hacmp aix

High Availability ClusterMulti-Processing XD (Extended Distance) for PPRC

HACMP/XD for Metro Mirror: Planning and Administration GuideVersion 5.4

SC23-4863-10

Tenth Edition (August 2006) Before using the information in this book, read the general information in Notices for HACMP/XD for Metro Mirror: Planning and Administration Guide.

This edition applies to HACMP/XD for AIX 5L v5.4, and to all subsequent releases of this product until otherwise indicated in new editions.

© Copyright International Business Machines Corporation 1998, 2006. All rights reserved.

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

Contents

About This Guide 9

Chapter 1: Overview of HACMP/XD for Metro Mirror 15Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15HACMP/XD for Metro Mirror Features . . . . . . . . . . . . . . . . . . . . . 16

High Availability and Disaster Recovery Support Features . . . . . . 16PPRC Mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Disk Subsystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17PPRC Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Optional Metro Mirror Configurations . . . . . . . . . . . . . . . . . . . . . . 19DSCLI Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19SVC PPRC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19eRCMF Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

HACMP/XD for Metro Mirror in an HACMP Cluster . . . . . . . . . 20PPRC Replicated Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20Resource Groups that Include PPRC Replicated Resources . . . . . . 21HACMP Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Fallover and Fallback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

Where You Go from Here . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Chapter 2: HACMP/XD for Metro Mirror General Planning 25HACMP/XD for Metro Mirror Management Types . . . . . . . . . . . 26

Coexistence of Two Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Planning Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Planning Worksheets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Planning HACMP Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Completing the HACMP/XD for Metro Mirror Site Worksheet . . 28Planning HACMP/XD for Metro Mirror Resource Groups . . . . . . 28Resource Group Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29Completing the PPRC Resource Group Worksheet . . . . . . . . . . . . 29

Where You Go from Here . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

Chapter 3: Installing HACMP/XD for Metro Mirror 31Installation Prerequisites for HACMP/XD Metro Mirror . . . . . . . 31

Next Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32Contents of the Installation Media . . . . . . . . . . . . . . . . . . . . . . . . . . 33

HACMP/XD for MetroMirror: Planning and Administration Guide 3

Contents

4

HACMP/XD for Metro Mirror Installation Choices . . . . . . . . . . . 35Installing from an Installation Server . . . . . . . . . . . . . . . . . . . . . . . 35Installing from a Hard Disk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35Installing the Software from the Installation Media . . . . . . . . . . . . 36

Upgrading to HACMP/XD 5.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Supported Upgrade Path to HACMP/XD 5.4 . . . . . . . . . . . . . . . . . 38Upgrade Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Getting Ready to Upgrade the HACMP/XD Software . . . . . . . . . . 39Upgrading the Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Verifying the Upgraded Cluster Definition . . . . . . . . . . . . . . . . . . . 39Recovering from a Failed Installation . . . . . . . . . . . . . . . . . . . . . . . 40Modifying Previous Cluster Snapshots . . . . . . . . . . . . . . . . . . . . . . 40

Addressing Problems during the Installation . . . . . . . . . . . . . . . . . 41Where You Go from Here . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

Chapter 4: HACMP/XD for Metro Mirror with Direct Management 43Planning for Direct Management . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

Planning Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44Sample Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44Planning Worksheets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45Planning Integration with PPRC . . . . . . . . . . . . . . . . . . . . . . . . . . . 46Planning for PPRC Replicated Resources . . . . . . . . . . . . . . . . . . . . 47Planning Tasks for the ESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Planning for ESS Disk Subsystems Used By Direct Management . 51Planning a PPRC Resource Group . . . . . . . . . . . . . . . . . . . . . . . . . 52Next Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

Installing HACMP/XD for Metro Mirror for Direct Management 52Installation Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52Software Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53Installing Filesets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54Addressing Problems during the Installation . . . . . . . . . . . . . . . . . 54

Configuring in a Direct Management Environment . . . . . . . . . . . . 55Configuration Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55Configuration Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55Configuring HACMP/XD for Metro Mirror for Direct Management Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56Configuring the HACMP/XD Metro Mirror Cluster for Direct Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61Improving the Performance of Volume Groups . . . . . . . . . . . . . . . 67Verifying and Synchronizing the PPRC Configuration . . . . . . . . . 68Configuring Resource Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Starting the Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69Changing PPRC Replicated Resources Configuration . . . . . . . . . . 69Changing the Configuration for Sites . . . . . . . . . . . . . . . . . . . . . . . 70Changing the Configuration for PPRC Replicated Resources . . . . 70

HACMP/XD for MetroMirror: Planning and Administration Guide

Contents

Chapter 5: HACMP/XD for Metro Mirror with DSCLI Management 71Overview of the DSCLI Management System . . . . . . . . . . . . . . . . 71Planning for DSCLI Management . . . . . . . . . . . . . . . . . . . . . . . . . . 72

Limitations and Restrictions for DSCLI Management . . . . . . . . . . 72Sample Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74Planning Worksheets for DSCLI Management . . . . . . . . . . . . . . . . 76Setting Up Volume Groups and File Systems on DSCLI-Protected Disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77Next Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

Installing the DSCLI Management for PPRC Filesets . . . . . . . . . 85Installing Prerequisite Software . . . . . . . . . . . . . . . . . . . . . . . . . . . 85Installing the DSCLI Management Filesets . . . . . . . . . . . . . . . . . . 86Installation Directories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86Upgrading to HACMP/XD for MetroMirror v 5.4 . . . . . . . . . . . . . 86

Configuring HACMP/XD for MetroMirror Using the DSCLI . . . 87DSCLI and HACMP Configuration Requirements . . . . . . . . . . . . . 87DSCLI Management Configuration Steps . . . . . . . . . . . . . . . . . . . 87Configuring DSCLI-Managed PPRC Replicated Resources . . . . . 87Configuring Resource Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

Verifying the DSCLI-Managed PPRC Configuration . . . . . . . . . . 92Synchronizing the Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94Starting the Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94Changing the DSCLI-Managed PPRC Replicated Resources Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

Changing the Configuration for Sites . . . . . . . . . . . . . . . . . . . . . . . 95Changing the Configuration of DSCLI-Managed PPRC Replicated Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

Chapter 6: HACMP/XD for Metro Mirror with SVC Management 97Overview of SVC Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

Definition of Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98Planning for SVC Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

Limitations and Restrictions for HACMP/XD: Metro Mirror for SVC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99Sample Configuration for SVC Management . . . . . . . . . . . . . . . . 100Planning PPRC-SVC Management Support . . . . . . . . . . . . . . . . . 103Setting up Volume Groups and Filesets on SVC PPRC-Protected Disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

Installing HACMP/XD: Metro Mirror for SVC . . . . . . . . . . . . . . 109Installing Prerequisite Software . . . . . . . . . . . . . . . . . . . . . . . . . . 109Installing HACMP/XD: Metro Mirror for SVC Filesets . . . . . . . 109

Configuring HACMP/XD: Metro Mirror for SVC . . . . . . . . . . . 109SVC and HACMP Configuration Requirements . . . . . . . . . . . . . 110Defining SVC Cluster Configurations . . . . . . . . . . . . . . . . . . . . . 110Defining SVC PPRC Relationships . . . . . . . . . . . . . . . . . . . . . . . 111


Contents

6

Defining SVC PPRC Resources . . . . . . . . . . . . . . . . . . . . . . . . . . 111Adding SVC PPRC Consistency Groups to HACMP Resource Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112Verifying and Synchronizing the HACMP/XD Cluster . . . . . . . . 112Starting the HACMP/XD SVC PPRC Cluster . . . . . . . . . . . . . . . 114

Changing an HACMP/XD for Metro Mirror SVC Configuration 114Changing an SVC Cluster Configuration . . . . . . . . . . . . . . . . . . . 115Removing an SVC Cluster Configuration . . . . . . . . . . . . . . . . . . . 115Changing SVC PPRC Relationships . . . . . . . . . . . . . . . . . . . . . . . 115Removing SVC PPRC Relationships . . . . . . . . . . . . . . . . . . . . . . 116Changing SVC PPRC Resources . . . . . . . . . . . . . . . . . . . . . . . . . 117Removing SVC PPRC Resources . . . . . . . . . . . . . . . . . . . . . . . . . 117Modifying SVC PPRC Consistency Groups in HACMP Resource Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

Troubleshooting HACMP/XD for Metro Mirror for SVC . . . . . 118SVC PPRC States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118Viewing SVC Clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Chapter 7: HACMP/XD for Metro Mirror with eRCMF Management 121Planning for eRCMF Management . . . . . . . . . . . . . . . . . . . . . . . . . 121

Planning Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121Limitations for eRCMF Support . . . . . . . . . . . . . . . . . . . . . . . . . . 121Sample Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122Planning Worksheets for eRCMF Environment . . . . . . . . . . . . . . 123Planning Integration with eRCMF . . . . . . . . . . . . . . . . . . . . . . . . 123Planning HACMP Configuration with eRCMF Management . . . 127Planning for Resource Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . 128Where You Go from Here . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

Installing eRCMF Management Filesets . . . . . . . . . . . . . . . . . . . . 129Software Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129eRCMF Management Filesets . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129Migrating from Direct Management to eRCMF Management Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130Where You Go from Here . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

Configuring eRCMF Management . . . . . . . . . . . . . . . . . . . . . . . . . 131Overview of Configuration for eRCMF management . . . . . . . . . 132Defining HACMP Sites for eRCMF Support . . . . . . . . . . . . . . . . 132Defining eRCMF Global Attributes . . . . . . . . . . . . . . . . . . . . . . . 134Defining eRCMF-Managed PPRC Replicated Resources . . . . . . 135Configuring Resource Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136Starting the Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136Changing the Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

Monitoring eRCMF Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138


Contents

Chapter 8: Maintaining and Troubleshooting HACMP/XD for Metro Mirror 139Logging Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139Maintaining HACMP/XD for Metro Mirror . . . . . . . . . . . . . . . . . 139

Using C-SPOC for Configuration Changes . . . . . . . . . . . . . . . . . 139Restarting ESS Copy Services Server on the Backup CSS . . . . . . 140Detecting ESS Failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140Using FlashCopy User Exit to Recover Data . . . . . . . . . . . . . . . . 140

Maintaining HACMP/XD for Metro Mirror with eRCMF Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

Security Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141Recovery Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

Troubleshooting DSCLI-Managed Clusters . . . . . . . . . . . . . . . . . 142Common cluster startup problems . . . . . . . . . . . . . . . . . . . . . . . . . 142Handy DSCLI Commands to Know . . . . . . . . . . . . . . . . . . . . . . . 143PPRC Instance States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143Handy AIX Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144

Appendix A: General HACMP Planning Worksheets 145

Appendix B: Planning Worksheets for Direct Management (ESS CLI) 151

Appendix C: Planning Worksheets for DSCLI Management 159

Appendix D: Planning Worksheets for SVC PPRC 169

Appendix E: Planning Worksheets for an eRCMF Environment 177

Glossary 187

Index 193


Contents

8

About This GuideThis guide provides information necessary to plan, install, configure, and maintain the High Availability Cluster Multi-Processing Extended Distance (HACMP/XD for Metro Mirror (previously known as Synchronous PPRC).

The following table provides version and manual part numbers for the HACMP/XD for Metro Mirror: Planning and Administration Guide.

Who Should Use This GuideThis guide is intended for system administrators and customer engineers responsible for planning hardware and software resources, installing, configuring, and maintaining HACMP/XD for Metro Mirror.

As a prerequisite for installing and configuring the HACMP/XD for Metro Mirror software, you should be familiar with:• HACMP™• IBM Enterprise Storage Server (ESS) (including disk devices, cabling, and network

interfaces)• IBM Enterprise Storage Server Copy Services and its Peer-to-Peer Remote Copy (PPRC)

component• Enterprise Remote Copy Management Facility (eRCMF)• The AIX 5L operating system, including the Logical Volume Manager subsystem• The System Management Interface Tool (SMIT)• Communications, including the TCP/IP subsystem.

HACMP Version Book Name Book Number

5.4 HACMP/XD for Metro Mirror: Planning and Administration Guide

SC23-4863-10

5.3: Last update 7/2006 HACMP/XD for Metro Mirror: Planning and Administration Guide

SC23-4863-09

5.3 HACMP/XD for Metro Mirror: Planning and Administration Guide

SC23-4863-07

5.2: Last update 10/2005 HACMP/XD for ESS PPRC: Planning and Administration Guide

SC23-4863-06

5.1 HACMP/XD for ESS PPRC: Planning and Administration Guide

SC23-4863-03

HACMP/XD for Metro Mirror: Planning and Administration Guide 9

10

HighlightingThis guide uses the following highlighting conventions:

ISO 9000ISO 9000 registered quality systems were used in the development and manufacturing of this product.

HACMP PublicationsThe HACMP software comes with the following publications:• HACMP for AIX 5L Release Notes in /usr/es/sbin/cluster/release_notes describe issues

relevant to HACMP on the AIX platform: latest hardware and software requirements, last-minute information on installation, product usage, and known issues.

• HACMP on Linux Release Notes in /usr/es/sbin/cluster/release_notes.linux/ describe issues relevant to HACMP on the Linux platform: latest hardware and software requirements, last-minute information on installation, product usage, and known issues.

• HACMP for AIX 5L: Administration Guide, SC23-4862• HACMP for AIX 5L: Concepts and Facilities Guide, SC23-4864• HACMP for AIX 5L: Installation Guide, SC23-5209• HACMP for AIX 5L: Master Glossary, SC23-4867• HACMP for AIX 5L: Planning Guide, SC23-4861• HACMP for AIX 5L: Programming Client Applications, SC23-4865• HACMP for AIX 5L: Troubleshooting Guide, SC23-5177• HACMP on Linux: Installation and Administration Guide, SC23-5211• HACMP for AIX 5L: Smart Assist Developer’s Guide, SC23-5210• IBM International Program License Agreement.

Italic Identifies new terms or concepts, or indicates emphasis.

Bold Identifies routines, commands, keywords, files, directories, menu items, and other items whose actual names are predefined by the system.

Monospace Identifies examples of specific data values, examples of text similar to what you might see displayed, examples of program code similar to what you might write as a programmer, messages from the system, or information that you should actually type.

HACMP/XD for Metro Mirror: Planning and Administration Guide

HACMP/XD PublicationsThe HACMP Extended Distance (HACMP/XD) software solutions for disaster recovery, added to the base HACMP software, enable a cluster to operate over extended distances at two sites. HACMP/XD publications include the following: • HACMP/XD for Geographic LVM (GLVM): Planning and Administration Guide,

SA23-1338• HACMP/XD for HAGEO Technology: Concepts and Facilities Guide, SC23-1922• HACMP/XD for HAGEO Technology: Planning and Administration Guide, SC23-1886• HACMP/XD for Metro Mirror: Planning and Administration Guide, SC23-4863.

HACMP Smart Assist Publications The HACMP Smart Assist software helps you quickly add an instance of certain applications to your HACMP configuration so that HACMP can manage their availability. The HACMP Smart Assist publications include the following: • HACMP Smart Assist for DB2 User’s Guide, SC23-5179• HACMP Smart Assist for Oracle User’s Guide, SC23-5178• HACMP Smart Assist for WebSphere User’s Guide, SC23-4877• HACMP for AIX 5L: Smart Assist Developer’s Guide, SC23-5210• HACMP Smart Assist Release Notes.

IBM AIX 5L PublicationsThe following publications offer more information about IBM technology related to or used by HACMP:• RS/6000 SP High Availability Infrastructure, SG24-4838• IBM AIX 5L v.5.3 Security Guide, SC23-4907• IBM Reliable Scalable Cluster Technology for AIX 5L and Linux: Group Services

Programming Guide and Reference, SA22-7888• IBM Reliable Scalable Cluster Technology for AIX 5L and Linux: Administration Guide,

SA22-7889• IBM Reliable Scalable Cluster Technology for AIX 5L: Technical Reference, SA22-7890• IBM Reliable Scalable Cluster Technology for AIX 5L: Messages, GA22-7891.

Accessing PublicationsUse the following Internet URLs to access online libraries of documentation:

AIX 5L, IBM eServer pSeries, and related products:

http://www.ibm.com/servers/aix/library

AIX 5L v.5.3 publications:

http://www.ibm.com/servers/eserver/pseries/library/


http://www.ibm.com/servers/aix/library/



http://www.ibm.com/servers/eserver/pseries/library/

12

WebSphere Application Server publications:

Search the IBM website to access the WebSphere Application Server Library

DB2 Universal Database Enterprise Server Edition publications:

http://www.ibm.com/cgi-bin/db2www/data/db2/udb/winos2unix/support/v8pubs.d2w/en_main#V8PDF

Tivoli Directory Server publications:

http://publib.boulder.ibm.com/tividd/td/IBMDirectoryServer5.1.html

IBM Welcomes Your CommentsYou can send any comments via e-mail to [email protected]. Make sure to include the following in your comment or note: • Title and order number of this book • Page number or topic related to your comment.

When you send information to IBM, you grant IBM a nonexclusive right to use or distribute the information in any way it believes appropriate without incurring any obligation to you.

TrademarksThe following terms are trademarks of International Business Machines Corporation in the United States or other countries:• AFS• AIX • AIX 5L• DFS•

• eServer Cluster 1600• Enterprise Storage Server• HACMP• IBM• NetView• pSeries• RS/6000 • Scalable POWERParallel Systems• Shark• SP • xSeries• WebSphere• Red Hat Enterprise Linux (RHEL)


http://www.ibm.com/cgi-bin/db2www/data/db2/udb/winos2unix/support/v8pubs.d2w/en_main#V8PDF

http://publib.boulder.ibm.com/tividd/td/IBMDirectoryServer5.1.html

• SUSE Linux Enterprise Server• RPM Package Manager for Linux and other Linux trademarks.

UNIX is a registered trademark in the United States and other countries and is licensed exclusively through The Open Group.

Linux is a registered trademark in the United States and other countries and is licensed exclusively through the GNU General Public License.

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


14

Chapter 1: Overview of HACMP/XD for Metro Mirror

High Availability Cluster Multi-Processing Extended Distance (HACMP/XD) for synchronous Peer-to-Peer Remote Copy ((PPRC), now known as and referred to as Metro Mirror increases data availability for IBM TotalStorage volumes that use to copy data to a remote site for disaster recovery purposes. HACMP/XD for Metro Mirror takes advantage of the following components to reduce downtime and recovery time during disaster recovery:• HACMP cluster management• PPRC fallover and fallback functions• Optional Components:

• DSCLI PPRC support• SAN Volume Controller (SVC) PPRC support• Enterprise Remote Copy Management Facility (eRCMF).

This chapter provides an overview of HACMP/XD for Metro Mirror. The chapter contains the following sections:• Prerequisites• HACMP/XD for Metro Mirror Features• PPRC Mirroring• Optional Metro Mirror Configurations• HACMP/XD for Metro Mirror in an HACMP Cluster• Where You Go from Here.

PrerequisitesBefore using HACMP/XD for Metro Mirror or reading this guide, you should be familiar with:• HACMP installation and administration• General background and Metro Mirror support for the type of implementation you intend

to configure:• Direct Management PPRC (ESS systems)• DSCLI PPRC (ESS or DS systems)• San Volume Controller (SVC) PPRC• eRCMF (Remote Copy Management Facility).

For more information, refer to the appropriate published or online documentation. For more information on how these products integrate with HACMP, refer to the chapters in this book dealing with individual product configuration.


Overview of HACMP/XD for Metro Mirror HACMP/XD for Metro Mirror Features

16

1

HACMP/XD for Metro Mirror FeaturesHACMP/XD for Metro Mirror extends HACMP cluster management for highly available applications and servers to support the disaster recovery mechanism supplied by PPRC. PPRC is a hardware mirroring technique that the IBM TotalStorage Enterprise Storage Server uses to replicate data. It allows mirroring to be suspended and restarted without affecting data integrity.

For all types of Metro Mirror support, HACMP helps to manage PPRC instances. Direct Management PPRC is the longest-running type of support through HACMP/XD, designed to provide basic PPRC management for ESS systems. In response to demand for simpler management interfaces, HAMCP/XD also provides more automated PPRC management (notably in the area of path and instance (pair) creation) via the following configurations:• DSCLI Management for HACMP/XD Metro Mirror allows additional flexibility by

supporting both ESS and DS storage subsystems with the same interface. • SAN Volume Contoller (SVC) Management for HACMP/XD Metro Mirror provides

storage virtualization and an additional layer of disaster recovery via the SVC cluster and hardware configuration.

• eRCMF Management provides a fully automated management solution, and continuing service during hardware or software outages.

In general, the styles of Metro Mirror support mentioned above (DSCLI, SVC, eRCMF) are intended to make Metro Mirror management easier to install and manage. Direct Management PPRC has the advantage of not having “middle man” software and hardware to maintain, but, in exchange, there is a relatively heavy amount of configuration and maintenance tasks for the administrator.

High Availability and Disaster Recovery Support FeaturesHACMP/XD for Metro Mirror provides high availability and disaster recovery with:• Automatic fallover of PPRC-protected volume pairs between nodes within a site• Automatic fallover of PPRC-protected volume pairs between sites• Automatic recovery/reintegration of PPRC-protected volume pairs between sites • Support for user-defined policy-based resource groups• Support for the following Inter-Site Management Policies for resource groups:

• Prefer Primary Site• Online on Either Site

• Support for VERBOSE_LOGGING and EVENT_EMULATION • Support for the Subsystem Device Driver (SDD)• Support for cluster verification and synchronization• Support for C-SPOC.


Overview of HACMP/XD for Metro Mirror PPRC Mirroring 1

PPRC MirroringPPRC is a mirroring technique used to maintain consistent copies of data across two ESS systems that are at different locations. It mirrors at the disk subsystem level, making it transparent to hosts. PPRC is also called Remote Copy and Mirror. The three types of remote copy and mirror are:• Metro Mirror—synchronous mirroring• Global Mirror—asynchronous mirroring• Global Copy—eXtended Distance mirroring.

In this book, the term PPRC is synonymous with synchronous mirroring.

See the Glossary for more detailed definitions.

Disk SubsystemsIBM TotalStorage Disk subsystems (Enterprise Storage System - ESS, or DS) are large disk subsystems that are typically configured to provide RAID-5 and RAID-10 volumes. They contain two pSeries servers connected to the system backplane to manage and monitor the system.

The disk drives are configured to provide groups of logical disks that are divided into partitions. A data partition in a disk group is referred to as a volume. Each volume is associated with a logical unit number (LUN) that is a subdivision of the physical disk on a storage system.

The disk subsystems have connections to:• pSeries (or other) host servers over SCSI or Fibre Channel links• Other disk subsystems over ESCON or Fibre Channel links. (PPRC uses these physical

links to mirror data.)• ESSNet, the local management Ethernet network for the ESS

Commands sent over this network configure and manage PPRC through ESS Copy Services.

• Nodes that access stored data.

Note: HACMP/XD for Metro Mirror requires that inter-storage links be available to carry data in both directions at the same time. If ESCON links are used, a minimum of two ESCON links is required because each link can carry data in only one direction at a time. You should have at least four links to improve throughput and to provide redundancy of the ESCON cables and adapters.

