks. Predicted
Utilization - 100

Policies and rules 111

Environment discovery in DPA

Change management
Change management analysis policies alert about changes in the environment. Table 29
on page 112 describes these jobs.

Table 29 Change management

Name Description Rule Parameters

Backup Client Generates alerts if the Backup client N/A

Configuration configuration of a backup configuration
Changed client has been modified. changed

Backup Device Generates alerts if the Backup device N/A

Configuration configuration of a backup configuration
Changed device has been modified. changed

Backup Group Generates alerts if the Backup group N/A

Configuration configuration of a backup configuration
Changed group has been modified. changed

Disk Firmware Level Generates alerts if the Disk firmware N/A

Changed firmware level of a disk has level changed
changed.

Disk Serial Number Generates alerts if a disk Disk serial N/A

Changed serial number has changed. number changed

Object Operating Generates alerts if the Object operating N/A

System Changed operating system of a object system changed
has changed.

RecoverPoint Active Generates an alert if the RecoverPoint N/a

RPA Changed active RPA has changed Active RPA
since the last analysis run. changed

RecoverPoint RPA Generates an alert if the RecoverPoint RPA N/a

Link Status status of the RPA link has Link Status
Changed changed since the last Changed
analysis run.

Tape Drive Firmware Generates alerts if the Tape drive N/A

Level Changed firmware level on a tape firmware level
drive has changed. changed

Tape Drive Serial Generates alerts if the serial Tape drive serial N/A
Number Changed number of a tape drive has number changed
changed.

112 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Configuration
The configuration analysis policies monitor the environment for device or application
configuration issues. Table 30 on page 113 describes these jobs.

Table 30 Configuration

Name Description Rule Parameters

Backup Client Generates alerts if a Backup client inactive N/A

Inactive backup client is not
scheduled to run.

Fileserver export Generates alerts if a Fileserver export and LUN N/A

and LUN on same fileserver export is on on same volume
volume the same volume as
a LUN.

LUN on vol0 Generates alerts if a LUN on given volume Volume - vol0

LUN has been
configured on vol0.

Network Generates alerts if IP autonegotiation N/A

Autonegotiation there is an mismatch
Mismatch autonegotiation
mismatch between a
host and its switch
port.

Network Duplex Generates alerts if IP duplex mismatch N/A

Mismatch there is a duplex
mismatch between
object and switch.

Virtual Memory Generates alerts if Not enough virtual N/A

Less Than 1.5x the amount of virtual memory
Physical Memory memory on a
computer is less than
1.5 times the amount
of physical memory.

Volume Priority Generates alerts Volume priority not N/A

Not Normal when volume priority normal
is set to something
other than normal.

Policies and rules 113

Environment discovery in DPA

Data protection
The data protection analysis policies monitor the environment for exceptions related to
backup and recovery issues. Table 31 on page 114 describes the monitored jobs.

Table 31 Data protection (page 1 of 2)

Name Description Rule Parameters

Application Restore Generates alerts if it is Application Recovery time

Time Estimate More estimated that it will take restore time objective - 12
Than 12 Hours more than 12 hours to estimate too high hours
restore an application.

Application Alert if an application has Application Recovery point

Unprotected For 72 not had a successful recovery point objective - 72
Hours backup in more than 72 objective missed hours
hours.

Backup Failed Alert generated if a backup Backup failed N/A

fails.

Backup Failed More Alert generated if a backup Backup failed Maximum

than Twice In A Row fails two consecutive times. many times failures - 2

Backup More Than Generates an Alert if a Backup larger Days of history

200% of Average backup Job is double its than average - 14 days
Size size of its average size over Deviation -
the last 14 days. 100%

Backup Not Alert is generated if a host Backup not Maximum days

Succeeded For At has not had a backup in the occurred for many not backed up
Least 3 Days last 3 days. days - 3

Backup Running at Generates an alert if there Backup Running None.

Same Time as TSM were any backups at Same Time as
Server Operation completed over a period Server Operation
that overlapped with any of
the following operations on
the backup server:
• Delete volumes
• Expirations
• Storage pool copies
• Moves
• Database backup
• Migrations
• Reclamations

Backup Spans More Alert is generated if a Backup spans Maximum

Than 3 Tapes backup spans more than 3 multiple tapes number of
tapes. tapes - 3

Full Backup Less Generates alerts if a Full Full backup Days of History
Than 50% of backup is less than 50% of smaller than - 14 days
Average Size its usual size. average Deviation -
50%

Full Backup Not Generates alerts if a host Full backup not Maximum
Succeeded for At has not had a successful occurred for many Days Not
Least 14 days full backup in the last 14 days Backed Up -
days. 14

114 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Table 31 Data protection (page 2 of 2)

Name Description Rule Parameters

Mirror Not Updated Generates alerts if a Mirror not Maximum

For At Least 2 Days Remote Disk Mirror has not updated for a Exposure - 48
been updated in at least 2 number of hours hours
days.

More than 7 Generates alerts if there Too many Maximum Non

Backups Without A have been more than seven backups without a Fulls - 7
Full runs of a backup Job since full
the last Full backup.

No NetWorker Generates an alert if there No NetWorker Maximum

Bootstrap has not been a NetWorker bootstrap hours without
Generated in Last 48 bootstrap ran in the last 48 generated bootstrap -
Hours hours. defaults to 48
hours

TSM Database Generates an alert if a TSM Database None.

Backup Running at database backup process Backup Running
Same Time as TSM completed while there was at Same Time as
Server Operation other activity on the backup Server Operation
server, including other
backups

TSM Database Alerts if there was a TSM TSM Database Time - 24

Backup Occurred in database backup in the last Backup Occurred Hours
Last 24 Hours 24 hours, or returns the last
TSM backup time if there
was no backup.

Licensing
The licensing analysis policies monitor the environment and generate alerts about
licensing issues. Table 32 on page 115 describes these policies in more detail.

Table 32 Licensing

Name Description Rule Parameters

License Expired Generates an alert if a License expired N/A

license in DPA has
expired.

License Will Generates an alert if a License nearing Minimum days

Expire in the license will expire in expiration before expiry -
Next Week the next week. defaults to 7 days

Policies and rules 115

Environment discovery in DPA

Performance
The performance analysis policies monitor the environment and generate performance
problem alerts. Table 33 on page 116 describes these jobs in detail.