Note: Refer to the specific chapters for DSCLI, SVC PPRC and eRCMF management for any configuration-specific cabling issues

PPRC OverviewWhen PPRC is activated, it establishes a mirror between specified volumes on two storage systems. The PPRC mirrored volumes are referred to as a PPRC pair, instance, or as a PPRC-protected volume.


Overview of HACMP/XD for Metro Mirror PPRC Mirroring

18

1

PPRC can mirror data in one of two ways:• Synchronous mirroring provides concurrent copy over ESCON links. This is supported by

HACMP/XD.Note that ESCON links transmit data in one direction and have distance limitations.

• Non-synchronous mirroring, also called Global Mirror mirroring, provides copying over longer distances. Extended distance mirroring ensures that volumes are consistent with each other at a point-in-time. This method is not supported in HACMP/XD.

Note: FlashCopy is an optional feature on the ESS. It is an instant (point-in-time) copy of the data.

Hardware Configuration for PPRC Mirroring

The following illustrations show sample hardware configurations that support PPRC-mirroring. The illustrations also show the connections from the ESS to:• Nodes over SCSI or Fibre Channel connections• Fibre channel or ESCON links that connect the two ESS systems to mirror data• The ESS Specialist and ESS Copy Services Specialist over ESSNet and a wide-area

network (WAN) that joins the two ESSNets..

ESS Configuration to Support PPRC


Overview of HACMP/XD for Metro Mirror Optional Metro Mirror Configurations 1

Optional Metro Mirror ConfigurationsThis section is a light overview of each of the optional HACMP/XD for Metro Mirror configurations available. For more detailed information, refer to the appropriate chapters in this book.

DSCLI ManagementDSCLI management is one type of optional extension to PPRC. It provides a simplified PPRC interface for both the ESS and DS storage hardware. The DSCLI interface provides simplified management of PPRC paths and instances in the following manners:• Provides a simplified interface to IBM TotalStorage PPRC services on ESS or DS storage

systems in order to allow management and reporting on PPRC instances and/or paths.• Monitors the status of PPRC relationships and consistency groups between the volumes

being mirrored. It reports any change in status, such as a volume moving to an offline state.

The DSCLI client software interfaces with the ESSNI server on the HMC, SMC or controller connected to the storage to which the DSCLI client is connected.

The following list highlights a few of the specific types of functions that you can perform with the DS command-line interface:• Check and verify your storage unit configuration. • Check the current Copy Services configuration that is used by the storage unit.• Create new logical storage and Copy Services configuration settings.• Modify or delete logical storage and Copy Services configuration settings.

For more information on DSCLI configuration, refer to the current DSCLI online or otherwise published documentation:

http://publib.boulder.ibm.com/infocenter/dsichelp/ds8000ic/index.jsp?topic=/com.ibm.storage.ssic.help.doc/f2c_cliesscli_1kx2so.html

See Chapter 5: HACMP/XD for Metro Mirror with DSCLI Management for information in this book.

SVC PPRCSVC PPRC is another optional extension to HACMP/XD for Metro Mirror. SVC provides both storage virtualization and PPRC management services in the following ways:• Provides virtualized storage that interfaces with the TotalStorage PPRC services to provide

management services and reporting for PPRC relationships (instances,) and/or consistency groups.

• Monitors the status of PPRC relationships and consistency groups between the volumes being mirrored. It reports any change in status, such as a volume moving to an offline state.

• Responds to a site failure (or possible site failure), by suspending all mirroring, and, if necessary, activating the mirror copy at the backup site for host access. All data changes are tracked until the mirrors can be synchronized again.

• Synchronizes the volumes.


http://publib.boulder.ibm.com/infocenter/dsichelp/ds8000ic/index.jsp?topic=/com.ibm.storage.ssic.help.doc/f2c_cliesscli_1kx2so.html

Overview of HACMP/XD for Metro Mirror HACMP/XD for Metro Mirror in an HACMP Cluster

20

1

For more information on SVC PPRC, refer to the San Volume Controller documentation and to Chapter 6: HACMP/XD for Metro Mirror with SVC Management in this document.

eRCMF ManagementeRCMF is an optional extension to PPRC. It integrates with the ESS to simplify PPRC management. eRCMF manages PPRC at two levels:• The enterprise level, to manage the full environment that includes two sites and the ESS

systems and Copy Services Server (CSS) at each site.• The volume set level, to manage volumes in the configured volume sets. A volume set is a

group of volumes that eRCMF manages as a unit.

eRCMF performs the following management tasks:• Monitors the status of PPRC pairs and links between the volumes being mirrored. It informs

you of any change in status, such as a volume moving to a state in which it is not synchronized.

• Responds to a site failure (or possible site failure), by suspending all mirroring, and, if necessary, activating the mirror copy at the backup site for host access. All data changes are tracked until the mirrors can be synchronized again.

• Manages the states of the PPRC and Flashcopy relationships. • Synchronizes the volumes.eRCMF runs on two dedicated AIX systems, referred to as Productivity Center Machines (PCMs). The PCM at each site runs the WebSphere Application Server and incorporates the Copy Services Server for PPRC in addition to running eRCMF. The master eRCMF processes eRCMF commands and monitors the PPRC pairs. If the active PCM fails, the master process switches to the backup PCM.

See Chapter 7: HACMP/XD for Metro Mirror with eRCMF Management and for complete information on eRCMF, see the eRCMF User Guide.

HACMP/XD for Metro Mirror in an HACMP ClusterHACMP/XD for Metro Mirror lets you include PPRC-mirrored volumes in an HACMP cluster. This requires two HACMP sites—an HACMP component to which you assign nodes. Cluster nodes access the same shared volume groups, but the nodes at each site access them from different physical volumes—the two volumes in a single PPRC pair. This is different from a single-site HACMP environment, in which all cluster nodes sharing volume groups have physical connections to the same set of disks.

PPRC Replicated ResourcesA PPRC replicated resource is an HACMP resource that manages a PPRC pair (has a primary and secondary instance that is copied from one site to another). The managing resource group definition includes the volume groups and filesystems built on top of the volume groups defined to a PPRC replicated resource.

HACMP/XD for Metro Mirror supports four types of replicated resources:


Overview of HACMP/XD for Metro Mirror HACMP/XD for Metro Mirror in an HACMP Cluster 1

• Direct Management PPRC replicated resources. A PPRC replicated resource is a PPRC pair associated with an HACMP site that is included in an HACMP cluster. The definition for a PPRC replicated resource contains a volume identifier and the name of the associated ESS. HACMP has knowledge of which volumes are mirrors of each other for each PPRC replicated resource. Synchronous mirroring is supported.

• DSCLI-managed PPRC replicated resources. A DSCLI-managed PPRC replicated resource is a definition of a set of volume pairs and the paths needed to communicate between them. The resource group definition includes the volume groups built on top of the PPRC replicated volumes. Synchronous mirroring is supported.

• SVC-managed PPRC replicated resources. An SVC-managed PPRC replicated resource contains the SVC virtual disk (vDisk) volume pair information combined with the SVC Cluster name bundled in PPRC relationships and consistency group. Synchronous mirroring is supported.

• eRCMF-managed PPRC replicated resources. An eRCMF-managed PPRC replicated resource is a PPRC replicated resource that eRCMF manages. The definition for an eRCMF-managed PPRC replicated resource contains the volume set name, the volume set type, and the production site name. eRCMF has knowledge of which volumes are mirrors of each other for each PPRC replicated resource. An eRCMF-managed PPRC replicated resource supports synchronous or asynchronous mirroring.

Resource Groups that Include PPRC Replicated ResourcesAn HACMP resource group is a collection of resources that comprise the operating environment for an application. Applications, as resources in a resource group, are made highly available.

Resource group management policies direct which node hosts the resource group during normal operation and when the host node fails or goes offline. With HACMP/XD for Metro Mirror, resource group configuration is the same as for other resource groups. In addition, the resource group includes• A shared volume group and the PPRC replicated resources associated with the individual

volumes in the volume group• An intersite management policy to handle a resource group during site recovery.

For information about HACMP resource groups, see the Planning Guide.

Limitations and Restrictions on Resource Groups that Include PPRC Replicated Resources

Refer to the section for the HACMP/XD Metro Mirror configuration you are implementing for additional Limitations lists. In general, the following restrictions apply to HACMP resource groups that will manage PPRC replicated resources because of the way that PPRC instances are managed (i.e. source site nodes will have I/O access, target site nodes will not.) The outcome of this is that any HACMP policy that allows a resource group to come online on more than one site at a time is not supported:• Inter-Site Management Policy Online Both Sides is not supported.• Startup Policies of Online Using Distribution Policy and Online on All Available Nodes

are not supported.• Fallover Policy Fallover Using Dynamic Node Priority is not supported.


Overview of HACMP/XD for Metro Mirror HACMP/XD for Metro Mirror in an HACMP Cluster

22

1

HACMP SitesHACMP/XD for Metro Mirror support requires the use of sites. HACMP supports two sites. The primary site is the active site, and the secondary site is the standby site.

The Inter-Site Management Policy for a resource group directs how a resource group and its resources fall over in response to an outage, and how they fall back if configured to do so. For each resource group, one site is an active production site and the other a backup site. If the nodes at the active production site become unavailable, the backup site becomes the active production site.

For non direct-management types of HACMP/XD for Metro Mirror support (DSCLI, SVC PPRC, eRCMF), each site contains at least one storage system and the nodes attached to it. For HACMP/XD Direct Management environments, each site contains only one ESS.

Resource groups have two types of management policies:• Resource group management policies determine fallover behavior if a node

becomes unavailable.• Site management policies determine fallover behavior if all of the nodes at a site are

not available.

Fallover and FallbackHACMP/XD for Metro Mirror handles the automation of fallover from one site to another in response to an outage at a production site, minimizing recovery time. When a site fails, the resource group configuration determines whether source volumes are accessible from the secondary site.

HACMP automates application recovery by managing replicated resources defined to HACMP resource groups in the following manner:• The fallover of nodes within a site based on node priority (as identified in the nodelist for

a resource group)• The fallover between sites (as specified by the site management policy for a

resource group)• The fallback of a resource group or site as configured.

When an application is running on an active production site:• Updates to the application data are made to the disks associated with the active

production site.• Data is mirrored using PPRC to the backup disks.

If the node or the disks at the production site become unavailable:• The application moves to a server at the backup site.• The application continues operation using the mirrored copy of the data.

When the initial production site becomes active again, resource group and site management policies determine whether or not the application moves back to the previous site:• The direction of mirroring may be reversed.• The application may be stopped and restarted on another node.


Overview of HACMP/XD for Metro Mirror Where You Go from Here 1

• Manual intervention may be required to bring the application back to a nominal functioning state.

For information about configuring resource group and site policies, see the HACMP Planning Guide.

Where You Go from HereThis next chapters of this guide document how to use HACMP/XD for Metro Mirror with PPRC managed by:• Direct management (ESS Copy Services Server)• DSCLI• SVC PPRC.• eRCMF.


Overview of HACMP/XD for Metro Mirror Where You Go from Here

24

1


Chapter 2: HACMP/XD for Metro Mirror General Planning

This chapter describes planning tasks required for all the types of HACMP/XD Metro Mirror support. The chapter contains the following sections:• HACMP/XD for Metro Mirror Management Types• Planning Overview• Planning Worksheets• Planning HACMP Sites• Where You Go from Here.


HACMP/XD for Metro Mirror General Planning HACMP/XD for Metro Mirror Management Types

26

2

HACMP/XD for Metro Mirror Management TypesHACMP/XD for Metro Mirror supports both synchronous and extended-distance PPRC mirrored volumes. The method used to manage PPRC depends on the hardware configuration available, or in the case of eRCMF, the management facility chosen. Using the table below, you can decide what type of PPRC configuration you will need to manage your cluster.

Mirror type Hardware Type

PPRC is managed by

How HACMP manages PPRC pairs

Synchronous ESS 800, etc. Copy Services Server (CSS, on storage controller)

Directly manages the fallover and resynchronization of the PPRC pairs by issuing commands directly to the ESS systems. In this book, this is referred to as Direct Management. In general, it refers to PPRC between ESS storage systems.

Synchronous and Non-synchronous (extended distance)

ESS (800, etc.) eRCMF management of Copy Services Server (on Productivity Center Machine)

Relies on eRCMF to manage PPRC pairs; communicates site events to eRCMF to initiate fallover or fallback of PPRC pairs. In this book, this is referred to as eRCMF management.

Synchronous ESS (800) or DS (8000, 6000) or intermix of any of these

DSCLI management, via ESSNI Server on either storage controller or Hardware Management Console (HMC)

Relies on the ESSNI server to manage PPRC pairs. Directly manages the fallover and resynchronization of the PPRC pairs by issuing commands directly to the storage systems. In this book, this is referred to as DSCLI management.

Synchronous vDisks as provided by SAN Volume Controller. (Hardware as supported by SVC services)

SVC management of PPRC services on SVC-specific hardware

Relies on the Copy Services Server in order to manage PPRC function. Directly manages the fallover and resynchronization of the PPRC pairs by issuing commands directly to the Copy Services Server. In this book, this is referred to as PPRC SVC management.


HACMP/XD for Metro Mirror General Planning Planning Overview 2

Coexistence of Two SolutionsHACMP/XD for Metro Mirror solutions can coexist on the same HACMP cluster only if the PPRC pairs are managed by one of the PPRC solutions at a time. Please refer to the latest support information for which PPRC solutions can successfully coexist on a single HACMP cluster.

Planning OverviewOnce you have decided what type of HACMP/XD for Metro Mirror implementation to set up, you can begin planning the configuration. At this stage, you should be familiar with the planning tasks for base HACMP. For information about planning HACMP clusters, see the HACMP for AIX: Planning Guide.

The following planning is required for all HACMP/XD for Metro Mirror implementations:• Identify sites for HACMP/XD for Metro Mirror.• Identify the resource groups needed to manage the PPRC replicated resources (optional at

this point, can be done at a later stage).• Identify storage systems to be used in the configuration.• Plan for connections to the storage units.• Plan the HACMP/XD for Metro Mirror configuration.• Identify the type of HACMP/XD for Metro Mirror support to be used.

Planning WorksheetsUse the planning worksheets in Appendix A: General HACMP Planning Worksheets to record the HACMP/XD for Metro Mirror configuration information. You can use this information when making configuration entries in SMIT. Sections in this chapter describe the information to provide on these worksheets.

Planning HACMP SitesWithin a resource group, the nodes at one site may handle the PPRC replicated resource differently than the nodes at the other site, especially in cases where the states of the volumes (suspended or full duplex) are different at the two sites.

HACMP/XD for Metro Mirror requires two HACMP sites for use within a resource group to control which volume in a PPRC pair a node can access. Although nodes at both sites can access a volume group, access is permitted to only one volume in a PPRC pair at a time—the source volume. This prevents nodes at different sites from accessing the same volume group at the same time. Typically, a number of volumes are mirrored through PPRC from one site to the other.


HACMP/XD for Metro Mirror General Planning Planning HACMP Sites

28

2

Completing the HACMP/XD for Metro Mirror Site WorksheetAn HACMP/XD for Metro Mirror configuration requires two sites. Each site contains one ESS and the nodes attached to it. Complete a HACMP Site Worksheet for each site.

To define an HACMP/XD for Metro Mirror site, enter the following information on the HACMP Site Worksheet:

1. Record the Site Name. Use no more than 20 alphanumeric characters and underscores. Do not begin the name with a number.

2. Record the names of the Site Nodes. The nodes must have the same names as those you define to the HACMP cluster. A site can have 1 – 7 nodes. A node can belong to only one site.

3. Record the Site Dominance. Set the dominance for one site to yes, and for the other site to no. It does not matter which value you assign to which site.

4. Record the Site Backup Communication Method as None.

Planning HACMP/XD for Metro Mirror Resource Groups Plan your resource groups using the guidelines in Chapter 6: Planning Resource Groups in the HACMP for AIX: Planning Guide. In addition to basic resource group planning, plan for the following resource group attributes:• Sites• PPRC replicated resources.

Note: An HACMP resource group can contain multiple volume groups, file systems, PPRC replicated resources, applications, etc.

When configuring a resource group to use a PPRC replicated resource, the resource group includes nodes in both cluster sites. For the resource group to manage nodes and resources across sites, you assign one of the following inter-site management policies:

• Prefer Primary Site. In a two-site configuration, replicated resources at startup are on the site with the highest priority, fall over to the other site, and then fall back to the site with the highest priority.

• Online on Either Site. Replicated resources are on either site at startup, fall over to the other site and remain on that site after fallover. This selection simplifies resource group takeover rules, which is helpful if you have a number of resource groups.

Note: Each PPRC pair will be included in a resource group whose home node is located at the site that is primary for that PPRC pair.

If you want to set up a mutual recovery configuration for ESS volumes, you configure one resource group in which one site is the active site and another resource group in which the other site is the active site. For an example of a mutual recovery configuration, see the section Sample Configuration in Chapter 4: HACMP/XD for Metro Mirror with Direct Management.


HACMP/XD for Metro Mirror General Planning Planning HACMP Sites 2

Resource Group LimitationsIn HACMP/XD for Metro Mirror configurations, all volume groups in a resource group must contain only PPRC-protected disks.

If HACMP/XD for Metro Mirror is not being used, all volume groups in a resource group must contain only non-PPRC-protected disks.

Each PPRC pair will be included in a resource group whose home node is located at the site that is primary for that PPRC pair.

Completing the PPRC Resource Group WorksheetComplete a Resource Groups for PPRC Replicated Resources Worksheet for each resource group that contains PPRC replicated resources.

To identify information required to include PPRC replicated resources in a resource group, enter the following information on the Resource Groups for PPRC Replicated Resources Worksheet:

1. Record the Resource Group Name for the resource group that will contain the PPRC replicated resource. The name can contain up to 32 characters and underscores. Do not begin the name with a number.

2. Record the startup, fallover, and fallback policies for the resource group.Resource groups that include PPRC replicated resources do not support the startup policy Online on All Available Nodes.

3. Record the Intersite Management Policy: Prefer Primary Site or Online Either Site.Prefer Primary Site is recommended for use with PPRC.

4. Record the names of the Nodes in the resource group, listing the nodes in the prioritized order in which they will appear in the nodelist for the resource group.

5. Plan and record the name of each PPRC replicated resource that will be managed, (and if you will be using Direct Management PPRC, specify whether it is primary or secondary.) You will use this name for the resources and assigned each as primary or secondary on the PPRC-Mirrored Volumes Worksheet appropriate for the HACMP/XD PPRC configuration you are working on.

6. Record the names of any related volume groups (and file systems).


HACMP/XD for Metro Mirror General Planning Where You Go from Here

30

2

Where You Go from HereDepending on what type of HACMP/XD for Metro Mirror configuration you are setting up, go to the following sections to continue planning and move forward into installing and configuring your cluster:

Chapter 3 contains general Installation and upgrading instructions. The chapters for each configuration include some specific installation information.

Management Type Chapter Reference

Direct Management Chapter 4: HACMP/XD for Metro Mirror with Direct Management

DSCLI Chapter 5: HACMP/XD for Metro Mirror with DSCLI Management

SVC PPRC Chapter 6: HACMP/XD for Metro Mirror with SVC Management

eRCMF Chapter 7: HACMP/XD for Metro Mirror with eRCMF Management


Chapter 3: Installing HACMP/XD for Metro Mirror

This chapter describes how to install base HACMP/XD filesets. Details for installing filesets and prerequisites for specific types of PPRC support are contained in subsequent chapters. This chapter contains the following sections:• Installation Prerequisites for HACMP/XD Metro Mirror• Contents of the Installation Media• HACMP/XD for Metro Mirror Installation Choices• Upgrading to HACMP/XD 5.4• Addressing Problems during the Installation• Where You Go from Here.

Note: At runtime, make sure you have at least 9 MB of free space in the /tmp directory. PPRC commands use this directory.

HACMP/XD for Metro Mirror is installed in addition to the HACMP base filesets. For more information on installing base HACMP, refer to the HACMP for AIX 5L Installation Guide.

In order to install HACMP/XD for Metro Mirror, you must install the cluster.xd.license fileset in addition to the base HACMP cluster.license fileset.

Installation Prerequisites for HACMP/XD Metro MirrorAt least the following base HACMP filesets should be installed on all cluster nodes: cluster.doc.en_US.es.html 5.4.0.0 C F HAES Web-based HTML cluster.doc.en_US.es.pdf 5.4.0.0 C F HAES PDF Documentation - U.S. cluster.es.client.lib 5.4.0.0 C F ES Client Libraries cluster.es.client.rte 5.4.0.0 C F ES Client Runtime cluster.es.client.utils 5.4.0.0 C F ES Client Utilities cluster.es.client.wsm 5.4.0.0 C F Web based Smit cluster.es.cspoc.cmds 5.4.0.0 C F ES CSPOC Commands cluster.es.cspoc.dsh 5.4.0.0 C F ES CSPOC dsh cluster.es.cspoc.rte 5.4.0.0 C F ES CSPOC Runtime Commands cluster.es.server.cfgast 5.4.0.0 C F ES Two-Node Configuration cluster.es.server.diag 5.4.0.0 C F ES Server Diags cluster.es.server.events 5.4.0.0 C F ES Server Events cluster.es.server.rte 5.4.0.0 C F ES Base Server Runtime cluster.es.server.testtool cluster.es.server.utils 5.4.0.0 C F ES Server Utilities cluster.license 5.4.0.0 C F HACMP Electronic License cluster.man.en_US.es.data 5.4.0.0 C F ES Man Pages - U.S. English


Installing HACMP/XD for Metro Mirror Installation Prerequisites for HACMP/XD Metro Mirror

32

3

cluster.msg.en_US.cspoc 5.4.0.0 C F HACMP CSPOC Messages - U.S. cluster.msg.en_US.es.client cluster.msg.en_US.es.server

Note: Fileset versions must reflect the version of HACMP you are installing.

In addition, refer to the following table for a list of filesets to install for each type of PPRC management support:

Since each type of PPRC management has different prerequisites, information on installing the particular filesets for specific support types (for example, cluster.es.pprc filesets) is deferred to the chapter specific to that PPRC management type.

Next StepsAt this stage, you can:• Install the base HACMP filesets. Then refer to instructions in this chapter when installing

the HACMP/XD Metro Mirror filesets• Read ahead and install the specific requirements for the option you select, and then refer to

the rest of this chapter to install the HACMP base and PPRC filesets at the same time:• Directly managed PPRC replicated resources: See Chapter 4: HACMP/XD for Metro

Mirror with Direct Management.

PPRC Management Type Filesets to install

Direct Management cluster.es.pprc.cmds

cluster.es.pprc.rte

cluster.msg.en_US.pprc (and/or other appropriate language message sets)

DSCLI cluster.es.pprc.cmds

cluster.es.pprc.rte

cluster.es.spprc.cmds

cluster.es.spprc.rte

cluster.msg.en_US.pprc (and/or other appropriate language message sets)

SAN Volume Controller (SVC)

cluster.es.svcpprc.cmds

cluster.es.svcpprc.rte

cluster.msg.en_US.svcpprc

(and/or other appropriate language message sets)

eRCMF cluster.es.ercmf.cmds

cluster.es.ercmf.rte

cluster.msg.en_US.ercmf

(and/or other appropriate language message sets)


Installing HACMP/XD for Metro Mirror Contents of the Installation Media 3

• DSCLI-managed PPRC replicated resources: See Chapter 5: HACMP/XD for Metro Mirror with DSCLI Management.

• SVC-managed PPRC replicated resources: See Chapter 6: HACMP/XD for Metro Mirror with SVC Management.

• eRCMF-managed PPRC replicated resources: See Chapter 7: HACMP/XD for Metro Mirror with eRCMF Management.

Perform the installation process as the root user.

Contents of the Installation MediaThe HACMP/XD for Metro Mirror installation media provides the following images for installation on each node in the cluster that can take over a PPRC mirrored volume group.

cluster.es.pprc

Installation directory: /usr/es/sbin/cluster/pprc

cluster.es.ercmf

Installation directory: /usr/es/sbin/cluster/ercmf

cluster.es.spprc

Installation directory: /usr/es/sbin/cluster/pprc/spprc

cluster.es.svcpprc

Installation directory: /usr/es/sbin/cluster/svcpprc

cluster.es.pprc.rte HACMP/XD for Metro Mirror runtime commands

cluster.es.pprc.cmds HACMP/XD for Metro Mirror commands

cluster.es.ercmf.rte HACMP/XD for Metro Mirror runtime commands for eRCMF support

cluster.es.ercmf.cmds HACMP/XD for Metro Mirror commands for eRCMF support

cluster.es.spprc.cmds HACMP/XD for Metro Mirror commands for DSCLI support

cluster.es.spprc.rte HACMP/XD for Metro Mirror runtime commands for DSCLI support

cluster.es.svcpprc.cmds HACMP/XD for Metro Mirror commands for SVC PPRC support

cluster.es.svcpprc.rte HACMP/XD for Metro Mirror runtime commands for SVC PPRC support


Installing HACMP/XD for Metro Mirror Contents of the Installation Media

34

3

Direct Management and DSCLI message sets

cluster.msg.ja_JP.pprc

cluster.msg.en_US.pprc

cluster.msg.Ja_JP.pprc

cluster.msg.En_US.pprc

SVC PPRC message Sets

cluster.msg.ja_JP.svcpprc

cluster.msg.en_US.svcpprc

cluster.msg.Ja_JP.svcpprc

cluster.msg.En_US.svcpprc

eRCMF message sets

cluster.msg.ja_JP.ercmf


cluster.msg.ja_JP.pprc HACMP/XD for Metro Mirror—Japanese IBM-eucJP

cluster.msg.en_US.pprc HACMP/XD for Metro Mirror messages—U.S. English

cluster.msg.Ja_JP.pprc HACMP/XD for Metro Mirror messages—Japanese

cluster.msg.En_US.pprc HACMP/XD for Metro Mirror messages—U.S. English

cluster.msg.ja_JP.svcpprc HACMP/XD for SVC PPRC-Japanese IBM-eucJP

cluster.msg.en_US.svcpprc HACMP/XD for SVC PPRC messages-U.S. English

cluster.msg.Ja_JP.svcpprc HACMP/XD for SVC PPRC messages-Japanese

cluster.msg.En_US.svcpprc HACMP/XD for SVC PPRC messages-U.S. English

cluster.msg.ja_JP.pprc HACMP/XD for Metro Mirror for eRCMF support—Japanese IBM-eucJP

cluster.msg.en_US.ercmf HACMP/XD for Metro Mirror messages for eRCMF support—U.S. English


Installing HACMP/XD for Metro Mirror HACMP/XD for Metro Mirror Installation Choices 3

cluster.msg.Ja_JP.ercmf

cluster.msg.En_US.ercmf

User Documentation for all Management Types

cluster.doc.en_US.pprc

HACMP/XD for Metro Mirror Installation ChoicesYou install the HACMP/XD for Metro Mirror software on each cluster node (server). The following sections describe how to install the software in three different ways:• From an installation server• From a hard disk to which the software has been copied• Directly from the installation media.

Installing from an Installation ServerTo install the HACMP/XD for Metro Mirror software in a cluster environment, you can create an HACMP/XD for Metro Mirror installation server (containing the HACMP/XD for Metro Mirror software images for installation) on one node and then load the images onto the remaining cluster nodes. This is the fastest way to install HACMP/XD for Metro Mirror.

HACMP/XD for Metro Mirror supports the Network Installation Management program and Alternate Disk Migration. For instructions on creating an installation server, see the AIX Installation Guide or the AIX Network Installation Management Guide and Reference.

Installing from a Hard DiskTo install HACMP/XD for Metro Mirror software from your hard disk, you copy the software from the installation media to the hard disk prior to installation.

To copy the HACMP/XD for Metro Mirror software to your hard disk:

1. Place the HACMP/XD for Metro Mirror CD into the CD-ROM drive.

2. Enter smit bffcreateThe Copy Software to Hard Disk for Future Installation panel appears.

cluster.msg.Ja_JP.ercmf HACMP/XD for Metro Mirror messages for eRCMF support—Japanese

cluster.msg.En_US.ercmf HACMP/XD for Metro Mirror messages for eRCMF support—U.S. English

cluster.doc.en_US.pprc.html HACMP/XD for Metro Mirror Web-based HTML documentation—U.S. English

cluster.doc.en_US.pprc.pdf HACMP/XD for Metro Mirror PDF documentation—U.S. English


Installing HACMP/XD for Metro Mirror HACMP/XD for Metro Mirror Installation Choices

36

3

3. Enter the name of the CD-ROM drive in the INPUT device / directory for software field and press Enter. If you are unsure of the input device name, press F4 to list available devices. Select the correct drive and press Enter. That drive name appears in the INPUT device / directory field as the valid input device.

4. Press Enter to display the Copy Software to Hard Disk for Future Installation panel.

5. Enter field values as follows:

6. Enter values for the other fields as appropriate for your site.

7. When you are satisfied with the entries, press Enter.SMIT responds: Are you sure?

8. Press Enter again to copy the software.

Installing HACMP/XD for Metro Mirror from the Hard Disk

After the HACMP/XD for Metro Mirror software is on your system (that is not an SP system), follow the instructions in the section Installing the Software from the Installation Media to install the software.

Installing the Software from the Installation MediaIf you install the HACMP/XD for Metro Mirror software from the CD-ROM, install the software directly onto each cluster node.

To install the HACMP/XD for Metro Mirror software on a server node:

1. If you are installing directly from the installation media, insert the CD into the CD-ROM drive.

2. Enter smit install_allSMIT displays the first Install and Update from ALL Available Software panel.

3. Enter the device name of the installation media or install directory in the INPUT device / directory for software field and press Enter. If you are unsure about the input device name or about the install directory, press F4 to list the available devices. Then select the correct device or directory and press Enter. The correct value is entered into the INPUT device / directory field as the valid input device.

4. Enter field values as follows. Press F1 for help on any field.

SOFTWARE name Press F4 for a software listing. Install the images for HACMP/XD for Metro Mirror.For a list of the HACMP/XD for Metro Mirror images, see the section Contents of the Installation Media.

DIRECTORY for storing software

Change the value to the storage directory that all nodes using HACMP/XD for Metro Mirror can access.


Installing HACMP/XD for Metro Mirror HACMP/XD for Metro Mirror Installation Choices 3

Note: You should use F4 to list the software before proceeding with the installation. This way you can install either the English or the Japanese message catalogs.

INPUT device / directory for software

This field shows the device or directory you specified earlier.

SOFTWARE to install Press F4 for a software listing.In the software list, use the arrow keys to locate all software filesets associated with an image. For a list of filesets, see the section Contents of the Installation Media earlier in this chapter. If you want to use eRCMF support, make sure that you install the cluster.es.ercmf image.Next press F7 to select either an image or a fileset. Then press Enter after making all selections. Your selections appear in this field.

PREVIEW only? If set to yes, the preview option checks and verifies that installation prerequisites are met, for instance that required software is installed and sufficient disk space is available. Press F1 for details.When you are ready to perform the actual installation, set this field to no.

COMMIT software updates?

This field applies only when you install software updates (PTFs). See the F1 help for details.

SAVE replaced files? This field applies only when you install software updates (PTFs). If you select no in response to commit software updates? select yes for this field. See F1 help for details.

AUTOMATICALLY install requisite software

Set this field to no if the prerequisite software for version 5.4 is already installed or if the OVERWRITE same or newer versions? field is set to yes; otherwise, set this field to yes to install required software. See F1 help for details.

EXTEND filesystems if space needed?

Select yes if you have adequate hard disk space, no if you have limited space. See F1 help for details.

OVERWRITE same or newer versions?

For new installations, leave this field set to no. Set it to yes if you are reinstalling the software. If you set this field to yes, set the Automatically install requisite software field to no. See the F1 help for details.

VERIFY install and check file sizes?

Select yes if you want the system to perform some checks on the software you installed. See F1 help for details.

DETAILED output? Select yes if you want a detailed log of all installation messages.


Installing HACMP/XD for Metro Mirror Upgrading to HACMP/XD 5.4

38

3

5. When you are satisfied with the entries, press Enter.SMIT responds Are you sure?

6. Press Enter to install the software.

7. After you install the software on a node, reboot that node.

Read the HACMP/XD for Metro Mirror 5.4 Release Notes in the /usr/es/lpp/cluster/doc directory for information that does not appear in the product documentation.

Upgrading to HACMP/XD 5.4This section discusses how to upgrade to HACMP/XD for Metro Mirror 5.4.

Supported Upgrade Path to HACMP/XD 5.4Upgrade is supported from HACMP Remote Copy 5.1 or HACMP/XD 5.2 or 5.3 to HACMP/XD for Metro Mirror 5.4.

Upgrade PrerequisitesBefore you upgrade to HACMP/XD for Metro Mirror 5.4, make sure you are familiar with the process for installing and configuring HACMP/XD 5.4. See the section Installing the Software from the Installation Media.• Make sure that your system meets the installation prerequisites. See the section Installation

Prerequisites for HACMP/XD Metro Mirror.• If you have not completed planning worksheets and diagrams, do so before continuing. You

should transfer all information about your existing installation, plus any changes you plan to make after the upgrade, to these new worksheets.

• Ensure that each cluster node has its own HACMP/XD license.• The HACMP/XD software uses 1 MB of disk space. • Perform the installation process as the root user.

Process multiple volumes?

Select this option if you want to enable the processing of multiple-volume CDs. See F1 for information.

ACCEPT new license agreements?

Select yes for this item to proceed with installation. If you choose no, installation may stop with a warning that one or more filesets require software license agreements. You accept the license agreement only once for each node.

Preview new license agreements?

Select yes to view the text of the license agreements. The text appears in the current window in the language defined on your system.


Installing HACMP/XD for Metro Mirror Upgrading to HACMP/XD 5.4 3

Getting Ready to Upgrade the HACMP/XD SoftwareBefore upgrading your cluster to HACMP/XD 5.4:

1. Archive any localized script and configuration files to prevent losing them during an upgrade.

2. Commit the installation (if it is applied but not committed) so that the HACMP software can be installed over the existing version. To see if your configuration is already committed, enter:lslpp -h cluster.*

3. If the word COMMIT is displayed under the Action header, continue to the next step. If not, run the smit install_commit utility before installing the version 5.4 software. SMIT displays the Commit Applied Software Updates (Remove Saved Files) panel.


5. Make a mksysb backup of each node’s configuration. If you restore a mksysb backup onto your system, you need to reset the SCSI IDs on the system.

6. Save any customized event information.

Upgrading the Software

Note: When upgrading do not leave the cluster at mixed versions for long periods. Mixed versions of the software in the cluster can impact availability within the cluster.

For information about installing the version 5.4 software, see the section Installing the Software from the Installation Media.The PPRC filesets must be upgraded at the same time as the other cluster filesets.

Verifying the Upgraded Cluster DefinitionAfter the HACMP/XD for Metro Mirror software is installed on all nodes, and all nodes have been rebooted, you should verify the configuration. Verification provides errors or warnings to ensure that the cluster definition is the same on all nodes.

To verify the cluster:

1. Enter smit hacmp

2. In SMIT, select HACMP Extended Configuration > HACMP Verification and Synchronization > Verify Changes only and press Enter.Verification verifies both HACMP and HACMP/XD for Metro Mirror configuration.

SOFTWARE name Press F4 for a software listing, and select all of the cluster.* filesets.

COMMIT old version if above version used it? Set this field to yes.

EXTEND filesystem if space needed? Set this field to yes.


Installing HACMP/XD for Metro Mirror Upgrading to HACMP/XD 5.4

40

3

Note: You cannot synchronize a mixed-version cluster. New functionality is available only when all nodes have been upgraded and the cluster has been synchronized. Do not expect commands like clfindres to supply correct information in a mixed cluster.

Recovering from a Failed InstallationWhen you install HACMP/XD, the cl_convert command runs automatically to convert the HACMP configuration database from a previous release to that of the current release.If the installation fails, run cl_convert from the command line to convert the database.

In a failed conversion, run cl_convert using the -F flag. For example, to convert from HACMP v 5.1 to HACMP v. 5.4, use the -F and -v (version) flags as follows:

cl_convert -F -v 5.1.0

To run a conversion utility requires:• root user privileges• The HACMP/XD version from which you are converting

The cl_convert utility logs conversion progress to the /tmp/clconvert.log file so that you can gauge conversion success. This log file is regenerated each time cl_convert or clconvert_snapshot runs.

For more information about cl_convert and clconvert_snapshot, see the respective man pages or see the Appendix on HACMP for AIX Commands in the Administration Guide.

Modifying Previous Cluster SnapshotsAfter you upgrade your HACMP/XD software to version 5.4, you may want to restore one or more of the previous version cluster snapshots you created using the Cluster Snapshot utility. The default directory path for storage and retrieval of a snapshot is /usr/es/sbin/cluster/snapshots; however, you may have specified an alternate path using the SNAPSHOTPATH environment variable. Check in these locations before using the /usr/es/sbin/cluster/conversion/clconvert_snapshot utility to convert the snapshot.

The snapshot is based on your full HACMP configuration—including the configuration to include PPRC replicated resources in a cluster. The clconvert_snapshot utility updates HACMP configuration data with new information for version 5.4.

To convert and apply a cluster snapshot, enter:clconvert_snapshot -v version# -s snapshot_file_name

where the -s flag is used with the snapshot file name you want to update or apply, and the -v flag is used with the version of the saved snapshot.

For example, if you want to convert an HACMP 5.1.0 snapshot named my_510_snapshot to HACMP 5.4, you would enter:clconvert_snapshot -v 5.1.0 -s my_510_snapshot

For more information about creating, restoring, and applying cluster snapshots in a cluster environment, see the chapter Saving and Restoring Cluster Configurations in the Administration Guide.


Installing HACMP/XD for Metro Mirror Addressing Problems during the Installation 3

Addressing Problems during the InstallationIf you experience problems during the installation, the installation program usually performs a cleanup process automatically. If, for some reason, the cleanup is not performed after an unsuccessful installation:

1. Enter smit install to display the Installation and Maintenance menu.

2. Select Install and Update Software.

3. Select Clean Up After a Interrupted Installation.

4. Review the SMIT output (or examine the /smit.log file) for the cause of the interruption.

5. Fix any problems and repeat the installation process.

Where You Go from HereAfter you successfully install the software on each cluster node, you are ready to configure HACMP/XD for Metro Mirror for one of the following:

Management Type Chapter Reference

Direct Management Chapter 4: HACMP/XD Metro Mirror with Direct Management

DSCLI Chapter 5: HACMP/XD Metro Mirror with DSCLI Management

SVC PPRC Chapter 6: HAMCP/XD Metro Mirror with SVC Management

eRCMF Chapter 7: HAMCP/XD Metro Mirror with eRCMF Management


Installing HACMP/XD for Metro Mirror Where You Go from Here

42

3


Chapter 4: HACMP/XD for Metro Mirror with Direct Management

This chapter describes the planning, installation, and configuration tasks for setting up HACMP/XD for Metro Mirror with Direct management. Using this method, HACMP directly manages PPRC pairs by communicating with the ESS Copy Services Server.

HACMP/XD for Metro Mirror with Direct Management is the oldest type of PPRC support HACMP provides. It has the simplest hardware configuration, but requires more work on the administrator's part to set up.

Briefly, HACMP provides support for direct management of ESS PPRC by managing specified pair and path tasks as defined by the user on their ESS storage systems Copy Services Server (CSS.) HACMP provides monitoring, fallover and fallback support for PPRC by issuing commands directly to the CSS via the ESS CLI interface. Refer to the general overview in Chapter 1 of this book for a full description of this type of configuration.

Direct Management can be used to support PPRC between ESS 800 storage systems.

The chapter contains the following sections:• Planning for Direct Management• Installing HACMP/XD for Metro Mirror for Direct Management• Configuring in a Direct Management Environment.

Planning for Direct Management You should be familiar with the planning tasks for HACMP. For information about planning HACMP clusters, see the Planning Guide.

You should have completed the planning steps in Chapter 2: HACMP/XD for Metro Mirror General Planning for HACMP sites, at least.

To continue the plan for an HACMP/XD for Metro Mirror with Direct Management environment:• Plan for connections to the ESS• Plan for ESS Copy Services• Plan the HACMP/XD for Metro Mirror configuration:

• Identify ESS systems to be included• Identify which resource groups will contain the PPRC replicated resources

• Identify PPRC replicated resources and provide information about the associated volumes• (Optional) Plan for any user-specific PPRC tasks.


HACMP/XD for Metro Mirror with Direct Management Planning for Direct Management

44

4

Planning PrerequisitesHACMP/XD PPRC with Direct Management manages PPRC resources by communicating with the Copy Services Server (CSS) on ESS systems via the ESS CLI. Accordingly, prior to configuring Direct Management, ensure the following:• The version of the ESS CLI shipped with the microcode for your storage system has been

installed on all HACMP cluster nodes. • You have access to the ESS Copy Services Web Interface for the storage systems involved

in your configuration.

ESS Copy Services

The ESS Copy Services Web Interface provides a configuration interface for setting up Direct Management PPRC. For complete information on PPRC and the ESS Copy Services, see the ESS Web Interface User's Guide on the IBM website at the following URL:

http://publibfp.boulder.ibm.com/epubs/pdf/f2bui04.pdf

For more information on Direct Management PPRC, refer to the appropriate chapters in this document.

See also Installation Prerequisites later in this chapter for more information on software and installation prerequisites.

Limitations

You can only use non-concurrent volume groups with PPRC.

Disk heartbeat networks are not supported in HAMCP for this configuration.

Sample ConfigurationYou can set up a mutual recovery configuration in which each site acts as a production site with the other site acting as an associated backup site. Implementing a mutual recovery configuration requires: • Two HACMP sites (the same as for a single recovery configuration)• Two resource groups.

The following figure shows a mutual recovery configuration with two resource groups in which Site_2 is the production site with Site_1 as the recovery site for Resource Group 2, and Site_1 is the production site with Site_2 as the recovery site for Resource Group 1. In Resource Group 1 the nodes n1 and n2 have a higher priority, and in Resource Group 2 the nodes n3 and n4 have a higher priority. The volume groups in both resource groups are included in PPRC replicated resources.



HACMP/XD for Metro Mirror with Direct Management Planning for Direct Management 4

Sample Configuration for Mutual Recovery of PPRC Replicated Resources

The order of the nodes in each resource group’s nodelist indicates which site is considered to be the production site and which site is considered to be the backup site for that resource group. The site that includes the nodes with the highest priority is considered the production site for the resource group.

In the preceding example, the configuration for resource group 1 would be:

In the preceding example, the configuration for resource group 2 would be:

Planning WorksheetsUse the planning worksheets in Appendix B: Planning Worksheets for Direct Management (ESS CLI) or to record the HACMP/XD for Metro Mirror configuration information. You can use this information when making configuration entries in SMIT. Sections in this chapter describe the information to provide on these worksheets.

You also need the HACMP site planning worksheet from Appendix A: General HACMP Planning Worksheets.

Ordered sitelist Site_1, Site_2

Primary volume volume100@ESS_1 (mirrored volume on ESS_1)

Secondary volume volume200@ESS_2 (mirrored volume on ESS_2)

Ordered nodelist n1, n2, n3, n4


Primary volume volume210@ESS_2 (mirrored volume on ESS_2)

Secondary volume volume110@ESS_1 (mirrored volume on ESS_1)




46

4

Planning Integration with PPRCBefore you configure HACMP/XD for Metro Mirror, you configure PPRC on the ESS. To configure PPRC, you set up:• PPRC paths, ESCON links that transfer mirrored data from one ESS to another• PPRC pairs, volumes mirrored from one ESS to another.

The PPRC paths connect PPRC pairs.

For information about configuring PPRC, see the ESS Web Interface User’s Guide on the IBM website at the following URL:


Planning Connections to the ESS

HACMP/XD for Metro Mirror communication with the ESS requires connections to each of the ESS subsystems through the ESSNet administrative network. This access lets HACMP/XD for Metro Mirror use the ESS Command Line Interface (CLI) to control the ESS subsystems during event processing. Commands are sent to the ESS Copy Services Server software that runs on one of the ESS subsystems in the environment.

The HACMP cluster nodes can either connect directly to the ESSNet or connect over a router or bridge. The interfaces that communicate with the ESSNet can be defined to HACMP, but IP address takeover (IPAT) should not be used on those interfaces. If the interfaces are defined to HACMP, ensure that the nodes can gain access to ESSNet at any time.

Note: The reliability of the connection between a node and the ESSNet directly affects the performance of HACMP/XD for Metro Mirror. A slow or unreliable network connection to ESSNet results in commands to initiate a fallover being processed slowly or unreliably.

Planning Copy Services Server on ESS

Considerations for using Copy Service Server differ depending on the PPRC version.

PPRC Versions before 2.2When you set up ESS Copy Services, you define one ESS cluster processor complex as the active Copy Services Server and one ESS cluster processor complex on another ESS as the backup Copy Services Server.

If the active Copy Services Server fails, a notification message is sent. Because PPRC versions prior to 2.2 do not support dual active Copy Services Servers, you manually reset the Copy Services Server to make the backup server assume the active role. HACMP cannot restart the Copy Services Server.

Set up the active Copy Services Server at the recovery site.

PPRC Version 2.2 or LaterWhen you set up ESS Copy Services, you define one ESS cluster processor complex in each ESS as a Copy Services Server. Because both Copy Services Servers are active at all times, you do not need to manually restart a Copy Service Server if one fails.




Configuring PPRC Paths

When configuring PPRC paths that connect PPRC pairs to be included in a PPRC replicated resource, keep in mind that each ESCON link:• Can carry up to 64 paths at the same time.• Carries data in one direction at a time.

HACMP/XD for Metro Mirror requires that ESCON links be available to carry data in both directions at the same time; therefore, you need a minimum of two physical links. To improve throughput and to provide redundancy for the ESCON cables and adapters, you should have at least four links. Note that the ESS manages the paths.