Table 33 Performance

Name Description Rule Parameters

Backup Less Than Generates an alert if the Backup slower Days of history -
50% of Average performance of a backup than average 14
Speed job is 50% less than its Deviation - 50%
average over the last 2
weeks.

Backup Running For Generates an alert if a Backup Max Runtime -

More Than 18 Hours backup has been running overrunning 18 hours
for more than 18 hours.

Fileserver Cache Hit Generates alerts if the Fileserver cache Minimum cache
Rate Below 80% cache hit rate of a fileserver hit rate low hit rate - 80%
drops below 80%.

Full Backup Ran At Generates an alert if a full Full backup Minimum

Less Than 300 backup ran at less than 300 succeeded but expected speed -
KB/sec KB/sec. slow 300 KB/sec

Provisioning
The provisioning analysis policies generate alerts about events that might require
provisioning operations. Table 34 on page 116 describes the jobs.

Table 34 Provisioning

Name Description Rule Parameters

File system Peak Generates alerts if the peak File system Days to examine
Snapshot Usage snapshot usage over the snapshot space usage - 14
Below 80% last 14 days is less than under utilized Minimum peak
80%. snapshot usage -
80%

116 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Resource utilization
Resource utilization analysis policies generate alerts about events that have occurred
because of resource utilization problems within the environment. Table 35 on page 117
describes these jobs in detail.

Table 35 Resource utilization (page 1 of 2)

Name Description Rule Parameters

Aggregate Snapshot Generates an alert if an Aggregate Maximum

Utilization Above aggregate snapshot snapshot aggregate
90% utilization is higher than a utilization high snapshot
specified threshold. utilization -
default is 90%

CPU Utilization At Generates an alert if the CPU pegged Maximum CPU

Least 90% For Last CPU Utilization on a host is utilization -
30 Minutes greater than 90% for last 30 defaults to 90%
minutes. Number of
minutes - 30
minutes

Disk Activity At Least Generates an alert if a disk Disk pegged Maximum Disk
90% For Last 30 on a host is greater than Busy Percentage
Minutes 90% busy for over 30 - 90%
minutes. Number of
minutes -
defaults to 30
minutes

Fibre Channel Port Generates an alert if a Fibre Fibre Channel Maximum

Above 70% of Rated Channel port exceeds 70% port utilization utilization -
Throughput of its max throughput. high 70%

Fibre Channel Port Generates an alert if a Fibre Fibre Channel N/A

Out of BB Credits Channel port has ran out of port no BB
buffer to buffer credits. credits

File system File Generates an alert if the File system file Maximum file
Utilization High number of files on a file utilization high system file
system is greater than 90% utilization -
of the max number allowed. 90%

File system Generates an alert if a file File system Maximum file

Snapshot Utilization systems snapshot snapshot system
Above 90% utilization is above 90%. utilization high snapshot
utilization -
defaults to 90%

File system Generates alerts if a file File system Maximum file

Utilization Above system utilization is above utilization high system
90% and Increasing 90% and is increasing. and increasing utilization -
defaults to 90%

Memory Utilization Generates an alert if Memory Maximum

Above 90% memory utilization on a utilization high memory
host is greater than 90%. utilization -
defaults to 90%

Network Interface Generates an alert if a Network Maximum

Above 70% of Rated network interface exceeds utilization high utilization -
Throughput 70% of its rated defaults to 70%
throughput.

Policies and rules 117

Environment discovery in DPA

Table 35 Resource utilization (page 2 of 2)

Name Description Rule Parameters

RecoverPoint Journal Generates an alert if the RecoverPoint Warning

Utilization High journal utilization for an Journal threshold
RPA is above a specified Utilization High Critical
warning or critical Threshold
threshold.

RecoverPoint RPA Generates an alert if the RecoverPoint Warning

SAN Usage High SAN utilization for an RPA is Journal threshold
above a specified warning Utilization High Critical
or critical threshold. Threshold

RecoverPoint RPA Generates an alert if the RecoverPoint Warning

WAN Usage High WAN utilization for an RPA RPA WAN Usage threshold
is above a specified High Critical
warning or critical Threshold
threshold.

RecoverPoint Generates an alert if the RecoverPoint Time Lag

Replication Lag High replication time or data lag Replication Lag Warning
is above a specified High threshold
warning or critical level. Time Lag Critical
Threshold
Data Lag
Warning
threshold
Data Lag Critical
Threshold

TSM Database Generates an alert if the TSM Database Maximum

Utilization Higher TSM Database utilization Utilization High Database
than 90% exceeds 90%. Utilization -
90%

TSM Expiration Generates an alert if the Expiration Percentage

Process Duration TSM Expiration process take Process Increase - 25%
Exceeds Expectation longer than an hour to run, Duration Period - 7
or more than 25% longer Exceeds Max Duration -
that the average expiration Expectation 1
process time over the last
seven days.

TSM Recovery Log Generates an alert if the TSM Recovery Maximum

Utilization Higher TSM Database utilization Log Utilization Recovery Log
than 90% exceeds 90% High Utilization -
90%

Service Level Agreements

Service Level Agreement (SLA) analysis policies generate alerts about SLA violations.
Table 36 on page 118 describes the SLA jobs.

Table 36 Service Level Agreement

Name Description Rule Parameters

Backup Succeeded Generates an alert if a Backup succeed but N/A

But Did Not Meet SLA backup was successful failed SLA
Requirements but outside of its requirements
backup window.

118 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Status
Status category analysis policies generate alerts when there is concern of the current
status of a monitored device or application match. Table 37 on page 119 describes status
jobs.

Table 37 Status (page 1 of 2)

Name Description Rule Parameters

Backup Server Generates an alert if a Backup server errors N/A

Errors backup server error is
logged (TSM only).

CPU Offline Generates an alert if a CPU CPU offline N/A

is offline.

Agent Heartbeat Generates an alert if an Agent heartbeat N/A

Failed agent fails to send in its failed
heartbeat.

Agent Log File Alerts on any message that Agent Log Messages N/A
Message appears in the agent log
files.

Disk Failed Generates an alert if a disk Disk failed N/A

has failed.

EDL Failover Generates an alert if one EDL Failover N/A

occurred EDL appliance fails over to Occurred
another.

Fan Inactive Generates an alert if a fan Fan inactive N/A

on a device is inactive.

Fibre Channel Port Generates an alert if a Fibre Fibre Channel port N/A
Changed State Channel port has changed changed state
state.

Less than 75% of Generates an alert if less Less than x% of Lowest

Backup Devices than 75% of the backup backup devices backup
Available devices on a backup server available device
are Up. availability -
defaults to
75%

More Than 3 Generates an alert if there Many backup Maximum

Backup Devices are more than 3 backup devices unavailable number of
Unavailable devices on a backup server downed
Down. devices - 3

Network Interface Generates an alert if Network interface N/A

Changed State network interface gets a changed state
link up or link down event.

Object Restarted Generates an alert if a host Object restarted N/A

has been rebooted.

Object Status not Generates an alert if a Object Status not Up N/A

Up object’s status changes to
anything except active.

PSU Inactive Generate an alert if a Power PSU inactive N/A

Supply Unit is not active.

Policies and rules 119

Environment discovery in DPA

Table 37 Status (page 2 of 2)

Name Description Rule Parameters

Publisher Hung Generates an alert if the Publisher Queue N/A

Publisher queue hasn’t Hung
changed since the last poll.

Server Log File Alerts on any messages Server Log N/A

Message appearing in server log Messages
files.

Tape Drive Needs Generates an alert if a tape Tape drive needs N/A
Cleaning drive needs cleaning. cleaning

Tape Drive Not Generates an alert if a tape Tape drive not okay N/A
Okay drive is reporting a status
other than OK.

Tape Library Not Generates an alert if a tape Tape library not okay N/A
Okay library is reporting a status
other than OK.

Thermometer Generates an alert if a Thermometer N/A

Inactive thermometer becomes Inactive
inactive.

Thermometer Generates an alert if a Thermometer N/A

Overheating thermometer on a device overheating
indicates that it is
overheating.

Waiting For Generates an alert if a Waiting for writable Maximum

Writable Tapes For backup server has been devices outstanding
More Than 30 waiting more than 30 devices -
Minutes minutes for a writable tape. defaults to 0
Minutes
before
alerting -
defaults to 30
minutes

Xsigo Fan Less Generates an alert if the Xsigo Fan Speed Percentage to
Than 90% of speed of a fan on a Xsigo Less than Expected Check -
Normal Speed Director falls below 90% of defaults to
the normal speed. 90%.

120 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Troubleshooting
The troubleshooting analysis policies provide help for troubleshooting problems with the
environment. Table 38 on page 121 describes these jobs.

Table 38 Troubleshooting (page 1 of 2)

Name Description Rule Parameters

Backup Failed Due to Generate an alert if a Backup failed due N/A

Client Network Errors backup failed on a client to client network
while it experienced an errors
increase in network
errors.

Backup Failed Due To Generate an alert if a Backup failed due Maximum

High Client CPU backup failed on a to high client CPU processor
Utilization client, while the CPU utilization utilization -
utilization on the defaults to
computer was greater 90%
than 90%.

Backup Failed Due To Generates an alert if a Backup failed due Maximum

High Client Memory backup failed on a client to high client memory
Utilization whilst the memory memory utilization utilization -
utilization on that client defaults to 90
was greater than 90%.

Backup Failed Due To Generates an alert if a Backup failed due Maximum

High Server CPU backup failed on a client to high server CPU processor
Utilization whilst the CPU utilization utilization -
utilization on the defaults to
backup server was 90%
greater than 90%.

Backup Failed Due to Generates an alert if a Backup failed due Maximum

High Server Memory backup fails whilst the to high server memory
Utilization memory utilization on memory utilization utilization -
the backup server is defaults to
greater than 90%. 90%

Backup Failed Due to Generates an alert if a Backup failed due N/A

Server Network Errors backup failed while to server network
there was an increase in errors
the number of network
errors on the backup
server.

Disk Failed For More Generates an alert if a Disk failed for a Maximum
Than 48 Hours disk is in a failed state number of hours failure time -
for more than 48 hours. defaults to 48
hours

Fibre Channel Port Generates an alert if a Fibre Channel port N/A

Reporting Errors Fibre Channel port is reporting errors
reporting errors.

Fibre Channel Port Generates an alert if Fibre Channel port Maximum

Reporting Errors On more than 1% of all reporting more than percentage
More Than 1% frames going through a x% errors errors -
Frames Fibre Channel port have defaults to 1%
errors.

Policies and rules 121

Environment discovery in DPA

Table 38 Troubleshooting (page 2 of 2)

Name Description Rule Parameters

Network Interface Generates an alert if Network interface N/A

Reporting Errors errors are being seen on reporting errors
a network interface.

Network Interface Generates an alert if Network interface Maximum

Reporting Errors On more than 1% of the reporting more than percentage
More Than 1% packets travelling x% errors errors -
Packets through a network defaults to 1%
interface have errors.

Tape Drive Reporting Generates an alert if Tape drive reporting Include

Errors there is an increase in errors. Recoverable
the number of errors Errors -
seen on a tape drive. defaults to
False

Protection policies
Protection policies are used to define service level agreements and exposure reporting to
calculate whether a backup ran in its backup window and to calculate whether an
application or host is meeting its recovery time objective (RTO) and recovery point
objective (RPO). Protection policies also determine how an application, host, or device
should be replicated or backed up. Policies are assigned to objects and consist of a set of
rules that dictate:
◆ For replication: the type of copy, the replication level, and the schedule.
◆ For backups: the level of backup and the schedule.
DPA reports then compare the protection policy for an object to the actual replication or
backup taking place to display the level of compliance with policy.

Recoverability checks
Recoverability checks are additional consistency checks that are performed by DPA on an
environment, if recoverability analysis is configured. A recoverability check verifies that
the storage and recoverability environment is configured to a user’s particular
requirement (for example, Disaster Recovery).
If enabled and an inconsistency is detected, a recoverability check generates an exposure
just like an exposure generated by a Protection Policy breach or a Recoverability request.
Recoverability check exposures are displayed in the Replication Analysis area and the
Exposure reports.
There are three system recoverability checks that identify gaps, as described in Table 39
on page 122.