Planning for PPRC Replicated ResourcesPPRC pairs are defined so that one volume in the pair resides on the ESS at one site, and the other volume in the pair resides on the ESS at the other site. In a PPRC pair, the volume that data is being written to is the source volume, and the volume that contains a mirrored copy of the data is the target volume. The definition for a PPRC replicated resource contains the volume identifier and the name of the ESS for the source and the target volumes. HACMP has knowledge of which volumes are mirrors of each other for each PPRC replicated resource.

To plan PPRC replicated resources for HACMP/XD for Metro Mirror, you should have a good understanding of the volume groups and volumes involved. When you complete the PPRC-Mirrored Volumes Worksheet, you provide information about all of the PPRC pairs that will be configured as PPRC replicated resources.

The terms primary and secondary are specific to each pair of PPRC-mirrored volumes as determined by the pair definition. In configurations that designate one site as a production site and the other as a backup site, the production site holds the primary volumes for the PPRC pairs, and the backup site holds the secondary volumes. Each PPRC pair is included in a resource group whose home node is located at the site that is primary for the PPRC pair.

In a mutual recovery configuration, in which nodes from both sites are active, each site contains primary volumes for some PPRC pairs and secondary volumes for others.

Note: PPRC copies the volume information, including the PVID, from one volume in a PPRC pair to the other. The volumes at both sites contain the same logical volumes and must therefore be imported with the same volume group name. This also allows single-name entries in a resource group definition.

Use ESS Copy Services to obtain information about the ESS configuration and the PPRC pairs that have been established. For information about using ESS Copy Services, see the ESS Web Interface User’s Guide on the IBM website at the following URL:





48

4

Identifying Volume Groups

You use the volume group name when configuring PPRC tasks in ESS Copy Services and when you identify the volume group to an HACMP node. Make sure that the name is the same in both places. When completing the PPRC-Mirrored Volumes Worksheet, you specify the volume group.

For information about defining shared volume groups, see the section Importing Mirrored Volume Groups at the Secondary Site in the section Configuring in a Direct Management Environment later in this chapter.

Note: The volume group and logical volume names must be the same on all nodes in each site where the volume group can be brought online.

To identify volume group information on the ESS:

1. Run the lspv command to view which volume group is associated with which hdisk.

2. Run the rsList2105.sh command to see which hdisks are associated with which serial numbers.

You can also use the PPRC rsPrimeServer command to assign names that are then visible in ESS Copy Services. For information about the rsPrimeServer command, see the ESS Command-Line Interface User’s Guide on the IBM website at the following URL:

http://www.storage.ibm.com/disk/ess/

Completing the PPRC-Mirrored Volumes Worksheet

Complete the PPRC-Mirrored Volumes Worksheet to record information that you use to define PPRC replicated resources.

Note: On the PPRC-Mirrored Volumes Worksheet, enter the volume information so that the information for each volume in a PPRC pair is on the same line.

To identify the information needed to define PPRC replicated resources, enter the following information on the PPRC-Mirrored Volumes Worksheet:

1. Record the Site Name for each of the HACMP sites in the cluster. You can copy this information from the information on the HACMP Site Worksheet.

2. Record the ESS Serial Number for the ESS at each site.You can copy this information from the information on the HACMP Site Worksheet.

3. For each pair of PPRC-mirrored volumes, provide the information in the following list for each volume in a PPRC pair.

Volume ID The identifier, or serial number, for the volume. To determine the volume ID, run the following command:lscfg -vl diskname

where diskname is the name of the hdisk that is presented to AIX. The Serial Number field contains the volume ID.





4. Specify Primary or Secondary for each volume. Assign the volume on one ESS as primary and the volume on the other ESS as secondary. For information about primary and secondary volumes, see the section Planning for PPRC Replicated Resources.

Note: Each PPRC pair will be included in a resource group whose home node is located at the site that is primary for that PPRC pair.

5. Record the Volume Group name for each group that contains PPRC-protected volumes.Volumes that are part of the same volume group should be located in the same LSS.For information about identifying the volume group, see the section Identifying Volume Groups.

6. Record the PPRC Replicated Resource Name, which is a management name to identify the resource. Use a name that is meaningful for your configuration.For example, a PPRC replicated resource could be named pprc4.1, with pprc signifying the type of replicated resource, 4 identifying the volume group to which the PPRC-protected volume belongs, and 1 enumerating a PPRC-protected volume in the volume group. Another PPRC replicated resource that includes a different volume in the same volume group could be named pprc4.2.

Planning Tasks for the ESSTasks are a set of actions to be executed on an ESS. They let you automate a series of steps that otherwise would be performed through the ESS Copy Services Web Interface. In a direct management environment, HACMP/XD for Metro Mirror relies on PPRC tasks for managing PPRC volume pairs and the associated PPRC paths. These tasks are used to establish and terminate PPRC pairs during fallover and fallback of volumes. To access these tasks, nodes in an HACMP cluster that supports HACMP/XD for Metro Mirror require access to ESSNet to execute commands.

Using User-Specific Task Names

Unless you have a configuration conflict (for example if you already have a task named with one of the recommended task names), you should name these tasks as listed in the section Configuring PPRC Tasks.

If you need to use user-specific task names, specify the names on the User-Specific Tasks Worksheets. Task names can be up to16 characters long. If possible, base the task name on the naming conventions described in the section Understanding Task Names.

Logical Subsystem ID The identifier for the logical subsystem (LSS).You can view this information from the Modify Volume Assignments Panel available in the ESS Copy Services Web Interface.Note: This number is used when setting up the paths on the ESS. You provide the information on the worksheet for reference only.



50

4

Note: When you specify a user-specific task name for any task, you need to specify task names for all of the tasks (whether or not they are different from the recommended names) for each volume group.

You configure these tasks in ESS Copy Services and define them to HACMP/XD for Metro Mirror. For information about defining a group of user-specific task names to HACMP, see the section Defining the PPRC Tasks to HACMP/XD for Metro Mirror for Direct Management later in this chapter.

Note: You should use the recommended task names, unless there is a very strong reason to do otherwise. If you do change names, ensure that the names are entered correctly in ESS Copy Services and in HAMCP/XD for PPRC.

Understanding Task Names

The recommended names for the PPRC tasks specify the volume group and PPRC tasks to manipulate the volume pairs in the volume group.

The recommended task names use the following naming conventions:

volume_ group_action_object_direction_modifier_system

A task name may contain some or all of the components listed in the preceding syntax line.



The following table shows the values for the parts of the task name:

Planning for ESS Disk Subsystems Used By Direct ManagementIn addition to planning the PPRC replicated resources, you need to plan:• Administrative information for the ESS subsystems• Copy Services.

Completing the ESS Disk Subsystem Worksheet

Complete an ESS Disk Subsystem Worksheet for each ESS system that contains PPRC replicated resources.

volume_ group The name of the volume group (less than 7 characters long)

_action The action to be executed:

EstSusTerDel

EstablishSuspendTerminateDelete

_object The pair or path for the volume group:

PrPt

PairPath

_direction The direction of the operation:

PSSP

Primary to secondarySecondary to primary

_modifier The type of copy:

NCFCSCFOFBF

No copyFull copySync changesFailoverFailbackForced (only for the delete action)

_system The type of system on which the action is being executed:

PS

PrimarySecondary


HACMP/XD for Metro Mirror with Direct Management Installing HACMP/XD for Metro Mirror for Direct Management

52

4

To define an ESS Disk Subsystem, enter the following information on the ESS Disk Subsystem Worksheet:

1. Record the Site Name, which is the name of the site where the system is located. Select the name of one of the sites identified in the section Planning a PPRC Resource Group.

2. Record the ESS Disk Subsystem Name. Enter a name with 30 or fewer characters.

3. Record the ESS IP Address in dotted decimal format. The Copy Services Server for an ESS uses this IP address.

Note: The IP address specified here is the ESSNet address of the ESS at this site.

4. Record the User ID, which is the administrative user ID that authenticates on this ESS.

5. Record the ESS Password, which is the password associated with the user ID.

Planning a PPRC Resource Group If you have not done so, plan your resource groups using the guidelines in Chapter 2: HACMP/XD for Metro Mirror General Planning as well as consulting Chapter 6: Planning Resource Groups in the HACMP Planning Guide.

Next StepsAfter you have planned the configuration for Direct Management, you are ready to install the HACMP/XD for Metro Mirror software on HACMP cluster nodes.

Installing HACMP/XD for Metro Mirror for Direct Management

This section describes how to install HACMP/XD for Metro Mirror for Direct Management. Details for installing base HACMP/XD from the installation media are included in Chapter 3: Installing HACMP/XD for Metro Mirror. This section contains the following subsections:• Installation Prerequisites• Software Requirements• Installing Filesets• Addressing Problems during the Installation.

Note: At runtime, make sure you have at least 9 MB of free space in the /tmp directory. PPRC commands use this directory.

Installation PrerequisitesBefore you install HACMP/XD Metro Mirror for Direct Management, make sure that the following is true for each cluster node:• The HACMP/XD software uses 1 MB of disk space. Ensure that the /usr filesystem has 0.8

MB free disk space for the upgrade.


HACMP/XD for Metro Mirror with Direct Management Installing HACMP/XD for Metro Mirror for Direct Management 4

• You have root access to each node.

Software RequirementsThe following software is required for HACMP/XD Metro Mirror:• AIX level as stated in the support flash• HACMP 5.4 plus all current APARs• Java Runtime Environment version appropriate for the ESS CLI version installed. (This

will be tied to and should be available with the microcode level on the ESS systems.)• ESS CLI as appropriate for your storage microcode (LIC) level:

The ESS Copy Services CLI software is closely tied to the ESS microcode level. CLI code is found on the MegaCDR that is included in the Customer Software packet.• IBM 2105 Command Line Interface (ibm2105cli.rte)or• IBM 2105 Command Line Interface (ibm2105esscli.rte) (also included on the

MegaCDR).Depending on which version of the ibm2105esscli fileset ships with your level of ESS microcode, the files will be installed in one of three locations:

a. /usr/opt/ibm2105cli

b. /usr/opt/ibm/ibm2105cli

c. /opt/ibm/ESScli HACMP/XD assumes a default directory for the ESS CLI executables. Therefore, once the CLI executables are installed in one of the directories listed, you must create the following link:/usr/opt/ibm2105cli -> <ESS cli installation location>so that HACMP/XD can find the executables.See the IBM TotalStorage support web pages for updated information on ESS CLI levels and installation site. http://www.ibm.com/servers/storage/support/software/cscli/

• (Optional but recommended) ESS microcode level vrmf 2.2.x.x. This version provides support for a dual active Copy Services Server.

Note: Install microcode vrmf 2.2.x.x on the ESS system to make support for a dual active Copy Services Server available. Otherwise, the Copy Services Server must be manually started on a backup Copy Services Server if the primary Copy Services Server is unavailable.


http://www.ibm.com/servers/storage/support/software/cscli/

HACMP/XD for Metro Mirror with Direct Management Installing HACMP/XD for Metro Mirror for Direct Management

54

4

Installing FilesetsIf you have not already done so, install the following filesets according to instructions listed in Chapter 3: HACMP/XD PPRC Installation.

Installation directory: /usr/es/sbin/cluster/pprc

cluster.doc.en_US.pprc

Install the appropriate message set for your locale:

cluster.msg.ja_JP.pprc

cluster.msg.en_US.pprc

cluster.msg.Ja_JP.pprc

cluster.msg.En_US.pprc

Addressing Problems during the InstallationIf you experience problems during the installation, the installation program usually performs a cleanup process automatically. If, for some reason, the cleanup is not performed after an unsuccessful installation:

1. Enter smit install to display the Installation and Maintenance menu.

2. Select Install and Update Software.

3. Select Clean Up After an Interrupted Installation.

4. Review the SMIT output (or examine the /smit.log file) for the cause of the interruption.

5. Fix any problems and repeat the installation process.

cluster.es.pprc.rte HACMP/XD for PPRC runtime commands

cluster.es.pprc.cmds HACMP/XD for PPRC commands

cluster.doc.en_US.pprc.html HACMP/XD for PPRC Web-based HTML documentation-U.S. English

cluster.doc.en_US.pprc.pdf HACMP/XD for PPRC PDF documentation-U.S. English

cluster.msg.ja_JP.pprc HACMP/XD for PPRC-Japanese IBM-eucJP

cluster.msg.en_US.pprc HACMP/XD for PPRC messages-U.S. English

cluster.msg.Ja_JP.pprc HACMP/XD for PPRC messages-Japanese

cluster.msg.En_US.pprc HACMP/XD for PPRC messages-U.S. English


HACMP/XD for Metro Mirror with Direct Management Configuring in a Direct Management Environment 4

Configuring in a Direct Management EnvironmentThis section describes how to configure the ESS system to support HACMP/XD for Metro Mirror, and how to configure an HACMP/XD direct management (ESS CLI) environment. This section contains the following subsections:• Configuration Prerequisites• Configuration Overview• Configuring HACMP/XD for Metro Mirror for Direct Management Support• Configuring the HACMP/XD Metro Mirror Cluster for Direct Management• Verifying and Synchronizing the PPRC Configuration• Configuring Resource Groups• Starting the Cluster• Changing PPRC Replicated Resources Configuration• Changing the Configuration for Sites• Changing the Configuration for PPRC Replicated Resources.

Configuration PrerequisitesBefore configuring HACMP/XD for Metro Mirror, ensure that:• PPRC is configured and running on the ESS systems.

HACMP/XD for Metro Mirror supports RS/6000 and pSeries servers that can support ESS systems.

• You have a good understanding of HACMP sites for PPRC replicated resources. For more information about sites, see the section HACMP Sites in Chapter 1: Overview of HACMP/XD for Metro Mirror.

• The HACMP/XD for Metro Mirror software is installed on each cluster node that will support HACMP/XD for Metro Mirror.For information about installing HACMP/XD for Metro Mirror, see Chapter 3: Installing HACMP/XD for Metro Mirror.

• The HACMP cluster is configured for:• Nodes• Networks and network interfaces• Service labels• Initial resource groups.

You can modify the attributes for a resource group later to accommodate PPRC replicated resources.

Configuration OverviewTo set up an HACMP/XD for Metro Mirror configuration:

1. On the ESS, configure PPRC support for HACMP/XD for Metro Mirror:

a. Configure PPRC tasks in ESS Copy Services


HACMP/XD for Metro Mirror with Direct Management Configuring in a Direct Management Environment

56

4

b. Import the volume groups on the nodes at the secondary site.

2. On cluster nodes, configure support for HACMP/XD for Metro Mirror:

a. Define HACMP sites for HACMP/XD for Metro Mirror.

b. Define ESS systems to be included.

c. Define PPRC replicated resources.

d. Configure resource groups to include PPRC replicated resources.

e. (Optional) If user-specific task names are used, define them to HACMP.

f. (Optional) If your cluster has a large number of volume groups, create LUN ID mapping files before startup as described in Improving the Performance of Volume Groups.

Configuring HACMP/XD for Metro Mirror for Direct Management SupportDirect Management configuration relies on tasks configured through ESS Copy Services and activated on the ESS system. For information about tasks, see the section Planning Tasks for the ESS earlier in this chapter.

Configure the ESS system in preparation for HACMP/XD for Metro Mirror before making configuration changes in HACMP/XD for Metro Mirror.

Configuring PPRC Tasks

You create 24 PPRC tasks for each PPRC-protected volume group in the cluster. This section lists each of these tasks and lists the options in ESS Remote Copy Services that you use to set up the task.

When you create tasks, use the ESS Copy Services Web Interface. In this interface, serial numbers identify volumes. You select disks by serial number, and then create tasks that start with the volume group name. Consult the PPRC-Mirrored Volumes Worksheet for the volume group name.

For more information on using the ESS Copy Services Web Interface, see the ESS Web Interface User's Guide on the IBM website at the following URL:


Note: Use the task names recommended, unless there is a very strong reason to do otherwise. If you do change names, ensure that these names are entered correctly in ESS Copy Services and in HACMP/XD for Metro Mirror.

If you name a task other than the recommended name, see the section Defining the PPRC Tasks to HACMP/XD for Metro Mirror for Direct Management.

Note: Enter the task names carefully. Direct Management does not work if any of the task names are inaccurate.

To configure tasks for a volume group to use HACMP/XD for Metro Mirror:

1. Log into ESS Copy Services as a user with administrative privileges.




2. Use the ESS Copy Services Web Interface to define the ESS tasks needed to manage the paths (ESCON links) that PPRC uses for the volume group.Create the tasks as shown in the following list. Use the task names shown here, substituting the name of the volume group for $vgname. This name can be up to seven characters long, for example vg1.

Note: Task names are case-sensitive.

Consult the PPRC-Mirrored Volumes Worksheet for the volume group name. For more information about volume group names, see Identifying Volume Groups.

3. Use the ESS Copy Services Web Interface to define the tasks needed to manage the pairs of volumes that are PPRC-protected. If you define multiple volumes in the volume group, you can use the Multiple Selection option on the ESS Copy Services Web Interface to define a single set of tasks to manage the entire volume group.

Task Name ActionOption in ESS Copy Services to Define Task

$vgnameEstPtPS Establish path(s) from the Logical Subsystem (LSS) in the primary ESS to the LSS in the secondary ESS.

Do not establish paths if they already exist

$vgnameDelPtPSF Remove paths from the LSS in the primary ESS to the LSS in the secondary ESS.

Force removal of PPRC path even if pairs exist

$vgnameDelPtPS Remove paths from the LSS in the primary ESS to the LSS in the secondary ESS.

----

$vgnameEstPtSP Establish paths from the LSS in the secondary ESS to the LSS in the primary ESS.

Do not establish paths if they already exist

$vgnameDelPtSP Remove paths from the LSS in the secondary ESS to the LSS in the primary ESS.

----

$vgnameDelPtSPF Remove paths from the LSS in the secondary ESS to the LSS in the primary ESS.

Force removal of PPRC path even if pairs exist



58

4

Create the tasks as shown in the following list. Use the task names shown here, substituting the name of the volume group for $vgname. Consult the PPRC-Mirrored Volumes Worksheet for the volume group name. For more information about volume group names, see Identifying Volume Groups.


$vgnameEstPrPSFC Establish a PPRC relationship from the source volume(s) in the primary ESS to the target volume(s) in the secondary ESS.

Copy Entire VolumePermit Read from Secondary

$vgnameEstPrPSNC Establish a PPRC relationship from the source volume(s) in the primary ESS to the target volume(s) in the secondary ESS.

Do Not Copy VolumePermit Read from Secondary

$vgnameEstPrPSSC Establish a PPRC relationship from the source volume(s) in the primary ESS to the target volume(s) in the secondary ESS.

Copy Out-of-Sync Cylinders OnlyPermit Read from Secondary

$vgnameEstPrPSFO Establish a PPRC relationship from the target volume(s) in the primary ESS to the source volume(s) in the secondary ESS.

PPRC Failover

$vgnameEstPrPSFB Establish a PPRC relationship from the target volume(s) in the primary ESS to the source volume(s) in the secondary ESS.

PPRC FailbackPermit Read from Secondary

$vgnameSusPrPSP Suspend a PPRC relationship between the source volume in the primary ESS and the target volume in the secondary ESS. This task is executed on the primary ESS.

----

$vgnameSusPrPSS Suspend a PPRC relationship between the source volume in the primary ESS and the target volume in the secondary ESS.This task is executed on the secondary ESS.

----



$vgnameTerPrPSP Terminate a PPRC relationship from the source volume in the primary ESS to the target volume in the secondary ESS. This task is executed on the primary ESS.

----

$vgnameTerPrPSS Terminate a PPRC relationship from the source volume in the primary ESS to the target volume in the secondary ESS.This task is executed on the secondary ESS.

----

$vgnameEstPrSPFC Establish a PPRC relationship from the source volume(s) in the secondary ESS to the target volume(s) in the primary ESS.

Copy Entire VolumePermit Read from Secondary

$vgnameEstPrSPNC Establish a PPRC relationship from the source volume(s) in the secondary ESS to the target volume(s) in the primary ESS.

Do Not Copy VolumePermit Read from Secondary

$vgnameEstPrSPSC Establish a PPRC relationship from the source volume(s) in the secondary ESS to the target volume(s) in the primary ESS.

Copy Out-of-Sync Cylinders OnlyPermit Read from Secondary

$vgnameEstPrSPFO Establish a PPRC relationship from the target volume(s) in the secondary ESS to the source volume(s) in the primary ESS.

PPRC Failover

$vgnameEstPrSPFB Establish a PPRC relationship from the target volume(s) in the secondary ESS to the source volume(s) in the primary ESS.

PPRC FailbackPermit Read from Secondary

$vgnameTerPrSPS Terminate a PPRC relationship from the source volume(s) in the secondary ESS to the target volume(s) in the primary ESS.This task is executed on the secondary ESS.

----




60

4

4. Run the ESS rsExecuteTask.sh command to verify that you can execute the tasks created in the previous step. Note that the command name is case-sensitive.The rsExecuteTask.sh command should return to a command prompt. If it does not, HACMP/XD for Metro Mirror does not work.If the command does not return to a command prompt:• Review task configuration and make corrections as needed.• Run the rsExecuteTask.sh command again.

Note: A host server IP address not being defined to the ESS causes a known configuration problem with the ESS. If this is present, the rsExecuteTask.sh command executes the ESS task, but the program does not return.

For information about the rsExecuteTask.sh command, see the ESS Command-Line Interface User’s Guide on the IBM website at the following URL:http://www.storage.ibm.com/disk/ess/

5. Repeat steps 1 – 4 for each volume group that will use HACMP/XD for Metro Mirror.

Importing Mirrored Volume Groups at the Secondary Site

If you have already created volume groups and are sure that volumes are mirroring correctly, skip this section.

$vgnameTerPrSPP Terminate a PPRC relationship from the source volume(s) in the secondary ESS to the target volume(s) in the primary ESS.This task is executed on the primary ESS.

----

$vgnameSusPrSPS Suspend a PPRC relationship between the source volume in the secondary ESS and the target volume in the primary ESS.This task is executed on the secondary ESS.

----

$vgnameSusPrSPP Suspend a PPRC relationship between the source volume in the secondary ESS and the target volume in the primary ESS.This task is executed on the primary ESS.

-----






If you have not created the volume groups, complete the procedure in this section. Refer to the PPRC-Mirrored Volumes Worksheet for the names of the volume groups. The secondary volume in a PPRC pair is not visible to the nodes at the secondary site when the PPRC relationship is active. As a result, use the following procedure to define the disks and volume groups for HACMP.

To define disks and volume groups to HACMP:

1. Log into ESS Copy Services as a user with administrative privileges.

2. Run the $vgnameEstPtPS task, where $vgname is the name of the volume group. This establishes the path(s) from the primary volume to the secondary volume.

3. Run the $vgnameEstPrPSFC task, where $vgname is the name of the volume group. This establishes the PPRC pair(s) from the primary volume to the secondary volume and synchronizes the two copies. This may take awhile.

4. Create the volume group using the AIX Logical Volume Manager (LVM) on nodes on the active site.

5. Run the $vgnameTerPrPSP task, where $vgname is the name of the volume group.This terminates the PPRC pair(s).

6. Run the $vgnameDelPtPS task, where $vgname is the name of the volume group.This terminates the path(s).

7. Run Configuration Manager (cfgmgr) on the backup host server to move ESS hdisks to the available state.

8. Import the volume group to all nodes on the secondary site.

The volume group can now be defined to HACMP/XD for Metro Mirror.