Table 39 Recoverability checks (page 1 of 2)

Recoverability check Description

Consistency Group Check Checks whether the devices of the recovery point are
configured in the same consistency group and the consistency
group is enabled. If no consistency group exists, a consistency
violation gap is generated for the recovery point.

122 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Table 39 Recoverability checks (page 2 of 2)

Recoverability check Description

Consistent Device Replication Checks whether the consistency option was used when the
Check images were created, when applicable. This is a best practice
check. If the consistency option was not used, then a
Consistency Violation gap is generated for the recovery point.

DR Host Visibility Check Checks whether the devices of a recovery point are mapped,
masked, and visible to the Disaster Recovery host. Otherwise, a
Consistency Violation gap is generated.

Chargeback policies
Chargeback reports provide the ability to perform a financial cost analysis for backups,
restores, and data protection replication operations in a customer’s environment. A cost is
calculated for each backup client and can be charged back to the business unit that is
responsible for that client or set of clients.
DPA calculates chargeback using two models: one for data backup and restore, and one
for the protection and replication of storage data by EMC RecoverPoint. Chargeback is
calculated for clients based on the inputs for each type.

Backup chargeback
Backup chargeback is broken out by cost per GB backed up and other backup costs.
Cost Per GB Backed Up uses the following inputs:
◆ Base Size - Baseline backup size in GB for base costing.
◆ Base Cost - Total cost for backup up to the base size specified.
◆ Cost of Each Additional GB - Additional cost per GB for backups greater than the base
size
Other Backup Costs are derived from the Chargeback Policy and use the following inputs:
◆ Cost Per Backup - the cost per backup (derived from the chargeback policy).
◆ Cost per GB Retained - the cost per gigabyte stored (derived from the chargeback
policy).
◆ Cost Per Restores - the cost per restore (derived from the chargeback policy).
◆ Cost per GB Restored - the cost per gigabyte restored (derived from the chargeback
policy).
◆ Cost Per Tape - the cost per tape used for backup (derived from the chargeback policy).

Storage chargeback
Storage chargeback is broken out by cost per GB stored, cost per GB replicated, and
snaps.
Cost Per GB Stored uses the following inputs:
◆ Cost Based On - chargeback calculated on either storage used or storage allocated.
◆ Base Size - Amount of base storage space allocated in GB
◆ Base Cost - A one-off price for the base size

Policies and rules 123

Environment discovery in DPA

◆ Cost of Each Additional GB - the price per GB after base size is exceeded
Cost Per GB Replicated uses the following inputs:
◆ Base Size - Amount of base storage space allocated in GB
◆ Base Cost - A one-off price for the base size
◆ Cost of Each Additional GB - the price per GB after base size is exceeded
Snaps uses the following inputs:
◆ Cost Per GB - the price per GB
A Chargeback Policy allows you to specify a value for each of these parameters. The total
cost for a client is calculated by adding each of the different cost elements. For example, if
you want to implement a chargeback model where you charge $5 for each backup that
took place and $0.20 for each GB that was backed up, then you can specify values for
these fields in the chargeback policy but not specify values for the other parameters.
Backup client objects are assigned a cost center, which allows Chargeback costs to be
calculated by cost center. A default cost center exists for objects that have not been
assigned a cost center.
Multiple chargeback policies can be created and different clients or groups of clients can
have different policies assigned to them. For example, if you wanted to calculate the
chargeback cost for one group of backup clients based on the number of backups
performed and another group based on the number of tapes used during the backup
process, two chargeback policies can be created and associated with each group of
clients.

Generating events from policies

When an analysis policy finds a matching condition, DPA generates an event. All events
are automatically logged in to the DPA Datastore server and can be viewed in the Advisor
section of the web console.
You can edit policies to:
◆ generate an email
◆ run a script
◆ send an SNMP trap
◆ write an event to a Windows Event Log
To edit actions for all the rules in the policy, go to Policies > Analysis Policies > Edit > Edit
Policy-based actions.
To edit actions on a per-rule basis:
1. Go to Policies > Analysis Policies > [select a policy] and click Edit.
2. Under Analysis Rules, highlight the rule name to edit, and click Edit Actions.
3. In the Edit Actions window, ensure that the Rule-based actions radio button is
selected.
Alternatively, edit all the rules in a policy or on a per-rule basis from the Inventory area.
This is applicable only to the roles that have permissions to edit the policy.

124 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

1. Go to Inventory and select the object.

2. Select Properties.
3. Within the object Details window, click the Policies tab.
4. Click Edit Override Settings. Edit Override Settings is available only if the role has
privileges to do so. Otherwise, the option is View Settings.
5. Within the object Override Policy Settings window, make applicable changes, either
on a per-rule level or at a policy level; and click OK when finished making changes.
The DPA online help provides more information on editing policy-based actions.
To monitor an object or group, you must assign an policy to an object in the inventory.
Policies that are assigned to groups are automatically inherited by the objects that are
children of that group.

Parameters for generating alerts from scripts

Scripts can be located in any directory. It might be useful to use the
<install-dir>/services/shared/ directory because in a clustered environment, you will need
to put the scripts just once. If you choose a different location, in a clustered environment
you will need manually to copy the scripts to every DPA Application server.
Table 40 on page 125 describes the parameters to the script to use to perform actions.

Table 40 Script field parameters

Parameter Description

Node Name of the node to which the alert applies.

Text Textual error message as defined in the ruleset.

Severity Severity of the alert.

Name Name of the analysis that triggered this alert.

Alert ID/Event ID ID that uniquely describes this alert.

First occurrence Timestamp that details the time that this alert first occurred.

Last occurrence Timestamp that details the time that this alert last occurred.

Count Number of times this alert has been issued.

Category Category of the analysis job that ran.

Description Alert description.

Policies and rules 125

Environment discovery in DPA

Table 41 on page 126 describes the arguments that are passed to a script in an alert
action.

Table 41 Script alert arguments

Argument Description

$0 Event node.

$1 Event message.

$2 Event severity (as set in the analysis properties).

$3 Name of analysis that caused the event.

$4 Alert ID (unique for this run of the script).

$5 Event ID (unique for this alert).

$6 First occurrence (timestamp).

$7 Last occurrence (always same as first occurrence).

$8 Count (always 1).

$9 Source view (the view to which the analysis was assigned).

$10 Source node (the node to which the analysis was assigned).

$11 Ruleset category.

Rule Template
A rule is the set of instructions used by the DPA Analysis Engine to determine whether a
condition has been met and if an alert is generated. For example, the file system filling up
rule contains the set of rules to determine if any file systems will exceed the threshold at a
certain point in the future.
A rule is used by an Analysis job to perform analysis and alerting based on information
within the DPA database. When DPA is installed, a number of pre-defined rules are
installed that can monitor for common problems that might occur in the environment. You
can use these rules as the basis for implementing an analysis policy. DPA provides a rules
editor that you can use to create entirely new rules.
The term rule template is used to differentiate the rule definition from the rule instance.
The rule template defines the rule's logic. When a rule template is added to an analysis
policy, it becomes a rule instance (or a rule) that the Analysis Engine will run. Also, when
rule templates are added to a policy, users can specify the values for any parameters. This
allows rules to be reused by different policies.
For example A Tier 1 policy might generate an alert when disk space is 80% utilized, while a Tier 2
policy can generate an alert when disk space is 90% utilized. This can be handled with the
same rule template that uses a parameter for utilization.

126 DPA 6.1 SP2 Installation and Administration Guide

 Environment discovery in DPA

Applying a policy
Policies can be applied directly to a group or an object. Policies applied directly to an
object always take precedence. When a policy is set at the group level, objects in the
group that do not have their own policies inherit the group's policy. The best practice is to
apply the policy at the highest group level. Policies cannot be applied to Smart Groups.
If an object is moved from one group to another group, the most recently applied policy is
implemented. For example, if an object is moved from Group A to Group B, the object
inherits the policy of Group B.
An administrator or any user with the Edit Node privileges can apply a policy to a group or
object.

Policies and rules 127

Environment discovery in DPA

128 DPA 6.1 SP2 Installation and Administration Guide

 Uninstallation

CHAPTER 4
Uninstallation

This chapter includes the following sections:

◆ Uninstalling the software ...................................................................................... 130

129
Uninstallation

Uninstalling the software

This section describes how to uninstall DPA in both UNIX/Linux and Windows
environments. Change to the install directory (_uninstall) by running the following
command:
cd <DPA_install_directory>/_uninstall

1. Initiate the uninstall process by running the following command:

Uninstall_Data_Protection_Advisor

When uninstalling the DPA Datastore, a warning indicating that the uninstaller will remove
the features that were installed during product installation appears indicating that the
database will be removed.

Silent uninstallations
◆ On UNIX/Linux machines, start a command shell, navigate to the _uninstall directory
and type the following command:
./Uninstall_Data_Protection_Advisor -i silent

◆ On Windows machines, type the following command through the command line:
Uninstall_Data_Protection_Advisor.exe -i silent

Uninstalling through user interface on Windows

1. Select Start > Control Panel > Programs and Features.
2. Uninstall EMC Data Protection Advisor from the list of installed applications.

Agent-only uninstallation
You cannot uninstall only the Agent from the DPA Application server or Datastore server
installation.
If you would like to upgrade the DPA Agent, upgrade the Agent only on the existing DPA
Application server or Datastore server installation. “” on page 24 provides information on
carrying out upgrades.

130 DPA 6.1 SP2 Installation and Administration Guide

 Troubleshooting

CHAPTER 5
Troubleshooting

This chapter includes the following sections:

◆ Troubleshooting installation.................................................................................. 132
◆ Viewing log files.................................................................................................... 133
◆ Troubleshooting data collection ............................................................................ 135
◆ Troubleshooting client discovery for replication analysis ....................................... 137
◆ Troubleshooting report output failure .................................................................... 141
◆ Troubleshooting report generation or publishing problems.................................... 141

131
Troubleshooting

Troubleshooting installation
DPA installation failure on Solaris
DPA fails to install on Solaris SPARC 64-bit with the following error message:
The included VM could not be unarchived (TAR). Please try to
download the installer again and make sure that you download using
’binary' modecopy. Please do not attempt to install this currently
downloaded copy.

Perform this procedure for installation of Solaris binaries for the DPA Agent, Server and
Migration Tools
Ensure that gnu tar is installed and available in the PATH ahead of the native Solaris
version before running the installer.
For example, through the command line,
1. cd /usr/sfw/bin/
2. ln -s gtar tar
If you are using Bourne shell, type,
export PATH=/usr/sfw/bin:${PATH}

IMPORTANT
The gtar version must be available on the Solaris host before executing the DPA 6.1
installer. Else, it will result in errors in the DPA 6.1 installation binaries.

This pertains to issue number DPA-21289.

DPA Datastore server on Linux or Solaris failure to start after installation

In certain circumstances the Kernel settings of the system running the DPA Datastore
server might need to be tuned for the Datastore server to start up correctly.
If the Datastore server fails to start and errors in the DPA log file reference shared memory
segments, then the values specified in the following file might need to be tuned according
to your system specifications.
◆ Linux: Investigate tuning values for SHMMAX and SHMMIN in the /etc/sysctl.conf
◆ Solaris: Investigate tuning values for shmsys:shminfo_shmmax and
shmsys:shminfo_shmmin in the /etc/system file

DPA web console start failure on Windows Server 2012

If the DPA web console fails to start on Windows Server 2012, check the following items:
◆ The Internet Explorer Enhanced Security Configuration (IE ESC) stops the DPA web
console from starting. Do not stop the notification of the block by clearing the
Continue to prompt when website content is blocked option because DPA never
comes past Starting services. Please wait.
The workaround for this is to disable the IE ESC.

132 DPA 6.1 SP2 Installation and Administration Guide

 Troubleshooting

◆ Internet Explorer in Windows server 2012 doesn't support Flash. The workaround for
this is to enable Desktop Experience in Windows server 2012.