Configuring the HACMP/XD Metro Mirror Cluster for Direct ManagementConfigure HACMP/XD for Metro Mirror after you:• Configure tasks for PPRC on the ESS systems• Define sites to support PPRC replicated resources in HACMP.

Defining PPRC Replicated Resources

Most of the configuration options for HACMP/XD for Metro Mirror are available from the Define PPRC Replicated Resources panel.

To define PPRC replicated resources:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resources Configuration > Configure PPRC Replicated Resources > Define ESS PPRC Replicated Resources and press Enter.From this panel you can:• Define ESS disk subsystems.• Define a PPRC replicated resource.



62

4

• Define PPRC tasks.• Synchronize the PPRC configuration.• Verify the PPRC configuration.

Defining the ESS Disk Subsystems to HACMP

You define the ESS subsystems included in the sites that support HACMP/XD for Metro Mirror to enable HACMP to process fallovers for PPRC replicated resources.

Consult your completed ESS Disk Subsystem Worksheet for the values planned for your cluster.

To define an ESS system to HACMP:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Configuration > Define PPRC Replicated Resources > Define ESS Disk Subsystem > Add an ESS Disk Subsystem and press Enter.

3. In the Add an ESS Disk Subsystem panel, enter field values as follows:

4. Press Enter.

5. Review the settings for the ESS subsystem.From the Define ESS Disk Subsystem panel, select Change/Show an ESS Disk Subsystem and select an ESS system.

6. Make changes to the configuration settings as needed.

Defining the PPRC Pairs to HACMP

You define the PPRC pairs, the primary volume to secondary volume mappings, to allow the HACMP/XD for Metro Mirror to manage them. A volume group with multiple disks should have been defined with a single set of tasks in the section Configuring PPRC Tasks. Define one PPRC pair for each mirrored volume in the volume group.

ESS Subsystem Name The name that identifies the ESS subsystem to HACMP. The name may consist of up to 32 alphanumeric characters, and may include underscores.

ESS site name The name of the site where the ESS resides. The site is already defined in HACMP for the site name to be available from the picklist.

ESS IP Address The IP address, in dotted decimal notation, that the Copy Services Server for an ESS uses.Note: The IP address specified here is the ESSNet address of the ESS at this site.

ESS User ID The user ID used to authenticate logging into the ESS.

ESS password The password associated with the specified ESS User ID.



Consult your completed PPRC-Mirrored Volumes Worksheet for the values planned for your cluster.

To define a PPRC Pair to HACMP/XD for Metro Mirror:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Configuration > Define PPRC Replicated Resources > Define a PPRC Replicated Resource > Add a PPRC Resource and press Enter.

3. In the Add a PPRC Resource panel, enter field values as follows:

PPRC Resource Name The name of the PPRC volume pair. The name may consist of up to 32 alphanumeric characters, and may include underscores.For example, a PPRC replicated resource could be named pprc4.1, with pprc signifying the type of replicated resource, 4 identifying the volume group to which the PPRC-protected volume belongs, and 1 enumerating a PPRC-protected volume in the volume group. Another PPRC replicated resource that includes a different volume in the same volume group could be named pprc4.2.

HACMP Sites The two HACMP sites, with the primary site for this PPRC pair listed first. The sites are already defined in HACMP. For information about defining sites, see the section Configuring the HACMP/XD Metro Mirror Cluster for Direct Management.

Primary Site ESS Info The identifier for the volume in the primary ESS in the form:volume_id@ESS_namewhere volume_id is the ESS volume ID of the disk, and ess_name is the name of the ESS subsystem containing this volume.The primary ESS is the ESS at the site that you chose as primary for this PPRC replicated resource.Consult the PPRC-Mirrored Volumes Worksheet for the volume ID, and the ESS Disk Subsystem Worksheet for the ESS name.For information about determining a volume ID, see the section Completing the PPRC-Mirrored Volumes Worksheet.



64

4

4. Press Enter.

5. Review the settings for the PPRC pairs. From the Define a PPRC Replicated Resource panel, select Change/Show a PPRC Resource and select a PPRC pair.


Defining the PPRC Tasks to HACMP/XD for Metro Mirror for Direct Management

You define PPRC tasks to HACMP/XD for Metro Mirror only in cases where you named the PPRC tasks differently than recommended in the section Configuring PPRC Tasks. If you used the recommended naming convention, you can skip this section.

Defining PPRC Path TasksIf you specified another name for a PPRC path task in ESS Remote Copy Services, identify that task to HACMP/XD for Metro Mirror.

Consult your completed User-Specific Tasks Worksheets for the values planned for your cluster.

To define a user-specific PPRC path task name to HACMP/XD for Metro Mirror:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Configuration > Define PPRC Replicated Resource > Define PPRC Tasks > Define PPRC Path Tasks > Add a Group of PPRC Path Tasks and press Enter.

3. In the Add a Group of PPRC Path Tasks panel, enter field values as follows:

Secondary Site ESS Info The identifier for a volume in the secondary ESS in the form:volume_id@ESS_namewhere volume_id is the ESS volume ID of the disk, and ess_name is the name of the ESS subsystem containing this volume.The secondary ESS is the ESS at the site that you chose as secondary for this PPRC replicated resource.Consult the PPRC-Mirrored Volumes Worksheet for the volume ID, and the ESS Disk Subsystem Worksheet for the ESS name.For information about determining a volume ID, see the section Completing the PPRC-Mirrored Volumes Worksheet.

Volume Group Name The name of the volume group associated with the PPRC path task.

Establish Path Pri - Sec The name of the task that establishes a PPRC path from the primary ESS to the secondary ESS as defined in ESS Copy Services.



4. Press Enter.

5. Review the settings for the PPRC path tasks, from the Define PPRC Path Tasks panel, select Change/Show a Group of PPRC Path Tasks.


Defining PPRC Pair TasksIf you specified another name for a PPRC pair task in ESS Remote Copy Services, identify that task to HACMP/XD for Metro Mirror.

Consult your completed User-Specific Tasks Worksheets for the values planned for your cluster.

Note: When you specify a user-specific task name, specify task names for all eighteen tasks (whether or not they are different from the recommended names) for each volume group.

To define a group of user-specific PPRC pair task names to HACMP/XD for Metro Mirror:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > Define PPRC Tasks > Define PPRC Pair Tasks > Add a Group of PPRC Pair Tasks and press Enter.

3. In the Add a Group of PPRC Pair Tasks panel, enter field values as follows:

Delete Path Pri - Sec The name of the task that deletes a PPRC path from the primary ESS to the secondary ESS as defined in ESS Copy Services.

Delete Path Pri - Sec FORCED

The name of the task that deletes a PPRC path from the primary ESS to the secondary ESS with the Forced option as defined in ESS Copy Services.

Establish Path Pri - Sec The name of the task that establishes a PPRC path from the secondary ESS to the primary ESS as defined in ESS Copy Services.

Delete Path Pri - Sec The name of the task that deletes a PPRC path from the secondary ESS to the primary ESS as defined in ESS Copy Services.

Delete Path Pri - Sec FORCED

The name of the task that deletes a PPRC path from the secondary ESS to the primary ESS with the Forced option as defined in ESS Copy Services.

Volume Group Name The name of the volume group associated with the PPRC pair task.

Direction of the operation: Primary to Secondary



66

4

Establish Pair Pri - Sec NO COPY

The name of the task that establishes a PPRC pair from the primary ESS to the secondary ESS with the No Copy option set as defined in ESS Copy Services.

Establish Pair Pri - Sec FULL COPY

The name of the task that establishes a PPRC pair from the primary ESS to the secondary ESS with the Full Copy option set as defined in ESS Copy Services.

Establish Pair Pri - Sec RESYNC

The name of the task that establishes a PPRC pair from the primary ESS to the secondary ESS with the Copy Out-of-sync Cylinders Only option set as defined in ESS Copy Services.

Establish Pair Pri - Sec FAILOVER

The name of the task that establishes a PPRC pair from the primary ESS to the secondary ESS with the Failover option set as defined in ESS Copy Services.

Establish Pair Pri - Sec FAILBACK

The name of the task that establishes a PPRC pair from the primary ESS to the secondary ESS with the Failback option set as defined in ESS Copy Services.

Suspend Pair Pri - Sec on Pri The name of the task executed on the primary ESS that suspends PPRC mirroring from the primary ESS to the secondary ESS as defined in ESS Copy Services.

Suspend Pair Pri - Sec on Sec The name of the task executed on the secondary ESS that suspends PPRC mirroring from the primary ESS to the secondary ESS as defined in ESS Copy Services.

Terminate Pair Pri - Sec on Pri The name of the task executed on the primary ESS that ends PPRC mirroring from primary ESS to secondary ESS as defined in ESS Copy Services.

Terminate Pair Pri - Sec on Sec The name of the task executed on the secondary ESS that ends PPRC mirroring from primary ESS to secondary ESS as defined in ESS Copy Services.

Direction of the operation: Secondary to Primary

Establish Pair Sec - Pri NO COPY

The name of the task that establishes a PPRC pair from the secondary ESS to the primary ESS with the No Copy option set as defined in ESS Copy Services.

Establish Pair Sec - Pri FULL COPY

The name of the task that establishes a PPRC pair from the secondary ESS to the primary ESS with the Full Copy option set as defined in ESS Copy Services.

Establish Pair Sec - Pri RESYNC

The name of the task that establishes a PPRC pair from the secondary ESS to the primary ESS with the Copy Out-of-sync Cylinders Only option set as defined in ESS Copy Services.



4. Press Enter.

5. Review the settings for the PPRC pair tasks, from the Define PPRC Pair Tasks panel, select Change/Show a Group of PPRC Pair Tasks.


Improving the Performance of Volume GroupsDuring resource acquisition, PPRC automatically creates LUN ID mappings. If your cluster contains a large number of volume groups, you can save time by creating these mapping files manually before bringing up the cluster.

To create these mapping files, run the /usr/es/sbin/cluster/pprc/utils/cl_store_LUNPairs command on each cluster node, and pass the names of your volume groups as parameters as shown in the following example: cl_store_LUNPairs MyVg1 MyVg2 MyVg3

This populates the /tmp directory with files named VolumePairs.VolumeGroupName, where VolumeGroupName corresponds with the volume group names you specified as parameters (for example, MyVg1, MyVg2, and so on). Each file contains pairs of LUN IDs that resemble the following:

Establish Pair Sec - Pri FAILOVER

The name of the task that establishes a PPRC pair from the secondary ESS to the primary ESS with the Failover option set as defined in ESS Copy Services.

Establish Pair Sec - Pri FAILBACK

The name of the task that establishes a PPRC pair from the secondary ESS to the primary ESS with the Failback option set as defined in ESS Copy Services.

Suspend Pair Sec - Pri on Sec The name of the task executed on the secondary ESS that suspends PPRC mirroring from the secondary ESS to the primary ESS as defined in ESS Copy Services.

Suspend Pair Sec - Pri on Pri The name of the task executed on the primary ESS that suspends PPRC mirroring from the secondary ESS to the primary ESS as defined in ESS Copy Services.

Terminate Pair Sec - Pri on Sec The name of the task executed on the secondary ESS that ends PPRC mirroring from the secondary ESS to the primary ESS as defined in ESS Copy Services.

Terminate Pair Sec - Pri on Pri The name of the task executed on the primary ESS that ends PPRC mirroring from the secondary ESS to the primary ESS as defined in ESS Copy Services.

PRIMARY_LUN_ID SECONDARY_LUN_ID

50122221 50122225

50222221 50222225



68

4

Verifying and Synchronizing the PPRC ConfigurationThe configuration changes that you have completed to this point need to be synchronized to the other cluster nodes.

Verifying configuration for PPRC replicated resources checks the configuration and reports on the following issues:• HACMP/XD for Metro Mirror classes in the HACMP Configuration Database are identical

on all nodes.• The PPRC Command Line Interface is installed correctly on each node.• The PPRC volume groups are not in concurrent volume groups that span nodes on different

sites.• Sites defined in the HACMP disk subsystem exist in an HACMP site.• The IP addresses of the ESS systems exist and are reachable.• The ESS systems for the PPRC replicated resources are defined as ESS disk subsystems to

HACMP.• The two volumes of a PPRC pair are on different ESS systems and on different HACMP

sites.• The volume IDs correspond to physical volumes defined to HACMP cluster nodes.• The PVIDs of the disks in a volume group at each end of a pair are the same.

Typically, you synchronize the HACMP/XD for Metro Mirror configuration with HACMP cluster configuration. For information about verifying and synchronizing a cluster, see Chapter 7: Verifying and Synchronizing Cluster Configuration in the Administration Guide.

You can also verify and synchronize only HACMP/XD for Metro Mirror configuration.

To verify and synchronize HACMP/XD for Metro Mirror configuration:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Configuration > Define PPRC Replicated Resources > and press Enter.

3. Select:• Verify PPRC Configuration

then• Synchronize PPRC Configuration

To synchronize HACMP/XD for Metro Mirror changes from the command line interface, you can use the cl_sync_pprc_config command.

To verify HACMP/XD for Metro Mirror changes from the command line interface, you can use the cl_verify_pprc_config command.

Configuring Resource GroupsAfter defining PPRC replicated resources, you can add them to a resource group.



Configure resource groups following the procedures in the chapter on Configuring HACMP Resource Groups (Extended) in the Administration Guide. Make sure that you understand site support for HACMP resource groups, see Chapter 6: Planning Resource Groups in the Planning Guide.

When you configure a resource group ensure that:• The site policy is set to Prefer Primary Site or Online on Either Site.• A Startup policy other than Online on All Available Nodes is specified.• The Resource Group Processing Ordering is set to serial.

Consult your completed Resource Groups for PPRC Replicated Resources Worksheet for the values planned for your cluster.

To add a PPRC replicated resource to a resource group:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Group Configuration > Change/Show Resources and Attributes for a Resource Group and press Enter.

3. In the Change/Show Resources and Attributes for a Resource Group panel specify:• The name of the PPRC replicated resource in the PPRC Replicated Resources field.• The volume groups associated with the individual PPRC replicated resources.

4. Verify and synchronize the cluster.For information about synchronizing a cluster, see Chapter 7: Verifying and Synchronizing a Cluster Configuration in the Administration Guide.

Starting the ClusterAfter verifying and synchronizing the HACMP/XD for Metro Mirror configuration changes, start the HACMP cluster. All PPRC pairs must be in a simplex-none-simplex-none state at the initial cluster startup. This means that no relationship exists between the disk volumes at cluster startup. To view and modify the state of the PPRC pairs, use ESS Copy Services.

For information about ESS Copy Services, see the ESS Web Interface User’s Guide on the IBM website at the following URL:


Changing PPRC Replicated Resources ConfigurationUse SMIT to change configuration for PPRC replicated resources.

Note: Changing resource configuration requires that HACMP services be stopped on all nodes at both sites in the cluster.

Any configuration changes you make to any of the following components effects the others listed:• Sites• PPRC replicated resources




70

4

• Volumes• Resource groups.

After you make a configuration change, verify and synchronize the configuration.

Changing the Configuration for Sites To change site configuration in HACMP:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Topology Configuration > Configure HACMP Sites > Change/Show a Site and press Enter.

For information about field values, see the section Configuring the HACMP/XD Metro Mirror Cluster for Direct Management.

Changing the Configuration for PPRC Replicated ResourcesTo change the configuration for PPRC replicated resources:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Configuration > Define PPRC Replicated Resources and press Enter.From this panel, select:• Define ESS Disk Subsystem

For information about field values, see the section Defining the ESS Disk Subsystems to HACMP.

• Define a PPRC Replicated ResourceFor information about field values, see the section Defining the PPRC Pairs to HACMP.

• Define PPRC TasksFor information about field values, see the section Defining the PPRC Tasks to HACMP/XD for Metro Mirror for Direct Management.

• Synchronize PPRC Configuration• Verify PPRC Configuration

3. After selecting a configuration option, select the Change/Show option for the value you want to change.


Chapter 5: HACMP/XD for Metro Mirror with DSCLI Management

This chapter describes the planning, installation and configuration tasks for HACMP/XD for Metro Mirror with DSCLI Management, from here on referred to as DSCLI Management. DSCLI Management simplifies how you manage PPRC replicated resources on IBM TotalStorage systems and how you can integrate PPRC replicated resources into an HACMP configuration.

You are not required to define tasks on the ESS Web Interface when you use this management system.

This chapter contains the following sections:• Overview of the DSCLI Management System• Planning for DSCLI Management• Installing the DSCLI Management for PPRC Filesets• Configuring HACMP/XD for MetroMirror Using the DSCLI• Verifying the DSCLI-Managed PPRC Configuration• Synchronizing the Cluster• Starting the Cluster• Changing the DSCLI-Managed PPRC Replicated Resources Configuration.• Plan which resource groups will contain the DSCLI-managed PPRC replicated resources

(if not already done in Chapter 2).

Overview of the DSCLI Management SystemDSCLI management provides the ability to execute Copy Services operations directly without relying on previously saved GUI tasks. The software dynamically manages DSCLI PPRC-controlled disks, providing a fully automated, highly available disaster recovery management solution.

The HACMP interface is designed to communicate with the DSCLI so that once the basic PPRC environment is configured, PPRC relationships are created automatically and there is no need for manual access to the DSCLI.

Integration of DSCLI and HACMP provides:• Support for the Subsystem Device Driver (SDD) (for the IBM ESS 2107 system).• Support for either Prefer Primary Site or Online on Either Site inter-site management

policies.• Flexible user-customizable resource group policies.• Support for cluster verification and synchronization.


HACMP/XD for Metro Mirror with DSCLI Management Planning for DSCLI Management

72

5

• Limited support for the HACMP Cluster Single Point Of Control (C-SPOC). See Installing the DSCLI Management for PPRC Filesets.

• Automatic fallover/reintegration of server nodes attached to pairs of PPRC-protected disk subsystem within and across sites. See Installing the DSCLI Management for PPRC Filesets.

• Management of PPRC for:• Fallover/failback of PPRC paths and instances for automatic movement of

PPRC-protected disks between HACMP sites.• Automatic fallover of PPRC-protected volume groups between nodes within a site. See

Installing the DSCLI Management for PPRC Filesets.

Using DSCLI allows HACMP to:• Automatically set up PPRC paths and instances that HACMP will manage.• Manage switching the direction of the PPRC relationships when a site failure occurs, so that

the backup site is able to take control of the HACMP-managed resource groups from the primary site.

Planning for DSCLI ManagementYou should be familiar with the planning tasks for HACMP. For information about planning HACMP clusters, see the Planning Guide. It is assumed at this point that:• HACMP Sites have been planned.• Basic DSCLI (and ESS CLI, if required for ESS storage,) support has been completely

configured. Refer to the appropriate documentation on how to install and configure each.

To plan for a DSCLI Management in an HACMP cluster, complete the following tasks:• Identify the Copy Services Servers (CSS) to be used.• Identify the disk subsystems to be used in the cluster.

• Identify the vpaths to be used in the configuration, including the volume IDs for each that corresponds to the storage unit and LUN.

• Identify the PPRC Replicated Resources to be used• Identify the Port Pairs to be used for PPRC Paths• Identify Volume Pairs (LUNs) • Identify the volume groups to be managed by PPRC Replicated Resources

• Plan which resource groups will contain the DSCLI-managed PPRC replicated resources (if not already done in the General Planning section).

Limitations and Restrictions for DSCLI Management The current release of HACMP/XD for Metro Mirror with DSCLI Management has the following limitations and restrictions. Check the IBM website for the latest information on TotalStorage models and HACMP support.

http://www.ibm.com/servers/storage/disk/index.html


http://www-1.ibm.com/servers/storage/disk/index.html

HACMP/XD for Metro Mirror with DSCLI Management Planning for DSCLI Management 5

Refer to the README packaged with the DSCLI Management filesets for the most up-to-date Limitations and Restrictions.

Volume Group Limitations1. A volume group must have the same volume major number across all cluster nodes. (It has

been known to cause problems during cluster function time, and it is not guaranteed to be fixed during cluster verification.)

2. Resource groups to be managed by HACMP cannot contain volume groups with both PPRC-protected and non PPRC-protected disks. For example:• VALID: RG1 contains VG1 and VG2, both PPRC-protected disks. • INVALID: RG2 contains VG3 and VG4, VG3 is PPRC-protected, and VG4 is not.• INVALID: RG3 contains VG5, which includes both PPRC-protected and

non-protected disks within the same volume group

3. Only non-concurrent volume groups can be used with PPRC.

Managed Resource Limitations

Resource groups cannot manage both DSCLI and Direct Management (ESS CLI)-managed PPRC resources simultaneously:

Note: ESS Storage resources (LSS and LUNs) are considered DSCLI PPRC Resources in this type of configuration because they are managed via the DSCLI interface and not the ESS CLI.

IBM TotalStorage Copy Services Functions Limitations

Only IBM TotalStorage Copy Services function Synchronous PPRC (Metro Mirror) is supported (No Global Copy or Global Mirror).

C-SPOC Limitations

C-SPOC operations on nodes at the same site as the source volumes successfully perform all tasks supported in HACMP 5.4.

C-SPOC operations will not succeed on nodes at the remote site (that contain the target volumes) for the following LVM operations:• Creating a volume group• Operations that require nodes at the target site to write to the target volumes (for example,

changing filesystem size, changing mount point, adding LVM mirrors) cause an error message in CSPOC. However, nodes on the same site as the source volumes can successfully perform these tasks. The changes are subsequently propagated to the other site via lazy update.

• For C-SPOC operations to work on all other LVM operations, it is highly recommended that you perform all C-SPOC operations when the cluster is active on all HACMP nodes and the underlying SVC consistency groups are in a consistent_synchronized state.

SNMP Notification Events Limitation

HACMP does not trap SNMP notification events.



74

5

Sample ConfigurationYou can set up a mutual recovery configuration in which each site acts as a production site with the other site acting as an associated backup site. Implementing a mutual recovery configuration requires:• Two HACMP sites (the same as a single recovery configuration)• Two resource groups.

A standard configuration includes two HACMP sites, each comprised of nodes attached to two PPRC-managed groups of ESSes spread across both sites.

A CLI client (ESSNI client) must be installed on the HACMP hosts. The ESSNI client is the interface between the application that intends to invoke PPRC commands and the ESSNI or HMC; it must be installed on all the ESSNI server nodes. PPRC services are invoked using the DSCLI within the ESSNI clients. The ESSNI client communicates with the ESSNI server.

The ESSNI Server runs in an HMC for the ESS 2107 and on the management server for ESS 1750. It runs directly on the ESS clusters for the 2105. The ESS Server, in turn, communicates the CLI commands to the ESS disk controllers.

A PPRC replicated resource contains the ESS disk volume pairs information. The HACMP resource group definition includes the volume groups built on top of the PPRC replicated volumes. HACMP manages PPRC processing by dynamically executing the DSCLI commands. You no longer have to define tasks on the ESS Web Interface.