Postinstallation memory tuning

[If you changed the memory settings to non-optimal values, run dpa ds tune command to
override any of the settings and set the optimal values. The value set to the dpa ds tune
command should be the available memory for the machine. So for a 8GB system, run:
dpa ds tune 8GB

Error messages during upgrades

If there is an error during the upgrade process, the DPA server stops. This could occur
under the following circumstances:
◆ Errors in SQL upgrade scripts
• Result: The server stops and does not continue.
• Suggested action: Contact EMC Technical Support.
◆ Errors in system metadata upgrade; for example, system reports, rule templates
• Result: The server stops, but you have the option to continue the upgrade.
• Suggested action: You can disregard this message and continue with the DPA
server upgrade. However, the DPA system might be unstable. If you do stop the
server upgrade, Contact EMC Technical Support
◆ Errors in the custom data upgrade; for example, custom analysis rules
• Result: An error message is thrown indicating the problem.
• Suggested action: You can disregard this message and continue with the DPA
server upgrade. However, you should expect the custom rule that failed to upgrade
not to work. An error is recorded in the log file.

Viewing log files

Log files provide important information when troubleshooting problems.

Note: The following section describes the log file locations for a standard DPA installation.
If the default installation directory was changed during installation, the location of the log
directory will be different.

By default, logs informational, warnings, and error messages. These might not provide
enough information when troubleshooting complex problems. To change the log detail
level, select Admin > System > Configure System Settings.

Viewing install log file

The Data_Protection_Advisor_Install_[two-digit date]_[two-digit month]__[year]_[two-digit
hour]_[two-digit minute]_[two-digit seconds].log file is generated during installation and
contains all log messages. For successful installations, this file can be found in the install

Viewing log files 133

Troubleshooting

directory (for example, /opt/emc/dpa/_install). For unsuccessful installations on UNIX

platforms, the file can be found in the root of the system drive. On Windows platforms, the
file can be found on the desktop.

Viewing server log files

The server log files are generated in the following locations:
◆ UNIX: /opt/emc/dpa/services/logs
◆ Windows: C:\Program Files\EMC\Data Protection Advisor\services\logs

Viewing agent log files

The agent log files are generated in the following locations:
◆ UNIX: /opt/emc/dpa/agent/logs
◆ Windows: C:\Program Files\EMC\Data Protection Advisor\agent\log\agent.log

Managing log files

When a log file reaches its maximum size, and the maximum number of log files exist in
the log file directory, the oldest log file for that process is deleted and a new log file
created. You can modify the maximum log file size and maximum number of log files. You
can also change the location of log files, if required.

Enabling alternative log rotation on VMs running Windows

There is a known issue on VMs running Windows that causes the logs not to rotate due to
the file being locked. To fix this, enable the alternative log rotation method. This will
change the way the logs are being used, where the highest numbered log is the latest and
not the agent.log file. This is pertaining to DPA-24288.
1. Create the following string registry:
HKLM\SOFTWARE\EMC\DPA\Agent\ALTLOGROTATE

2. Set the value to true.

Erroneous memory data in installer log file

The Free Memory and Total Memory data indicated at the top of the installation log files is
erroneous. The correct Free Memory and Total Memory data is located further down in the
log file, under STDERR ENTRIES.
The Corrected Total Memory data indicated under Executing IAUpdatePostgesconfFile:
[INFO] refers to data being used for the DPA Datastore service.

Running a DPA Agent request in debug mode using DPA web console
The DPA Agent request in debug mode, also sometimes called a modtest, is a support
tool. If you are encountering problems with a Data Collection Defaults, an EMC Technical
Support Engineer might ask you to run the Agent request debug mode from the DPA web
console. You can run DPA Agent request in debug mode, download the zip file directly from

134 DPA 6.1 SP2 Installation and Administration Guide

 Troubleshooting

the DPA web console with no need of going to DPA Server to retrieve the zip file, and send
the zip file for analysis. The Agent request debug mode runs the selected request and
retrieves the output and the log messages, in debug log level, and stores that report xml
as a zip file to the following location: <DPA_HOME>\services\shared\modtests, where
<DPA_HOME> is the location of the DPA installation.
Consider the following when running DPA Agent request in debug mode using DPA web
console:
◆ The test cannot be run if the Collection Request is disabled.
◆ The test cannot be run if the Collection Request isn’t applicable on the object.
◆ If you are running Google Chrome: you should change the default security setting for
the URL to low:
Go to Trusted Sites, add the URL to Trusted Sites list, and set security to low.
To run a DPA Agent request in debug mode using DPA web console:
1. In the web console, select Inventory > Object Library.
2. In the Object Library, select the DPA server under All hosts.
3. In the host details window, select the Data Collection tab.
4. In Data Collection, select the Request.
5. Right-click Run and select Run in Debug.
6. Click Close to the a dialog box that appears confirming that the test is running.
7. Click History to view collected tests. The rows highlighted in orange indicate results
from a DPA Agent request in debug mode.
8. Click the test result. If a Windows Security Login appears, enter your DPA server
credentials and click OK.
9. Access the saved test result zip file in the default browser folder.
Alternatively, to access the successfully collected tests, go to
<DPA_HOME>\services\shared\modtests.

Default modtest deletion schedule

Deletion of modtest files from the DPA server is executed weekly on Sunday at 4:00 a.m.
All test results files older than seven days are removed. This schedule is not configurable.

Troubleshooting data collection

This section describes the steps that can be taken to diagnose problems when trying to
gather data. The following scenario is assumed:
◆ DPA was successfully installed.
◆ The Discovery Wizard was successfully run to create the object to monitor.
◆ Requests have been assigned to the object and the agent has been reloaded.
◆ Sufficient time (15 minutes) has passed to allow the agent to gather data.

Troubleshooting data collection 135

Troubleshooting

◆ An appropriate report has been run that returns no data when data should exist for the
object.

Data collection troubleshooting first actions

1. Verify that the time period selected for the report is correct.
2. Check that the correct requests have been assigned to the object.
Select Inventory > Object Library > [select host] >Data Collection. Select the policy that
contains the object > Edit to ensure that the request is enabled.
3. Rerun the report.
Review any errors returned by the Agent Errors report and take corrective action if
possible (for example, an authentication problem that needs to be resolved).

Data collection troubleshooting second actions

If the first round of steps do not resolve the problem, perform the following:
1. If no resolvable agent errors are reported, select Admin > System, click Configure
System Settings, and verify the data collection agent settings.
2. If the status shows that the agent is active, verify that the process is active on the
operating system on which the agent is installed.
3. Run the Agent log reports in the web console followed by the Agent Status, and then
the Data Collection History report.
4. Rerun the report. If the report continues to show no data, open the agent log and look
for any problems. For example, was an incorrect value entered during agent
installation. "Viewing log files" on page 133 describes how to view the log files.