The example shows a typical implementation of two ESS model 2107s with PPRC in a four-node HACMP geographic cluster. The cluster consists of four pSeries nodes. Each ESS is connected to each node (server) via a SCSI or Fibre Channel connection. Two PPRC links (ESCON or FC) between the ESSes provide the basic level of redundancy.

One link carries data from the source Logical Subsystem (LSS) on the Primary site to the target LSS on the Secondary site, and the other link carries data in the other direction (source is on the Secondary site, target is on the Primary site).

The configuration also includes point-to-point networks for heartbeating to connect the cluster nodes.



Four-Node HACMP/XD PPRC Configuration with ESS 2107 Disk Subsystems Managed by DSCLI

Example Listing for DSCLI-Managed Configuration for Mutual Takeover

An example of the configuration information for a DSCLI-managed mutual takeover configuration:DS Subsystems-------------m222:

Cluster1 IP: 9.22.22.22 Cluster2 IP: 9.44.44.44ESS Storage ID: IBM.2107-2222222Associated CS Server: m222h

m555: Cluster1 IP: 9.55.55.55 Cluster2 IP: 9.77.77.77ESS Storage ID: IBM.2107-5555555Associated CS Server: m555h

Copy Services Servers---------------------m222h: IP address 9.112.112.2m555h: IP address 9.115.115.2



76

5

Available IO ports------------------m222: I0002, I0012m555: I0005, I0015

Volumes (LUNs) to be used-------------------------m222: 1200, 1201m555: 1200, 1201

LSS to be used--------------m222: 12 m555: 12

DSCLI-Managed PPRC Replicated Resource for Resource Group RG1---------------------------------------PPRC Resource Name: sample_res1HACMP Sites: SiteA SiteBVolume Pairs: 1200->1200ESS Pair: m222 m555LSS Pair: 12 12PPRC Type: mmirPRI-SEC PortPairIDs 0002->I0005SEC-PRI PortPairIDs I0015->I0012PPRC Link Type: fcpPPRC Recovery Action AUTOMATEDVolume Group sample_VG1

DSCLI-Managed PPRC Replicated Resource for Resource Group RG2---------------------------------------PPRC Resource Name: sample_res2HACMP Sites: SiteB SiteAVolume Pairs: 1201->1201ESS Pair: m555 m222LSS Pair: 12 12PPRC Type: mmirPRI-SEC PortPairIDs I0005->I0002SEC-PRI PortPairIDs I0012->I0015PPRC Link Type: fcpPPRC Recovery Action AUTOMATEDVolume Group sample_VG2

Note that the definitions for RG2 of volume pairs, ESS Pair, LSS Pair and the PortPairIDs are listed in reference to the source of the PPRC instance that will be used by the resource: RG2 is intended to come ONLINE with the HACMP secondary site as the “source.” RG1 is intended to come ONLINE with the HACMP primary site as “source.”

From this point, resource groups RG1 and RG2 will be configured to include the DSCLI-Managed PPRC Replicated Resources sample_res1 and sample_res2 respectively.

Planning Worksheets for DSCLI ManagementIf you have not already done so, use the planning worksheets in Appendix A: General HACMP Planning Worksheets to plan HACMP sites necessary for this configuration.



For this section, use the planning worksheets in Appendix C: Planning Worksheets for DSCLI Management to record the configuration information. You can use this information when making HACMP configuration entries in SMIT. Sections in this chapter describe the information to provide on these worksheets.

Setting Up Volume Groups and File Systems on DSCLI-Protected DisksIt is suggested, although not required, that you complete these steps prior to planning. These steps must be completed prior to the initial HACMP verification to avoid verification errors.

1. Make sure that the hdisks and corresponding vpaths made available to your nodes are visible at those nodes. If they are not, and you can verify that the nodes have been cabled and configured correctly to make the vpaths available, reboot the node and run cfgmgr to make the disks viewable.

2. Based on the LUNs you have selected for a given PPRC relationship, determine which vpaths and hdisks correspond. Use the following utility on a node at both the primary and backup HACMP sites:/usr/sbin/lsvpcfg

Note: It is likely (although not required) that the LUNS will be different at each site, and it is also likely (although not required) that the vpaths will be different on each node in each site.

The output will look something like this: (this output would be from an HACMP node in siteA of the example configuration. Assume more than one type of storage unit is connected to this node.)smithers) /usr/sbin/lsvpcfg

vpath12 (Avail ) 13AABKK1602 = hdisk14 (Avail ) hdisk44 (Avail ) hdisk74 (Avail ) hdisk104 (Avail )vpath13 (Avail ) 13AABKK1603 = hdisk15 (Avail ) hdisk45 (Avail ) hdisk75 (Avail ) hdisk105 (Avail )vpath14 (Avail ) 13AABKK1604 = hdisk16 (Avail ) hdisk46 (Avail ) hdisk76 (Avail ) hdisk106 (Avail )vpath15 (Avail ) 22222221100 = hdisk17 (Avail ) hdisk47 (Avail ) hdisk77 (Avail ) hdisk107 (Avail )vpath16 (Avail pv sample_VG1) 22222221200 = hdisk18 (Avail ) hdisk48 (Avail ) hdisk78 (Avail ) hdisk108 (Avail )vpath17 (Avail pv sample_VG2) 22222221201 = hdisk19 (Avail ) hdisk49 (Avail ) hdisk79 (Avail ) hdisk109 (Avail )pilot> lshostvol.sh

The third column in this output corresponds to the storage system ID and LUN associated with the vpath listed in column one.

Examples

vpath12 above (which has no volume group created on it yet) is on storage system IBM.XXXX-13AABKK, LSS 16, LUN 002.

vpath17 (which has volume group sample_VG2 created on it) is on the same system, with a different LSS/LUN: IBM.XXXX-2222222, LSS 12, LUN 001.

To fill in the 'XXXX' from the system ID above: either from your own documentation, or use lsdev -Ccdisk | grep <hdisk associated with the vpath in question> to display the underlying disk type: IBM FC 1750 or IBM FC 2107.

Example:

smithers) /usr/sbin/lsdev -Ccdisk | grep hdisk14

hdisk14 Available 2A-08-02 IBM FC 1750



78

5

You see that vpath12 is on IBM.1750-13AABKK, LSS 16, LUN 002.

3. Create volume groups and file systems at the Primary HACMP site.

a. On one of the nodes at the Primary HACMP site, on the Vpaths) that correspond to the volume pairs for a given PPRC Relationship, set up the volume group(s) and file system(s) to be managed by HACMP. Ensure that the Volume Major Numbers for the volume groups can be used on all HACMP cluster nodes, and that the physical volume name for the filesystem can also be used across all HACMP cluster nodes.

Note: Use the /usr/sbin/lvlstmajor command on each node in the cluster to list the available volume major numbers and select a free number on all nodes.

b. After successfully creating all necessary volume groups and file systems on the first node, import the data to all other nodes at the same site.

4. Use PPRC to mirror disk to the backup HACMP site

5. Create a temporary PPRC relationship in order to copy the volume group/file set information to the remote disk.Run these commands to set up the PPRC path and instance, and to copy over the local disk information (refer to the DSCLI documentation, or run dscli help <command> for more details):

a. mkpprcpath/opt/ibm/dscli/dscli -user <userid> -passwd <password> -hmc1 <local hmc name> mkpprcpath -dev <local storage device ID> -remotedev <remote storage device ID> -srclss -tgtlss -remotewwnn <WWNN> <local port>:<remote port>

Example, where IBM.2107-2222222 is at the Primary HACMP site, IBM.2107-5555555 is at the Backup HACMP site: /opt/ibm/dscli/dscli -user sluggo -passwd batterup -hmc1 m222h mkpprcpath -dev IBM.2107-2222222 -remotedev IBM.2107-5555555 -srclss 12 -tgtlss 13 -remotewwnn 6005076303FFC354 I0002:I0005

b. mkpprc/opt/ibm/dscli/dscli -user <userid> -passwd <password> -hmc1 <local hmc name> mkpprc -dev <local storage device ID> -remotedev <remote storage device ID> -type <mmir| -mode full <local LUN>:<remote LUN>

Example:/opt/ibm/dscli/dscli -user sluggo -passwd batterup -hmc1 m222h mkpprc -dev IBM.2107-2222222 -remotedev IBM.2107-5555555 -type mmir -mode full 1002:1002

At this point, you should have a PPRC instance available, and in a copying state:/opt/ibm/dscli/dscli -user <userid> -passwd <password> -hmc1 <local hmc name> lspprc -dev <local storage device ID> -remotedev <remote storage device ID> <local LUN>:<remote LUN>

c. rmpprcOnce the PPRC relationship has completed copying, delete the relationship (if you do not, then the Backup HACMP site nodes will not have write access to the LUNs, and so you will not be able to import the new volume group(s):



/opt/ibm/dscli/dscli -user <userid> -passwd <password> -hmc1 <local hmc name> rmpprc -quiet -dev <local storage device ID> -remotedev <remote storage device ID> <local LUN>:<remote LUN>

This step is necessary in order for the next LVM operations to complete successfully.

d. Using SMIT or the command line on the backup HACMP site (the site that is connected to the remote disk subsystem), import the volume group(s) created in step b.) At this point, the volume groups and filesystems necessary to configure HACMP have been created.

Planning Copy Services Servers (CSS)

The first step in configuring DSCLI Management is to plan which CSS systems will interact with which HACMP sites, and record some basic information about the CSS systems. (It is assumed at this point that HACMP sites have already been planned and/or defined.)

Use the Copy Services Server Worksheet in Appendix C: Planning Worksheets for DSCLI Management to complete this next section. The table is constructed so that CSS can be easily identified compared to the HACMP cluster they will communicate with primarily.

Record the following information, one entry per CSS (or HMC, SMC, whichever is appropriate for your configuration):

Planning DS/ESS Disk Subsystems

In the worksheet, plan which Disk Subsystems will interact with the HACMP sites. (It is assumed that HACMP sites have already been planned and/or defined.)

It is suggested that you do this step prior to setting up any volume groups, as it may assist in keeping things straight during that process.

CSS site name Name of the HACMP site where the CSS resides. The site should already be defined in HACMP.

CSS Subsystem Name Short name for the system hosting CSS. This is the host name of the CSS or controller the services reside on.

CLI Type For DSCLI Management configurations, this entry must be DSCLI. Note: This field toggles but the second entry, ESSCLI is not applicable for this configuration; it is used for Direct Management configurations.

CSS IP Address IP address, in dotted decimal notation, of the system hosting CSS; it is used by HACMP to submit DSCLI commands

CSS User ID User ID used to authenticate logging into the CSS; necessary to complete ADSCLI commands.

CSS password Password associated with the specified CSS User ID, necessary to complete ADSCLI commands.



80

5

Use the DS/ESS Disk Subsystem Worksheet in Appendix C: Planning Worksheets for DSCLI Management to complete this section

1. Record the HACMP Sites on a copy of the worksheet.

2. Record the following information on a copy of the worksheet, one entry per storage system to be configured. The worksheet is set up so that storage systems sharing PPRC connections can be written up together. (Optional)

Note: The term ESS found on the worksheets also refers to DS systems for this context.:

3. Repeat for all disk subsystems to be included in the configuration.

Planning Primary and Secondary Site Layout for DSCLI-Managed PPRC Replicated Resources in HACMP Resource Groups

In order for a resource group to come ONLINE correctly at cluster startup time, the definition of the Primary site must be the same for the PPRC replicated resource and the resource group that contains that resource.

Defining the Primary Site for a DSCLI-Managed PPRC Replicated ResourceThe order that data is entered into the SMIT panel fields when defining a PPRC Replicated Resource indicates which site will be primary. The following entries are order-sensitive. In all cases (unless otherwise specified), the first entry corresponds to information for the primary site.• HACMP Sites• PPRC Volume Pairs

ESS site name Name of the HACMP site where this storage system will be directly connected.

ESS Cluster 1 IP Address

IP address, in dotted decimal notation of the first controller on the storage system.


IP address, in dotted decimal notation of the second controller on the storage system.

ESS User ID User ID used to authenticate logging into the ESS, if available.

ESS password Password associated with the specified ESS User ID, if available.

Full ESS Storage ID Enter the fully qualified ESS storage image ID. This includes the manufacturer, device type, model and serial numbers (MTMS). The format is: manufacture.type-model-serial number. For example: IBM.2107-921-75FA120

List of CS Servers List of CSS that will communicate with this Disk Subsystem. This information was recorded in the Copy Services Server worksheets in the previous step.



Example for a resource with multiple volume pairs: [1200->1300 1201->1301]

Primary site LUNS are 1200 & 1201. Secondary site LUNS are 1300 & 1301• ESS Pair• LSS Pair• Pri-Sec Port Pair IDs

For both Pri-Sec and Sec-Pri, the following format is correct when entering data in the SMIT panels:

[I0022->I0052 I0023->I0053]

where I0022 & I0023 correspond to Ports on the storage system directly connected to the site that will be primary for the PPRC Replicated Resource, and I0052 & I0053 are Ports on the storage system directly connected to the secondary site.• Sec-Pri Port Pair IDs

If you were going to use the same Port pairs as in the example above, the Sec-Pri listing would look like this:

[I0052->I0022 I0053->I0023]

where the Ports are on the same systems as described above.

For more examples and information, see the Sample Configuration section of this chapter.

Defining the Primary Site for an HACMP Resource GroupWhen defining the Primary and Secondary site nodes on creating an HACMP resource group, select nodes from the site where you want the resource group to come ONLINE during cluster startup as the Primary Site nodes.

Note that combining the Online on Either Site inter-site management policy with the Online on First Available Node startup policy will allow the resource group to come ONLINE on a node other than those defined as belonging to its defined Primary site.

Example:HACMP Site 1: node11, node12

HACMP Site 2: node21, node22

To define an HACMP Resource Group that will come up on Site 1:

Participating Nodes from Primary Site [node11, node12]

Participating Nodes from Secondary Site [node21,node22]

To define an HACMP Resource Group that will come up on Site 2:

Participating Nodes from Primary Site [node21, node22]

Participating Nodes from Secondary Site [node11,node12]

For more examples and information, see the Sample Configuration section of this chapter.



82

5

Planning DSCLI-Managed PPRC Replicated Resource

For this step, complete a DSCLI Replicated Resource Worksheet, found in Appendix C: Planning Worksheets for DSCLI Management. Also on this sheet, you will find a column for entering the HACMP Resource Group that will manage this Replicated Resource. This is to serve as a cross-reference for the next step in Planning.

Note: When configuring DSCLI-managed PPRC Replicated Resources, the layout of the information is pertinent to where (which HACMP site) a resource will come ONLINE during cluster startup. See Planning Primary and Secondary Site Layout for DSCLI-Managed PPRC Replicated Resources in HACMP Resource Groups and the Names of the DS or ESS Storage systems Logical Subsystems that the LUN pairs exist on. section of this chapter for more information.

For each PPRC Replicated Resource to be configured, enter the following:

PPRC Resource Name

Name that will be used by HACMP to create the DSCLI Volume Group. Use no more than 32 alphanumeric characters and underscores.

HACMP Sites Enter the name of the HACMP primary site followed by the name of the HACMP secondary site.The primary site is the one where the PPRC Replicated Resource will come ONLINE initially during cluster startup.

PPRC Volume Pairs LUN pairs that correspond to the vpaths that will support managed volume groups. The format is:Primary Volume ID:Secondary Volume ID.• All PPRC volume pairs in a PPRC replicated resource consist

of volumes from the same LSS pair.• Volumes in a PPRC replicated resource must be from the

same volume group.• A volume group can span more than one LSS.• A volume group can span more than one disk subsystem.

ESS pair Names of the DS or ESS storage systems that contain the LUNs the volume pairs correspond to. The first name in the list is the primary and the second is the secondary.

LSS Pair Names of the DS or ESS Storage systems Logical Subsystems that the LUN pairs exist on. The first name in the list is the primary LSS and the second is the secondary LSS.



PPRC Type Indicates whether the PPRC volume relationships will be Metro Mirror (Synchronous) mmir or Global Copy (Extended Distance) gep relationships. (Global Copy is not yet supported for HACMP/XD.)• Metro Mirror (Synchronous) maintains the PPRC relationship

in a consistent manner. I/O write completion status is returned to the application once the updates are committed to the target ESS.

• Global Copy (Extended Distance) maintains the PPRC relationship in a non-synchronous manner. I/O write completion status is returned to the application once the updates are committed to the source ESS. Updates to the target volume are performed at a later point in time. The original order of updates is not strictly maintained.

Pri-Sec Port Pair IDs I/O Ports that will be used for PPRC communication between the primary and secondary storage systems named in the ESS Pair entry. The source and target port must be a Fibre Channel / ESCON I/O port that is configured for point-to-point or switch fabric topology. You can have up to 8 PPRC Path Port Pair IDs defined for each pair of LSSs. Example of three Port pairs:I1A10:I2A20<space>I1A11:I2A21<space>I1A12:I2A22

Sec-Pri Port Pair IDs I/O Ports that will be used for PPRC communication between the secondary and primary storage systems named in the ESS Pair entry. The example here shows the reverse source and target order of the pri-sec pair IDs (12 > 11 instead of 11 > 12):I2A10:I1A20<space>I2A11:I1A21<space>I2A12:I1A22

PPRC Link Type Select ESCON or FCP depending on the physical connection you are using between the two storage systems listed in the ESS Pair entry.

PPRC Critical Mode This option is used to write-protect the source volume. If the last path fails between the pairs, resulting in the inability to send information to the target, the source becomes write-protected. Current updates and subsequent attempts to update the source will fail, with a unit check on S/390 or a check condition on SCSI. Values are ON (set critmod) or OFF (do not set critmod).The default is OFF.

PPRC Recovery Action

Select the HACMP recovery action for a site failure for PPRC-XD type volume pairs. MANUAL: User action required or AUTOMATED: No user action required.



84

5

Completing the Resource Group Worksheet

Complete a Resource Groups Worksheet (See Appendix A: General HACMP Planning Worksheets) for each resource group that includes DSCLI-managed PPRC replicated resources.

To identify information required to include DSCLI-managed PPRC replicated resources in a resource group, enter the following information on the Resource Groups Worksheet:

1. Record the Resource Group Name for the resource group that will contain DSCLI-managed PPRC replicated resources. The name can contain up to 32 characters and underscores, but without a leading numeric character.

2. Record the startup, fallover, and fallback policies for the resource group.

Note: PPRC enforces that disks at the secondary site are not I/O accessible. Because of this, the following Resource Group policies are not supported:

• Inter-Site Management Policies: Online on Both Sites• Startup Policy: Online Using Node Distribution Policy, Online On All Available

Nodes• Fallover Policy: Fallover Using Dynamic Node Priority

3. Record the Inter-Site Management Policy: Prefer Primary Site or Online on Either Site.

4. Record the names of the nodes in the resource group, listing the nodes in the prioritized order in which they will appear in the node list for the resource group.

5. Record the name of each DSCLI-managed PPRC replicated resource (volume set) to be included in the resource group.

6. Record the names of the volume groups associated with the volume sets to include in the resource group.

Next StepsAfter you plan the configuration for DSCLI-managed PPRC replicated resources in an HACMP cluster:

1. Install the base HACMP/XD for Metro Mirror software on HACMP cluster nodes. (Refer to Chapter 3: Installing HACMP/XD for Metro Mirror for information on installation methods and fileset locations.

2. Install the DSCLI Management filesets (see Installing the DSCLI Management for PPRC Filesets section this chapter.

3. Configure the HACMP cluster and DSCLI Management (see Configuring HACMP/XD for MetroMirror Using the DSCLI in this chapter.

Volume Group Volume group that will exist on this PPRC instance (replicated resource). The volume group can include volumes pairs from different LSSes as well as from different ESSes.


HACMP/XD for Metro Mirror with DSCLI Management Installing the DSCLI Management for PPRC Filesets 5

Installing the DSCLI Management for PPRC FilesetsYou must be logged in as root to perform installation tasks. See Chapter 3: Installing HACMP/XD for Metro Mirror for general installation instructions.

Installing Prerequisite SoftwareBefore installing HACMP/XD DSCLI Management for PPRC v.5.4 (spprc filesets), the following software must be installed on cluster nodes:

Note: Check the README for updates on versions supported after the first release of this software.

1. AIX 5.2.0 ML4 or higher with all applicable APARs or AIX 5.3.0 ML8 with all applicable APARs.

2. HACMP/XD Version 5.4.0 (cluster.es.server.rte 5.4.0.0) or higher. (Refer to Chapter 3: Installing HACMP/XD for Metro Mirror for more information)

3. IBM Subsystem Device Driver (SDD) for the storage system you are using. Check the documentation and/or website for the currently approved SDD version to use with a given microcode version. Make sure the following filesets are installed:

a. devices.fcp.disk.ibm.rte (1.0.0.0)

Note: It may not be clear during installation of the SDD filesets that this first fileset is necessary, but it is critical for correct fallover behavior.

b. devices.sdd.**.rte (latest version)

c. devices.ibm2105.rte(latest version)

Note: This fileset provides connection scripts for all of the ESS and DS disk types, including ESS 800, DS 8000 and DS 6000. Refer to the storage website, www.storage.ibm.com for more information on current levels of this driver.

4. DSCLI client software and other configuration specific prerequisites, as shipped with the microcode for the storage systems to be used in this cluster. Refer to the DSCLI documentation for more information on how to install and configure this software.

5. (Optional) If ESS storage systems are going to be included in your cluster) ESS CLI software is shipped with the microcode for the storage hardware. DSCLI Management code expects the ESS CLI to be installed in the following (non-standard) directory/opt/ibm/ibm2105cli

so you may have to create a link from the actual installation location to this location. Be mindful of this, as it will cause problems during both the verification and run-time for your cluster.


HACMP/XD for Metro Mirror with DSCLI Management Installing the DSCLI Management for PPRC Filesets

86

5

Installing the DSCLI Management FilesetsThere is a fileset dependency between the DSCLI management filesets and the base PPRC filesets for this configuration. Install these filesets on each cluster node:• cluster.es.pprc, which contains:

• cluster.es.pprc.cmds

• cluster.es.pprc.rte

• cluster.es.spprc, which contains: • cluster.es.spprc.cmds

• cluster.es.spprc.rte

• cluster.msg.pprc.en_US, which contains message sets for both filesets listed above.

Refer to Chapter 3: Installing HACMP/XD for Metro Mirror for more details on installing these and the base HACMP/XD filesets.

Installation DirectoriesAll HACMP-PPRC programs and scripts are located in the following directory and sub-directories:/usr/es/sbin/cluster/pprc/usr/es/sbin/cluster/pprc/spprc

All DSCLI programs and scripts are located in the following directory and sub-directories:/opt/ibm/dscli

All ESSCLI programs and scripts are expected to be located in the following directory:/opt/ibm/ibm2105cli

Upgrading to HACMP/XD for MetroMirror v 5.4When upgrading from a previous version of HACMP/XD for Metro Mirror, you can choose to upgrade the base pprc filesets only, or you can add the cluster.es.spprc filesets (to add DSCLI Management).

Note: Currently there is no migration path from eRCMF Management or SVC Management to DSCLI Management. The statements here apply only to Direct Management.