Preparing a log file for submission to EMC Support

When submitting a log file to EMC Support, steps can be taken to ensure that the log file
contains the relevant information and unnecessary log entries are minimized.
To prepare a clean log file:
1. Set the Log Level of the process to Debug in System Settings, as described in "Viewing
log files" on page 133.
2. Stop the agent process.
3. Navigate to the directory in which the log file is stored. Rename or remove all existing
log files for the process.
4. Restart the process.
Restarting an agent reloads all the requests assigned to that agent and starts the data
gathering routine. This ensures that all requests have been attempted. Starting a new
log file removes the need to search through unnecessarily long log files for a problem.
5. Select Inventory > Object Library > [select host] >Data Collection and then select
History. Alternatively,run an Agent History report.
6. Rerun the report to confirm data is not being gathered.

136 DPA 6.1 SP2 Installation and Administration Guide

 Troubleshooting

7. Set the Log Level of the process back to Info in System Settings.
8. Make a copy of the log for submission to EMC Support.

Troubleshooting client discovery for replication analysis

This section describes the steps that can be taken to diagnose problems when trying to
configure VNX/CLARiiON or Symmetrix storage arrays for replication analysis. The
following scenario is assumed:
◆ DPA was successfully installed.
◆ The DPA server and the storage array host meet the requirements as specified in the
EMC Data Protection Advisor Software Compatibility Guide.
◆ EMC Solutions Enabler has been successfully installed.

Client discovery using remote execution

Table 1 on page 137 describes possible problems and solutions encountered when
attempting to discover clients remotely (without a DPA agent).

Table 1 Client discovery problems and solutions (page 1 of 2)

Problem Solutions

Client discovery fails: No authentication defined, or • Create a credential in DPA (Admin > System > Manage
unable to log in. Credentials) and assign it to the client.
• Check that the username and password supplied with the
credential is able to connect the client.
• Ensure su or sudo are not needed to connect, and if so that the
correct parameters are supplied in the credential.

Client Discovery fails: Failed to connect to client using • Check that the username and password supplied with the
RPC, or a specified logon session does not exist. credential is able to connect the client.
• Ensure you supplied the username along with the domain
name:
<domain>\<username > for remote computers, <computer
name>\<username> for local computers. In most cases
localhost\<username> can be used.
• Check if the host is accessible from the DPA server using the
admin share: \\hostname\Admin$
• If the error exists after you tried all the preceding actions,
change the Log on as value for the DPA server service from local
system to any other user with administrator privileges. A local
administrator also can be set.

Client Discovery failed: Failed to connect to client using • Check if the client's Name, IP, or Alias is defined correctly and
RPC. The network path was not found. reachable from the DPA server.
• Check if the host is accessible from the DPA server using the
admin share: \\hostname\Admin$
If the share is not accessible, check that it is not blocked by a
firewall.

Troubleshooting client discovery for replication analysis 137

Troubleshooting

Table 1 Client discovery problems and solutions (page 2 of 2)

Problem Solutions

Client Discovery failed: User does not have enough • Follow the system requirements for remote execution
privileges to get the device mapping information. permissions.
• Assign a user with administrator capabilities to the credential.
• Check that the user that connects to the client has write and
execute privileges to the path /var/tmp. (UNIX)

Client Discovery failed: Failed to send file for discovery Check for free disk space in /var/tmp.
to client using SCP
or
Failed to send file for discovery to client using FTP.

If the DPA Server is installed on Solaris host and you Ensure the DPA server is resolvable from the Windows Proxy host.
would like to run a request on a Windows host without The DPA server name might need to be added to the hosts file.
installing a local agent, you must run the request
trough a proxy (a Windows host that has an agent
installed on it).
If the proxy host does not resolve the DPA server name
correctly (DNS issues), the following error will appear in
the request history log:
"Error Running Request: No Data"
(The following error will also appear in the Agent Log:
Detail: get host by name failed in tcp_connect() )

Client Discovery fails: Error (977). Overlapped IO Verify that there is no antivirus software installed on the host.
operation in progress. Antivirus software might be blocking the irxsvs.exe operation.
Disable antivirus blocking by authorizing the irxsvs.exe file in the
anti-virus software.

Client discovery fails with the following error: Ensure that the following services are running: Server, Computer
<client_name> irx errMsg: Unable to connect Browser, and Workstation.
host:<client_name> with user:<domain>\<username>
using RPC irx output: Error (1203): No network provider
accepted the given network path.

When using sudo, the Host Config request might fail to This occurs only when the credentials are configured to use sudo.
return volume group information on AIX hosts with the Add the following line to the sudoers file: Defaults env_keep +=
following message: "ODMDIR"
SymMapVgShow exited with code 161
(SYMAPI_C_VG_NOT_AVAILABLE) SessionId: 0 - for
VG:<vg_name> with type: 2(AIX LVM) VolumeGroup
information will not be parsed.

Client discovery with agent

Table 2 on page 138 describes possible problems and solutions encountered when
attempting to discover clients using the DPA agent.

Table 2 Client discovery with agent problems and solutions

Problem Solution

Client Discovery request is using remote execution 1. Ensure that the agent is installed on the host.
instead of using the installed agent. 2. Ensure that the DPA server is defined as the controller for the
agent.
3. Restart the agent service.

138 DPA 6.1 SP2 Installation and Administration Guide

 Troubleshooting

General client discovery

Table 3 on page 139 describes general possible problems and solutions encountered
when attempting to discover clients from DPA.

Table 3 General client discovery problems and solutions (page 1 of 2)

Problem Solution

Client discovery finished with warnings: • Check if the application is running and available for
Failed to discover application storage objects for connection.
application <application_name> on client • Check if the user configured in the DPA credential has enough
<client_name>. privileges to query the application's system data.

Client discovery failed: can’t connect to any IP. • Check that port 25011 from the host to the DPA server or port
135 from the Windows Proxy Collector is not blocked by a
firewall.

Client discovery finished with warnings: 1. Select Admin > System > Manage Credentials.
Home directory was not found for application. 2. Click Edit to edit the Credential.

Unsupported File System Type encountered: This file system type is not supported by DPA.
<filesystem_name>. To avoid this warning in the next client discovery, you can ignore
discovery for this file system.

Note: DPA will not display recovery data for this file system.

Client discovery failed with the error: Ensure that there is enough disk space in the host's root file
Verify that you have enough disk space and write system according to the system requirements.
permission.
or
Failed to unpack file on client <client_name>.

Client discovery finished with warnings: • Check if <home_dir> that was discovered exists on the client.
Can't find or no permission to execute file <home_dir>. • Check if the file sqlplus in the home directory that was
discovered has enough permissions for being executed by
DPA.

Client discovery finished with error: 1. In DPA, select Admin > System > Configure System Settings.
Timeout waiting for agent response on client 2. Select Server.
<client_name>.
3. Change the parameter Timeout(s) from the default 120 to a
larger value.
or
Check if the DPA server has two network cards enabled and the
client is able to reach both of them. If the client is not able to
reach one of the cards, disable this network card.

When connecting to ECC 6.1, client discovery finished 1. Run a batch file containing the following command:
with error: %ECC_INSTALL_ROOT%\tools\JRE\Nt\latest\bin
Error Import Clients for w2k3-96-52.dm1nprlab.com \java -cp
finished with errors. %ECC_INSTALL_ROOT%\ECCAPIServer\class;%ECC
Check previous error messages for further information. _INSTALL_ROOT%\ECCAPIServer\ecc_inf\exec\e
Unable to logon (Connection refused). ccapiclient.jar;
com.emc.ecc.eccapi.client.util.EccApiPopul
ateRandomPassword ApiClient

The additional classpath parameter is only required if the

command is not used from within the ECC classes directory.

Troubleshooting client discovery for replication analysis 139

Troubleshooting

Table 3 General client discovery problems and solutions (page 2 of 2)

Problem Solution

SYMAPI and CLARAPI discovery requests fail when run 1. Run the discovery requests separately. Make sure that one
together for the first time on a Solaris DPA Server. CLARAPI or SYMAPI discovery is successful before running the
other.
2. After successful requests have been run sequentially first,
they can be run together.

Client discovery is not correctly resolving LUNs on Configure the DNS on the ESX properly or add the VM name and
VFMS on ESX 4.1. IP to the ESX hosts file.
When trying to correlate the virtual devices to the
remote storage they reside on, and the hosting ESX of
the virtual machine cannot resolve the name of the VM
(DNS configuration), the correlation fails and the
virtual devices are displayed as local devices.

Importing CLARiiON information request fails with the Run the SYMCLI command on the SE host for this CLARiiON:
following error message: symcfg sync -clar
"An error occurred while data was being loaded from a
Clariion ClarEventGet exited with code 3593
(SYMAPI_C_CLARIION_LOAD_ERROR) "

Synchronizing Incorrect Recovery Point times

If there is a time difference between the DPA server and the storage array being monitored,
recovery points might be displayed with times that do not match the expected times. For
example, a system administrator initiates a recovery point at 2:00 but the recovery point is
displayed at 4:00 in DPA.
The discovery requests have a Time Offset option that accounts for time differences and
allows recovery points to be displayed with consistent times. You must calculate the exact
offset between the DPA server and the storage array host.
In the following instructions, connector refers to the DPA host which has the
SYMAPI/CLARAPI Engine Discovery request assigned.
The time offset is calculated in seconds.

VNX/CLARiiON
To calculate the time offset between the VNX/CLARiiON and the DPA server:
1. Query the VNX/CLARiiON's time using the navicli command getsptime.
2. At the same time, query the connector's time.
3. If the connector time and the DPA server time are equal (no Time Zone difference):
TimeOffset = Connector time - VNX/CLARiiON time.
4. Otherwise, if there is a time difference also between the connector time and the DPA
server time:
TimeOffset = (DPA server time - Connector time) - VNX/Clariion's time.
5. Set the time offset for the request. "Configuring the time offset" on page 141 provides
information on this.

140 DPA 6.1 SP2 Installation and Administration Guide

 Troubleshooting

Symmetrix
To calculate the time offset between the Symmetrix and the DPA server:
1. Query the DPA server time.
2. At the same time, query the connector's time.
3. If the Connector time and DPA server time are not equal:
TimeOffset = DPA server time - Connector Time.
4. Otherwise, no time offset for Symmetrix needs to be set.
5. Set the time offset for the request."Configuring the time offset" on page 141 provides
information on this.

Configuring the time offset

After you have calculated the time offset, set the value for the request. To set the time
offset value:
1. Select Inventory > Object Library > [select host] >Data Collection.
2. Select the policy containing the request and click Edit.
3. Select the time offset you calculated Client-server Time Difference (in seconds or
minutes).
4. Click Apply.

Troubleshooting report output failure

If reports are hanging after you save them with the message Please wait while
generating report,and you are using Internet Explorer, it could be because you do
not have the XMLHTTP option enabled. This is in relation to DCE-1546.
To enable the XMLHTTP option:
1. Go to Internet Options > Advanced
2. Scroll to Security and select Enable Native XMLHTTP Support, then click OK.

Troubleshooting report generation or publishing problems

If scheduled reports fail to generate, or if they generate properly but fail to publish,
perform the following actions:
◆ If a custom report, check that report template has been designed correctly in Report
Editor.
◆ Check that report template runs properly in Run Report Wizard.
◆ Check that report template properly saved (exported) in desired format.
◆ Check errors/warnings in server.log regarding scheduled reports.
If these actions do not resolve the issue, contact EMC Technical Support.

Troubleshooting report output failure 141

Troubleshooting

Synchronizing the system clock

As part of the User Authentication process, DPA relies on the system clock times on the
client machine and the server differing by less than one minute. In the event that clock
times are unsynchronized, the following error message is displayed:
User Authentication failed due to the times on the client and server not matching. Ensure
that the times are synchronized.
To resolve this issue, ensure that the system clock times on the client and server are
synchronized.
It is also advised that NTP be used to synchronize the DPA Server and all the DPA Agent
hosts as well. This is imperative for accurate data collection.

142 DPA 6.1 SP2 Installation and Administration Guide