You can install the cluster.es.spprc filesets in your current base (Direct Management) PPRC environment and continue to operate the current environment. This is possible because the base pprc and spprc configuration information is stored in different configuration databases (ODMs).

The spprc filesets are installed automatically if you already have a previous version of them on your system when you install via smitty update_all.


HACMP/XD for Metro Mirror with DSCLI Management Configuring HACMP/XD for MetroMirror Using the DSCLI 5

Configuring HACMP/XD for MetroMirror Using the DSCLIThis section explains how to configure the DSCLI Management with HACMP.

DSCLI and HACMP Configuration RequirementsBefore configuring HACMP/XD for Metro Mirror with the DSCLI management interface, ensure that:• PPRC is configured and running on the storage systems.• ESSNI client and server software is installed (DSCLI software on all HACMP cluster

nodes, for example).• You have a good understanding of HACMP sites for PPRC replicated resources. For more

information about sites, see the section HACMP Sites in Chapter 1: Overview of HACMP/XD for Metro Mirror.

• Both the base HACMP/XD for Metro Mirror and DSCLI Management filesets are installed on each cluster node.

• The HACMP cluster is configured for:• Nodes• Sites• Networks and network interfaces• Service labels, application monitors, etc.• Initial resource groups

You can modify the attributes for a resource group later to accommodate PPRC replicated resources.

DSCLI Management Configuration StepsSteps for setting up the DSCLI Management interface:

1. Configure PPRC-Managed Replicated Resources (use the SMIT panels at the bottom of the main HACMP PPRC-Managed Replicated Resources menu):

a. Configure Copy Services Servers

b. Configure disk systems to be included

c. Configure DSCLI-Managed PPRC replicated resources.

2. Configure HACMP resource groups to include PPRC-Managed replicated resources.

Configuring DSCLI-Managed PPRC Replicated ResourcesTo define DSCLI-Managed PPRC replicated resources:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Configuration > Configure PPRC Replicated Resources and press Enter.


HACMP/XD for Metro Mirror with DSCLI Management Configuring HACMP/XD for MetroMirror Using the DSCLI

88

5

From this panel you can:• Configure Copy Services Server• Configure DS ESS disk subsystems• Configure DSCLI-Managed PPRC replicated resources.

Configuring Copy Services Server

To configure the Copy Services Server:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > HACMP Extended Resources Configuration > Configure PPRC Replicated Resources > Configure Copy Services Server > Add a Copy Services Server and press Enter.

3. In the Add a Copy Services Server panel, enter field values as follows:

4. Press Enter.

5. Repeat these steps for the CSS on the other site.

Defining the DS ESS Disk Subsystems to HACMP

You define the DS ESS subsystems included in the sites that support HACMP/XD for MetroMirror to enable HACMP to process fallovers for PPRC replicated resources.

Consult your completed ESS Disk Subsystem Worksheet for the values planned for your cluster.

To define a DS ESS system to HACMP:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > Configure PPRC Replicated Resources > Configure DS ESS Disk Subsystem > Add an ESS Disk Subsystem and press Enter.

CSS Subsystem Name Name that identifies the Copy Service Server. The name may consist of up to 32 alphanumeric characters, and may include underscores.

CSS site name Name of the HACMP site where the CSS resides. The site must already be defined in HACMP for the site name to be available from the picklist.

CLI Type Select DSCLI if you are using the ESS 2107. Select ESSCLI if you are using the ESS 2105.

CSS IP Address IP address, in dotted decimal notation, that the Copy Services Server uses. (This is different from the ESS IP address.)

CSS User ID User ID used to authenticate logging into the CSS.

CSS password Password associated with the specified CSS User ID.



3. In the Add an ESS Disk Subsystem panel, enter field values as follows:

4. Press Enter.

5. Review the settings for the ESS subsystem.From the Configure an ESS Disk Subsystem panel, select Change/Show an ESS Disk Subsystem and select an ESS system to view. Make changes if necessary and press Enter.

6. Repeat these steps to enter the information for the DS ESS at the second site.

Adding a DSCLI-Managed PPRC Replicated Resource

To add a PPRC replicated resource:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > Configure PPRC Replicated Resources > DSCLI-Managed PPRC Replicated Resource Configuration > Add a PPRC Replicated Resource and press Enter.

ESS Subsystem Name Name that identifies the ESS subsystem to HACMP. The name may consist of up to 32 alphanumeric characters, and may include underscores.

ESS site name Name of the site where the ESS resides. The site is already defined in HACMP for the site name to be available from the picklist.


IP address, in dotted decimal notation of the ESS disk subsystem in cluster 1.Note: The IP address specified here is the ESSNet address of the ESS at this site.


IP address, in dotted decimal notation of the ESS disk subsystem in cluster 2.Note: The IP address specified here is the ESSNet address of the ESS at this site.

ESS User ID User ID used to authenticate logging into the ESS, if available.

ESS password Password associated with the specified ESS User ID, if available.

Full ESS Storage ID Enter the fully qualified ESS storage image ID. This includes the manufacturer, device type, model and serial numbers (MTMS). The format is: manufacture.type-model-serial number. For example: IBM.2107-921-75FA120

List of CS Servers From the list, select the CSS that will manage the PPRC of this Disk Subsystem.


HACMP/XD for Metro Mirror with DSCLI Management Configuring HACMP/XD for MetroMirror Using the DSCLI

90

5

3. In the Add a Replicated Resource panel, enter field values as follows:

PPRC Resource Name

Enter a name of the set of PPRC volume pairs that make up the PPRC replicated resource. Use no more than 31 alphanumeric characters and underscores.

HACMP Sites Enter the names of the HACMP sites (already defined to HACMP). Enter the name of the primary site followed by the secondary site.

PPRC Volume Pairs List of PPRC volume pairs that are contained in this PPRC replicated resource. The format is:Primary Volume ID:Secondary Volume ID.• All PPRC volume pairs in a PPRC replicated resource consist

of volumes from the same LSS pair.• Volumes in a PPRC replicated resource must be from the

same volume group.• A volume group can span more than one LSS.• A volume group can span more than one ESS disk subsystem.

ESS pair Set of ESSs associated with this PPRC resource. The first name in the list is the primary ESS and the second is the secondary ESS.

LSS Pair Set of LSSs associated with this PPRC resource. The first name in the list is the primary LSS and the second is the secondary LSS.

PPRC Type Indicates whether the PPRC volume relationships will be Metro Mirror (Synchronous) mmir or Global Copy (Extended Distance) gep relationships. • Metro Mirror (Synchronous) maintains the PPRC relationship

in a consistent manner. I/O write completion status is returned to the application once the updates are committed to the target ESS.

• Global Copy (Extended Distance) maintains the PPRC relationship in a non-synchronous manner. I/O write completion status is returned to the application once the updates are committed to the source ESS. Updates to the target volume are performed at a later point in time. The original order of updates is not strictly maintained.



4. Press Enter.

5. Repeat as necessary to define more DSCLI-Managed PPRC replicated resources.

Configuring Resource GroupsAfter defining PPRC replicated resources, you can add them to a resource group.

Pri-Sec Port Pair IDs List of PPRC Path Port Pair IDs of PPRC Links between a primary LSS and secondary LSS. The source and target port must be a Fibre Channel / ESCON I/O port that is configured for point-to-point or switch fabric topology. A PPRC Path Port Pair ID consists of two Port IDs, one designated as the source port and the other as the target port for a PPRC path. The first Port ID is the designated source Port. The second LSS ID is the designated target Port. Separate the two Port IDs of a PPRC Path Port Pair ID with a colon and no whitespace.You can have up to 8 PPRC Path Port Pair IDs defined for each pair of LSSs. Use a white space to separate multiple PPRC Path Port Pair IDs. Example of three Port pairs:I1A10:I2A20<space>I1A11:I2A21<space>I1A12:I2A22

Sec-Pri Port Pair IDs Same as above. The example here shows the reverse source and target order of the pri-sec pair IDs (12 > 11 instead of 11 > 12):I2A10:I1A20<space>I2A11:I1A21<space>I2A12:I1A22

PPRC Link Type Select ESCON or FCP depending on the connection you are using for the PPRC path.

PPRC Critical Mode This option is used to write-protect the source volume. If the last path fails between the pairs, resulting in the inability to send information to the target, the source becomes write-protected. Current updates and subsequent attempts to update the source will fail, with a unit check on S/390 or a check condition on SCSI. Values are ON (set critmod) or OFF (do not set critmod).The default is OFF.


Select the recovery action for PPRC-XD volume pairs. MANUAL: User action required or AUTOMATED: No user action required.

Volume Group Volume group that contains the PPRC volume pairs included in this PPRC replicated resource. The volume group can include volumes pairs from different LSSes as well as from different ESSes.


HACMP/XD for Metro Mirror with DSCLI Management Verifying the DSCLI-Managed PPRC Configuration

92

5

Configure resource groups following the procedures in Chapter 4: Configuring HACMP Resource Groups (Extended) in the Administration Guide. Make sure that you understand site support for HACMP resource groups, see Chapter 6: Planning Resource Groups in the Planning Guide.

When you configure a resource group ensure that:• The site policy is set to Prefer Primary Site or Online on Either Site.• A Startup policy other than Online on All Available Nodes is specified.


To add a PPRC replicated resource to a resource group:

1. Enter smit hacmp


3. In the Change/Show Resources and Attributes for a Resource Group panel specify:• The name of the PPRC replicated resource in the PPRC Replicated Resources field.• The volume groups associated with the individual PPRC replicated resources.

The PPRC Replicated Resources entry is a picklist that displays the resource names created in the previous step. Make sure that the volume groups selected on the Resource Group configuration screen match the volume groups used in the PPRC Replicated Resource.

4. See the next section on verifying the cluster. You must verify before synchronizing the cluster.

Verifying the DSCLI-Managed PPRC ConfigurationVerifying the configuration for DSCLI-Managed PPRC replicated resources checks the configuration and reports on the following issues:• SPPRC information in the HACMP Configuration Database (ODM) is identical on all

nodes.• The DSCLI Command Line Interface is installed correctly on each node.• The PPRC volume groups are not in concurrent volume groups that span nodes on different

sites.• Sites are properly defined for the HACMP and PPRC configuration.• The IP addresses of the ESS systems exist and are reachable.• The ESS systems for the PPRC replicated resources are defined as ESS disk subsystems to

HACMP.• The two volumes of a PPRC pair are on different ESS systems and on different HACMP

sites.• The volume IDs correspond to physical volumes defined to HACMP cluster nodes.• The PVIDs of the disks in a volume group at each end of a pair are the same.


HACMP/XD for Metro Mirror with DSCLI Management Verifying the DSCLI-Managed PPRC Configuration 5

• All PPRC volumes pairs in a PPRC replicated resource can only consist of volumes from the same LSS pair.

• Volumes in a PPRC replicated resource must be from the same volume group• Correct PPRC links and their Port IDs have been defined• The defined CLI path and the ESSNI client jar file exist on all the HACMP servers• The volume pairs have volumes that do exist on the ESSs defined to HACMP.

To verify the HACMP/XD DSCLI-managed PPRC configuration:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Verification and Synchronization and press Enter.

Since the cluster is inactive, the following options appear. Make sure to select Verify in the first field (not both, you do not synchronize the configuration yet):

3. Press Enter. The verification output appears in the SMIT Command Status window.

4. If any error messages appear, make the necessary changes and run the verification procedure again. You may see Warnings if the configuration has a limitation on its availability; for example, only one interface per node per network is configured.

Verify Synchronize or Both Select Verify only.

Automatically correct errors found during verification?

No is the default. HACMP performs corrective actions without prompting you to perform any action. If you select Interactively, during verification you will be prompted when HACMP finds a problem it can correct related to the following, for example:• Importing a volume group• Exporting and re-importing shared volume groups

(mount points and filesystems issues)You then choose to have the action taken or not. For more information, see Chapter 7 of the HACMP for AIX Administration Guide.

Force synchronization if verification fails?

No is the default. If you select Yes, cluster verification runs but verification errors are ignored and the cluster is synchronized. Use the default.

Verify changes only? No is the default. (Run the full check on resource and topology configuration.) Use the default.

Logging Standard is the default. You can also select Verbose. Verification messages are logged to /var/hacmp/clverify/clverify.log.


HACMP/XD for Metro Mirror with DSCLI Management Synchronizing the Cluster

94

5

Synchronizing the ClusterTo synchronize the HACMP/XD DSCLI-managed PPRC configuration:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Verification and Synchronization and press Enter.

3. The HACMP Verification and Synchronization panel appears. Select either both or Synchronize in the first field and press Enter.

The cluster is synchronized.

Starting the ClusterVerification runs automatically at cluster startup unless you turn this option off. After completing the steps above that set volume groups to show up only as vpaths, cluster verification will fail. To avoid this issue, set the ignore verification errors? field to true on the Start Cluster Services SMIT panel:.

1. Enter the fastpath smit cl_admin

2. In SMIT, select Manage HACMP Services > Start Cluster Services.

3. Make the selections for the fields on this panel, setting the field Ignore verification errors? to true.

4. Press Enter to start cluster services.

HACMP will start up and manage the PPRC resources.

See the HACMP for AIX Administration Guide for more details on starting a cluster.

Changing the DSCLI-Managed PPRC Replicated Resources Configuration

Use SMIT to change configuration for PPRC replicated resources.


Configuration changes you make to any of the following components affects the others listed:• Sites• PPRC replicated resources• Volumes• Resource groups.

After you make a configuration change, verify and synchronize the configuration.


HACMP/XD for Metro Mirror with DSCLI Management Changing the DSCLI-Managed PPRC Replicated Resources Configuration 5

Changing the Configuration for Sites To change site configuration in HACMP:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Topology Configuration > Configure HACMP Sites > Change/Show a Site and press Enter.

For information about field values, see the section Configuring DSCLI-Managed PPRC Replicated Resources.

Changing the Configuration of DSCLI-Managed PPRC Replicated Resources

To change or remove the configuration for DSCLI-Managed PPRC replicated resources:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Configuration > Configure PPRC Replicated Resources and press Enter.From this panel, select:• Configure Copy Services Server

For information about field values, see the section Configuring Copy Services Server.• Configure DS ESS Disk Subsystem

For information about field values, see the section Defining the DS ESS Disk Subsystems to HACMP.

• Configure a DSCLI-Managed PPRC Replicated ResourceFor information about field values, see the section Adding a DSCLI-Managed PPRC Replicated Resource.

3. After selecting a configuration option, select the Change/Show option for the value you want to change, or the Remove option for the value you want to remove.


HACMP/XD for Metro Mirror with DSCLI Management Changing the DSCLI-Managed PPRC Replicated Resources Configuration

96

5


Chapter 6: HACMP/XD for Metro Mirror with SVC Management

This chapter presents information for planning, installing, and configuring an HACMP/XD: Metro Mirror for SAN Volume Controller (SVC) cluster. SVC virtualizes storage, assists in managing Peer to Peer Remote Copy (PPRC) services on TotalStorage systems and simplifies how you can integrate PPRC replicated resources into an HACMP configuration. SVC Management allows SVC PPRC resources to be managed by HACMP with a minimum of user configuration.

IBM TotalStorage SAN Volume Controller versions 1.2 and 2.1 are supported.

This chapter contains the following sections:• Overview of SVC Management• Planning for SVC Management• Installing HACMP/XD: Metro Mirror for SVC• Configuring HACMP/XD: Metro Mirror for SVC• Changing an HACMP/XD for Metro Mirror SVC Configuration• Troubleshooting HACMP/XD for Metro Mirror for SVC.

For more information on the IBM TotalStorage SAN Volume Controller, see:

http://www.redbooks.ibm.com/redbooks/pdfs/sg246423.pdf

Overview of SVC ManagementThe IBM TotalStorage SAN Volume Controller (SVC) is a virtualization appliance solution that maps virtualized volumes visible to hosts and applications to physical volumes on storage devices. It is an in-band implementation that minimizes the dependency on unique hardware and software, telecopying the storage functions expected in a SAN environment from the storage subsystems and managing storage resources. In addition, it provides advanced copy services for data migration and business continuity similar to the Peer to Peer Remote Copy function in the ESS and DS4000 (formerly named the FAStT) Storage subsystems. Since the copy services operate on the virtual volumes, dramatically simpler replication configurations can be created using the SAN Volume Controller, rather than replicating each physical volume in the managed storage pool.

HACMP/XD for Metro Mirror with SVC management enhances PPRC’s ability to provide a fully automated, highly available disaster recovery management solution by taking advantage of SVC’s ability to provide virtual disks derived from varied disk subsystems. The HACMP interface is designed so that once the basic SVC environment is configured, PPRC relationships are created automatically; no additional access to the SVC interface is needed.


http://www.redbooks.ibm.com/redbooks/pdfs/sg246423.pdf

HACMP/XD for Metro Mirror with SVC Management Planning for SVC Management

98

6

HACMP configuration information required for SVC PPRC resource management is simple; it is easily acquired from the existing SVC configuration.

The integration of HACMP and SVC PPRC provides:• HACMP management of SVC PPRC for automatic fallover/reintegration of SVC

PPRC-protected virtual disks between HACMP sites• Support for user-defined policy-based resource groups• Support for the following Inter-Site Management Policies for resource groups:

• Prefer Primary Site• Online on Either Site.

• Support for the Subsystem Device Driver (SDD) (for the 2145 (SVC) system)• Support for cluster verification and synchronization• Flexible user-customizable resource group movement policies• Limited support for C-SPOC (HACMP Single Point of Control utility)• Automatic fallover/reintegration of server nodes attached to SVC-provided virtual disk

pairs within sites• Management of automatic fallover of PPRC-protected volume groups between nodes

within a site• SVC management to switch the SVC PPRC relationships over so that the backup site can

take control of the HACMP-managed resource groups from the primary site in case of a site failure

• SVC PPRC Command Line Interface and/or GUI to manually manage SVC PPRC consistency groups and relationships.

Definition of TermsIt is important to understand how the terms master and auxiliary are used in the HACMP/XD: Metro Mirror for SVC environment. In general, master and auxiliary refer to the SVC virtual disks that are on either end of a SVC PPRC link. Primary and Secondary refer to the HACMP sites that host the resource groups that manage the SVC PPRC replicated resources that contain those SVC PPRC links.

The terms master and auxiliary can also refer to the SVC clusters themselves. In general, the Master SVC cluster is connected to the HACMP Production Site, and the Auxiliary SVC Cluster is connected to the HACMP Backup/Recovery Site.

See the Installation Guide and Glossary, as well as Chapter 1: Overview of HACMP/XD for Metro Mirror for more definitions used in HACMP and PPRC.

Planning for SVC ManagementYou should be familiar with the planning tasks for HACMP. For information about planning HACMP clusters, see the HACMP Planning Guide. It is assumed at this point that:• HACMP Sites have been planned.• Basic SVC and SVC PPRC support has been completely configured. Refer to the SVC

documentation on how to install and configure SVC and SVC PPRC support.


HACMP/XD for Metro Mirror with SVC Management Planning for SVC Management 6

• SVC Clusters and native PPRC support on those SVC Clusters has already been configured.

To plan for SVC-managed PPRC replicated resources in an HACMP cluster, complete the following planning tasks:• Identify the SVC Clusters and which HACMP sites they belong to.• Identify SVC relationships.• Identify the volume groups associated with SVC virtual disks (VDisks) to be used in the

relationships.• Identify SVC consistency groups.• Identify the relationships to be used in the consistency groups.

Note: The term 'Consistency Group Name' can be, in this instance, interchanged with the term 'PPRC Replicated Resource Name.' In the context of SVC, they are the same.

• Plan which resource groups will contain the SVC-managed PPRC replicated resources.

Limitations and Restrictions for HACMP/XD: Metro Mirror for SVCThe current release of HACMP/XD SVC PPRC has the following limitations and restrictions.• Size of the cluster. In this release, only two-node HACMP cluster configurations are fully

supported.• SVC Host Naming Convention. Although SVC Host Name Aliases are arbitrary, for

HACMP/XD SVC PPRC, they must match the node names you define for the HACMP/XD Sites. This ensures that the ssh commands used to execute SVC tasks are completed for the correct nodes.

• SSH must be installed and configured. HACMP uses commands to communicate with the SVC PPRC cluster that require ssh. Therefore, some version of ssh must be installed and configured on all SVC PPRC cluster nodes.

• Volume Groups. Resource Groups to be managed by HACMP cannot contain volume groups with both SVC PPRC-protected and non SVC-PPRC protected disks.For example:• VALID: RG1 contains VG1 and VG2, both PPRC-protected disks. • INVALID: RG2 contains VG3 and VG4, VG3 is PPRC-protected, and VG4 is not.• INVALID: RG3 contains VG5, which includes both PPRC-protected and

non-protected disks within the same volume group.• SVC PPRC Global Mirror function is not supported.• SVC Cluster names and HACMP Host aliases. The Host Aliases on your SVC Clusters

must match the node names (short names) used in HACMP to define each cluster node.• CSPOC Operations. You cannot use C-SPOC for the following LVM operations to

configure nodes at the remote site (that contain the target volumes):• Creating a volume group



10

6

• Operations that require nodes at the target site to write to the target volumes (for example, changing filesystem size, changing mount point, adding LVM mirrors) cause an error message in CSPOC. This includes functions such as changing filesystem size, changing mount points, and adding LVM mirrors. However, nodes on the same site as the source volumes can successfully perform these tasks. The changes are subsequently propagated to the other site via lazy update.For C-SPOC operations to work on all other LVM operations, it is highly recommended that you perform all C-SPOC operations when the cluster is active on all HACMP nodes and the underlying SVC consistency groups are in a consistent_synchronized state.

Sample Configuration for SVC ManagementYou can set up a mutual recovery configuration in which each site acts as a production site with the other site acting as an associated backup site. Implementing a mutual recovery configuration requires:• Two HACMP sites (the same as a single recovery configuration)• At least two resource groups.

Basic SVC PPRC Cluster Configuration and Operation

The infrastructure we consider here is a wide-area cluster system whereby the set of local cluster nodes are directly attached to local storage systems but with no access to the geographically separated remote storage systems. In a typical setup with AIX server nodes in a wide-area HACMP cluster configuration, the server nodes at each HACMP cluster site are connected directly to the local SVC Clusters. Two or more inter-cluster links are required between the two SVC clusters. The HACMP nodes at the same geographical site access the same shared volume groups, but the nodes at each site access them from different physical volumes.

The SVC PPRC maintains separate identical local copies of the application data on two separate back end storage subsystems. Several virtual disks are mirrored via master SVC cluster PPRC from the primary site to the standby site. When node or site failure occurs, access to disk resources are not passed from one node to another. Instead, all highly available applications are restarted at the standby site using data copy on secondary volumes. Under normal operation, the application is active on a server at the production site, and all updates to the application data are automatically replicated to the backup disk subsystem by the SVC PPRC framework. PPRC protects the backup copy from inadvertent modification. When a total site failure occurs, the application is restarted on a backup server at the remote site.

Prior to restarting the application, which, in this context, means not only the application that end users interact with, but also all dependent data base software or other middleware, HACMP/XD initiates SVC actions to ensure that the backup disk volumes are in the appropriate state to allow application access.

HACMP support for geographic clustering is based on the concept of replicated resources. These are defined as a resource type that has a primary and secondary instance, corresponding to the source and target of data copies that are replicated across two locations. In SVC PPRC, a PPRC consistency group, comprised of a list of SVC relationships defined to SVC and associated with an HACMP geographically separated cluster definition and defined to HACMP

0 HACMP/XD for Metro Mirror: Planning and Administration Guide


is referred to as an SVC PPRC replicated resource. The definition for an SVC PPRC replicated resource contains the VDisk name and the HACMP volume groups associated with the VDisk pairs. SVC recognizes which volumes mirror each other for each PPRC replicated resource.

In the example shown here, the HACMP production site includes:• Server A and Server B• The ESS labeled Primary ESS• A SAN Volume Controller connected by Fibre Channel PPRC links to the SVC at the

secondary site and two connections to the Primary ESS.

The HACMP recovery site includes:• Server C and Server D• The ESS labeled Secondary ESS• A SAN Volume Controller connected by Fibre Channel PPRC links to the SVC at the

primary site and two connections to the Primary ESS.

The HACMP resource group contains:• The four server nodes• IP networks to connect the server nodes• One or more shared volume groups• The PPRC replicated resource associated with the volumes in the volume group.

The configuration also includes point-to-point networks for heartbeating to connect the cluster nodes.

Note that all nodes also require connections to the SVC. At least one storage controller is connected to each SVC cluster via a SCSI or Fiber Channel connection.



10

6

Four-Node HACMP/XD SVC PPRC Configuration with a Pair of SVC-managed ESS Disk Subsystems

Example HACMP/XD: Metro Mirror for SVC Mutual Takeover Configuration

The example lists the information for a two-node SVC PPRC cluster configured for mutual takeover.SVC Clusters------------svc_Cl1, svc_Cl2

SVC PPRC Relationships Configuration -------------------------------------Relationship Name = sample_rel1Master Vdisk info = volume_id1@svc_Cl1Auxiliary Vdisk info = volume_id2@svc_Cl2

SVC PPRC Relationships Configuration -------------------------------------Relationship Name = sample_rel2Master Vdisk info = volume_id3@svc_Cl1Auxiliary Vdisk info = volume_id4@svc_Cl2

SVC PPRC-Replicated Resource Configurationfor Resource Group RG1------------------------------------------SVC PPRC Consistency Group Name = CG1Master SVC Cluster Name = svc_Cl1Auxiliary SVC Cluster Name = svc_Cl2List of Relationships = sample_rel1, sample_rel2



**SVC PPRC Relationships Configuration for Consistency Group CG3-------------------------------------Relationship Name = sample_rel3Master Vdisk info = volume_id5@svc_Cl2Auxiliary Vdisk info = volume_id6@svc_Cl1

SVC PPRC-Replicated Resource Configurationfor Resource Group RG2------------------------------------------SVC PPRC Consistency Group Name = CG1Master SVC Cluster Name = svc_Cl2Auxiliary SVC Cluster Name = svc_Cl1List of Relationships = sample_rel3

From this point, resource groups RG1 and RG2 are configured to include the SVC PPRC-Replicated Resources CG1 and CG2 respectively.

Note: Since nodes located at each site treat the pairs differently in some situations, using PPRC pairs in resource groups that have no site preference could lead to unpredictable results.

Planning PPRC-SVC Management SupportIf you have not already done so, use the planning worksheets in Appendix A: General HACMP Planning Worksheets to plan HACMP sites necessary for this configuration. For this section, use the worksheets in Appendix D: Planning Worksheets for SVC PPRC to record the configuration information. You can use this information when making HACMP configuration entries in SMIT. Sections in this chapter describe the information to provide on these worksheets.

Before you configure SVC management option, you configure SVC itself to manage PPRC mirrored volumes. Ensure that your SVC configuration:• Makes a consistent point-in-time copy of data available at all times• Has virtual disks (VDisks) made available from both SVC Clusters to be used in the

HACMP configuration.

For information about configuring SVC, see the appropriate SVC documentation. See also the Limitations and Restrictions for HACMP/XD: Metro Mirror for SVC for any naming convention restrictions.

Setting up Volume Groups and Filesets on SVC PPRC-Protected DisksAs part of planning your HACMp/XD PPRC SVC environment, decide what vDisks will be used to support volume groups and file systems on the HACMP cluster nodes.

Regardless of whether you simply plan the vDisks at this time or perform the steps listed below, you must compete these steps prior to the initial HACMP SVC PPRC verification step. That step creates the final SVC PPRC relationships.



10

6

This procedure assumes that the hdisks and corresponding vpaths made available to your nodes are visible at those nodes. If they are not, and you can verify that SVC has been configured correctly to make the virtual disks (vDisks) available, reboot the node and run cfgmgr to make the disks viewable.

Note: We recommend making a copy of the SVC PPRC Mirrored Volumes Worksheet from Appendix D: Planning Worksheets for SVC PPRC, to record the information gathered below.

Discover AIX vpaths associated with SVC vDisks1. Select vDisks from your SVC configuration to be used in SVC PPRC relationships (that

will be grouped in consistency groups) that will be managed by HACMP.

2. Based on the vDisks you have chosen to use in a given SVC PPRC relationship, find out which vpaths and hdisks correspond.

a. On an HACMP node, run:lsdev -Ccdisk | more

From this output, you can discover which hdisks are associated with the SAN Volume Controller:HAnode1> lspv…

hdisk32 Available 81-08-02 SAN Volume Controller Devicehdisk33 Available 81-08-02 SAN Volume Controller Devicehdisk34 Available 81-08-02 SAN Volume Controller Devicehdisk35 Available 81-08-02 SAN Volume Controller Devicehdisk36 Available 81-08-02 SAN Volume Controller Devicehdisk37 Available 81-08-02 SAN Volume Controller Devicevpath0 Available Data Path Optimizer Pseudo Device Drivervpath1 Available Data Path Optimizer Pseudo Device Drivervpath2 Available Data Path Optimizer Pseudo Device Drivervpath3 Available Data Path Optimizer Pseudo Device Driver

In the example listing above, hdisk32 - hdisk37 are all SAN Volume Controller Devices.

b. Then run one of the following commands: `lsvpcfg | grep <hdisk>`, to list the vpath for a specific hdisk`lsvpcfg | grep vpath`, to list information for all vpaths.

SVC-related vpaths have an output similar to this:[HAnode1][/usr/es/sbin/cluster/svcpprc/utils]> lsvpcfg | grep vpath

vpath2 (Avail pv ) 6005076801898009980000000000012A = hdisk4 (Avail ) hdisk11 (Avail ) hdisk24 (Avail ) hdisk31 (Avail )

vpath3 (Avail pv ) 60050768018980099800000000000129 = hdisk5 (Avail ) hdisk12 (Avail ) hdisk25 (Avail ) hdisk32 (Avail )

c. Using the volume ID number from lsvpcfg: (ie: 6005076801898009980000000000012A )

Run the following command to get any associated vDisk value. (If there is doubt as to the output, run the command without the trailing 'grep' and 'cut'):ssh admin@<SVC Cluster IP address> svcinfo lshostvdiskmap -delim : | grep <SVC volume ID value> | cut -f5 -d":"



Example:HAnode1> ssh [email protected] svcinfo lshostvdiskmap -delim : |grep 6005076801898009980000000000012C |cut -f5 -d":"

vDisk1

HAnode1>

The value returned should be the name given to the vDisk during its creation on the SVC system.

Set Up Volume Group and Filesystem for HACMP Management3. On the vpath(s) that corresponds to the master vDisk(s) for a given SVC PPRC

Relationship, set up the volume group and filesystem to be managed by HACMP. Ensure that the Volume Major Number for the volume group can be used on all HACMP cluster nodes, and that the physical volume name for the file system can also be used across all HACMP cluster nodes. Use the lvlstmajor command on each cluster node to display the available Volume Major Numbers.

4. Vary off (using varyoffvg) the newly created volume group and import it to all nodes in the local HACMP site.

5. Create a temporary SVC PPRC relationship to copy the volume group/file set information to the auxiliary vDisk. Run the following commands. (Refer to the SVC CLI documentation for more details.)

ssh admin@<master SVC Cluster IP> svctask mkrcrelationship -master <vDisk_name>-aux <vDisk_name> -cluster <Aux Cluster name> -name <relationship name>

ssh admin@<master SVC Cluster IP> svctask startrcrelationship <relationship name>

At this point, wait until the relationship moves from inconsistent_copying to consistent_synchronised state. Check the state by running:

ssh admin@master<master SVC Cluster IP> svcinfo lsrcrelationship [relationship name]

6. Once the SVC PPRC relationship has completed copying, delete the relationship:ssh admin@<master SVC Cluster IP> svctask rmrcrelationship <relationship name>

This step is necessary in order for the next LVM operations to complete successfully.

7. Using SMIT or the command line on the Backup HACMP site (the site that is connected to the Auxiliary SVC Cluster,) import the volume group(s) created in step 2c.

8. Ensure that on ALL cluster nodes, the AUTO VARYON feature for volume groups is set to NO:

On each node run:

chvg -a 'n' -Q 'y' <volume group name>

HACMP will attempt to auto-correct this during verification but will not be able to do so in the case of remote PPRC. (If done at this point, it will save you time later.)



10

6

At this point, the volume groups and file systems necessary to configure HACMP have been created.

Planning SVC Clusters

The first step in planning SVC Management is to plan which SVC clusters will interact with HACMP sites. (HACMP sites should have already been planned and/or defined.)

Use the SVC Cluster Worksheet in Appendix D: Planning Worksheets for SVC PPRC to complete this next section.

1. Record the following information on a copy of the SVC Cluster Worksheet, one entry per SVC cluster

Planning SVC PPRC Mirrored VolumesIt is recommended to do this step prior to setting up any volume groups, as it may assist in keeping things straight during that process.Use the SVC PPRC Mirrored Volumes Worksheet in Appendix D: Planning Worksheets for SVC PPRC to complete this section.

1. Record the HACMP Sites and corresponding SVC Cluster Names on a copy of the worksheet.

2. Record the following information on a copy of the worksheet, one entry per volume group to be managed by HACMP. Some of this information will be used for configuring SVC Relationships. Some of it is for record-keeping purposes that will make creating and cross-referencing the SVC Relationships and AIX vpaths easier in troubleshooting situations.

HACMP site The HACMP site associated with this SVC cluster.

SVC Cluster Name Enter the same name for the SVC cluster. This name cannot be more than 20 alphanumeric characters and underscores.

SVC Cluster Role Select Master or Auxiliary. The Master SVC Cluster is usually defined at the Primary HACMP site, the Auxiliary SVC Cluster at the Backup HACMP site.

SVC Cluster IP Address IP address of this cluster. used by HACMP to submit PPRC management commands.

Remote SVC Partner The name of the SVC Cluster that will be hosting vDisks from the other side of the SVC PPRC link.

Volume Group Name Volume group that will be used to host file systems, applications, etc. that will be mirrored over the PPRC link

Volume Group Major Number Major Number associated with the volume group. (for record keeping, and importing the volume groups onto other cluster nodes.)



3. For each HACMP site listed, enter the following:

4. Continue to fill in the table for all volume groups to be managed in this cluster

Planning SVC PPRC Consistency Groups (SVC PPRC Replicated Resources)

The definition for an SVC PPRC replicated resource contains the name of the consistency group, the list of SVC relationships to be contained in the consistency group, and the Master and Auxiliary SVC Clusters this consistency group will reside on.

During cluster verification, HACMP/XD creates these consistency groups via SVC CLI commands, which will then allow the SVC system to manage it.

For this step, complete an SVC PPRC Replicated Resource Worksheet, found in Appendix D: Planning Worksheets for SVC PPRC. Also on this sheet, you will find a column for entering the HACMP Resource Group that will manage this Replicated Resource. This is to serve as a cross-reference for the next step in Planning.

1. For each HACMP site listed, enter the following:

AIX vpath name See Discover AIX vpaths associated with SVC vDisks below in this section for steps on how to discover the vpath associated with a given SVC vDisk. Record the vpath associated with the vDisk chosen for this volume group in this column in the worksheet.

SVC PPRC Relationship Name Name that will be used to create this SVC PPRC relationship.

SVC vDisk Name vDisk chosen to host this volume group. The SVC PPRC Relationship as entered in the SMIT panels will be a combination of this name and the SVC cluster name, that is, vDisk@SVCclusterName.

Role Master or Auxiliary. Master denotes that this vDisk will be the first one SVC will attempt to make the source (that is, I/O accessible side) of the PPRC pair when it initially comes online. Auxiliary denotes the target member of the pair.

SVC Consistency Group Name

Name that will be used by HACMP to create the SVC consistency groups used to manage PPRC resources

HACMP Resource Group Name

HACMP resource group that will manage this consistency group.

Master SVC Cluster Name

Name of the SVC Cluster that will be the first to have the source side of the PPRC pair available at cluster startup. (Primary side.)

Associated HACMP Site

(For record-keeping.) HACMP site that the SVC Cluster listed in the previous column communicates with (is local to).



10

6

Completing the Resource Groups Worksheet

Complete a Resource Groups Worksheet (See Appendix A: General HACMP Planning Worksheets) for each resource group that includes SVC-managed PPRC replicated resources.

To identify information required to include SVC-managed PPRC replicated resources in a resource group, enter the following information on the Resource Groups Worksheet:

1. Record the Resource Group Name for the resource group that will contain SVC-managed PPRC replicated resources. The name can contain up to 32 characters and underscores, but without a leading numeric character.


3. Record the Site Management Policy: Prefer Primary Site or Online on Either Site.Prefer Primary Site is recommended for use with PPRC.

4. Record the names of the nodes in the resource group, listing the nodes in the prioritized order in which they will appear in the nodelist for the resource group.

5. Record the name of each SVC-managed PPRC replicated resource (volume set) to be included in the resource group.


Next Steps

After you plan the configuration for SVC-managed PPRC replicated resources in an HACMP cluster:

1. Install the HACMP/XD for Metro Mirror software on HACMP cluster nodes.

2. Configure SVC Management support for your HACMP cluster.

Auxiliary SVC Cluster Name

Name of the SVC cluster that will take over the target side of the PPRC pair at cluster startup. (Secondary)

Associated HACMP Site

(For record-keeping.) HACMP site that the Auxiliary SVC Cluster communicates with (is local to).

List of SVC Relationships

SVC Relationships that will be managed through this consistency group. SVC commands will be issued to the consistency group so it will fall over or fall back as a single entity. Ensure that the Master and Auxiliary SVC Cluster name for the relationships listed here match the Master and Auxiliary SVC Cluster names for the consistency group.


HACMP/XD for Metro Mirror with SVC Management Installing HACMP/XD: Metro Mirror for SVC 6

Installing HACMP/XD: Metro Mirror for SVCYou must be logged in as root to perform installation tasks.

Installing Prerequisite SoftwareBefore installing HACMP/XD: Metro Mirror for SVC v.5.4, the following software must be installed on cluster nodes:

1. AIX 5.2.0 ML 4 or higher.

2. openssh version 3.6.1 or higher (+ license), for access to the SVC interface.

3. HACMP/XD Version 5.4.0 (cluster.es.server.rte 5.4.0.0) or higher.

4. IBM 2145 Subsystem Device Driver (SDD). Check the SVC documentation and/or website for the currently approved SDD version to use with a given SVC version. Make sure the recommended versions of the following filesets are installed:

a. devices.fcp.disk.ibm.rte (1.0.0.0)

b. devices.sdd.**.rte (latest version)

c. devices.ibm2105.rte (latest version)

Note: devices.fcp.disk.ibm2145.rte (the SVC fcp device driver.was the original fileset listed for SVC PPRC support. The device listed above in c. has since been named as its successor.

It may not be clear during installation of the sdd filesets that this first fileset is necessary, but it is critical in order to ensure that failover works correctly.

Installing HACMP/XD: Metro Mirror for SVC FilesetsInstall these filesets on each cluster node that may take over a SVC PPRC-mirrored volume group:• cluster.es.svcpprc

• cluster.msg.svcpprc.en_US.

All SVC PPRC filesets are installed in the /usr/es/sbin/cluster/svcpprc directory.

Configuring HACMP/XD: Metro Mirror for SVCAfter you install SVC following the procedures outlined in its documentation, configure the cluster following these steps:

1. Check SVC and HACMP configuration requirements.

2. Configure HACMP and SVC PPRC.

3. Verify the HACMP and SVC PPRC configuration.

4. Synchronize the cluster.


HACMP/XD for Metro Mirror with SVC Management Configuring HACMP/XD: Metro Mirror for SVC

11

6

5. Start the cluster.

SVC and HACMP Configuration RequirementsMake sure the following requirements are met:

1. Configure ssh for each node that will communicate with the SVC Clusters. Public/private key pairs must be created on the nodes and distributed to the SVC clusters in order for HACMP/XD: Metro Mirror for SVC to function. (See the prerequisite list above for fileset name and version.)The SVC PPRC replicated methods used with HACMP/XD rely heavily on remote commands executed using ssh. It must be possible for example, to remotely execute a command via ssh on all cluster nodes that are part of a resource group that contains a PPRC replicated resource. Private/public key pairs must be installed on both the nodes that will access the SVC clusters, and the SVC clusters themselves in order for the SVC PPRC commands to work correctly.See the SVC documentation for directions on how to complete this task.

2. Make sure the Host Aliases on your SVC Clusters match the node names (short names) used in HACMP. To get a list of the current host to vDisk maps on the master and auxiliary SVC clusters, use the SVC CLI commands:

ssh admin@<ip_for_MASTER_SVC_Cluster> svcinfo lshostvdiskmap | moressh admin@<ip_for_AUXILIARY_SVC_Cluster> svcinfo lshostvdiskmap | more

Check the host names listed there against your HACMP node names. If they differ, then refer to the SVC documentation on how to change the names to match, either via the SVC CLI or GUI interface.

Defining SVC Cluster ConfigurationsUse the information in the SVC Cluster Worksheet to complete this part of the configuration. Using HACMP SMIT to define the SVC cluster:

1. smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resources Configuration > Configure SVC-PPRC Replicated Resources > SVC Clusters Definition to HACMP > Add an SVC Cluster and press Enter:


SVC Cluster Name Enter the same name used by SVC. This name cannot be more than 20 alphanumeric characters and underscores.

SVC Cluster Role Select Master or Auxiliary.

HACMP site Select the HACMP site associated with this SVC cluster.

SVC Cluster IP Address IP address of this cluster


HACMP/XD for Metro Mirror with SVC Management Configuring HACMP/XD: Metro Mirror for SVC 6

4. Press Enter after making your selections.

Defining SVC PPRC RelationshipsOnce you have defined the SVC cluster, define the SVC PPRC relationships. Use the values you entered on the SVC PPRC Mirrored Volumes Worksheet:

1. Enter smit hacmp

2. Select Extended Configuration > Extended Resource Configuration > HACMP Extended Resources Configuration > Configure SVC-PPRC Replicated Resources > SVC PPRC Relationships Definition > Add an SVC PPRC Relationship and press Enter.


4. Press Enter when you have finished the definition.

5. Create as many SVC PPRC Relationships as will be necessary to manage all the vpaths used for volume groups that will be managed by HAMCP.

Defining SVC PPRC ResourcesUse the information you enter on the SVC Replicated Resource Worksheet to define the SVC PPRC resources:

1. Enter smit hacmp

Remote SVC Partner Name of the SVC Cluster that will be hosting vDisks from the other side of the SVC PPRC link

Relationship Name The name used by both SVC and HACMP for configuration of SVC PPRC Relationships. Use no more than 20 alphanumeric characters and underscores.

Master VDisk Info The Master and Auxiliary VDisk names use the format:vdisk_name@svc_cluster_nameSee the section Setting up Volume Groups and Filesets on SVC PPRC-Protected Disks for steps on how to discover what virtual disks (VDisks) correspond to the vpaths used for your configuration.The Master VDisk is the disk that resides at the primary site for the resource group the SVC PPRC Relationship will be a part of.

Auxiliary VDisk Info The Auxiliary VDisk is the disk at the backup site for the resource group the SVC PPRC Relationship will be a part of.


HACMP/XD for Metro Mirror with SVC Management Configuring HACMP/XD: Metro Mirror for SVC

11

6

2. Select Extended Configuration > Extended Resource Configuration > HACMP Extended Resources Configuration > Configure SVC-PPRC Replicated Resources > > SVC PPRC-Replicated Resource Configuration > Add an SVC PPRC Resource, and press Enter.


4. Press Enter.

Adding SVC PPRC Consistency Groups to HACMP Resource GroupsAdd the SVC PPRC Consistency Groups created in the previous step to HACMP resource groups:

1. Enter smit hacmp

2. Select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Group Configuration > (*)

(*) Select Change/Show or Add a Resource Group depending on whether you are working with an existing resource group or creating a new resource group.

The SVC PPRC Replicated Resources entry appears at the bottom of the page in SMIT. This entry is a picklist that displays the resource names created in the previous step. Make sure that the volume groups selected on the Resource Group configuration screen match the vDisks used for the SVC PPRC Replicated Resource.

3. Press Enter.

Verifying and Synchronizing the HACMP/XD ClusterAt this point, provided you have configured all the other aspects of your resource group (volume groups being the most important,) verify the cluster and SVC PPRC configuration by taking the following steps:

1. Verify the HACMP cluster configuration (using SMIT).

2. Verify the SVC PPRC configuration (using CLI).

3. Synchronize the HACMP Cluster configuration (using SMIT).

SVC PPRC Consistency Group Name

The name to be used by SVC and also to be used in the resource group configuration. Use no more than 20 alphanumeric characters and underscores.

Master SVC Cluster Name Name of the Master cluster is the SVC Cluster connected to the HACMP Primary Site

Auxiliary SVC Cluster Name Name of the SVC Cluster connected to the HACMP Backup/Recovery Site

List of Relationships List of names of the SVC PPRC relationships


HACMP/XD for Metro Mirror with SVC Management Configuring HACMP/XD: Metro Mirror for SVC 6

Verifying the HACMP Cluster Configuration (using SMIT)

Be careful to run verification only. You must use the Extended Configuration SMIT path to get the correct menu.

1. Enter smit hacmp

2. Select Extended Configuration > Extended Verification and Synchronization and press Enter.

Enter the field values as follows:

3. Press Enter. The output from the verification is displayed in the SMIT Command Status window.

4. If you receive error messages, make the necessary changes and run the verification procedure again. You may see Warnings if the configuration has a limitation on its availability; for example, only one interface per node per network is configured.

Verifying the SVC PPRC Configuration (using the CLI)

To run the SVC PPRC verification, execute the following command:

/usr/es/sbin/cluster/svcpprc/utils/cl_verify_svcpprc_config

If any configuration errors appear, go back into the HACMP SMIT panels and correct the errors, then re-run this script.

All the SVC PPRC Relationships and Consistency Groups named in the previous steps are created during this step. If there were any errors during execution of this script, the SVC PPRC consistency groups and relationship may not be created.

To verify that the configuration exists, run the following two commands against either of the SVC clusters:

ssh admin@<SVC Cluster IP> svcinfo lsrcrelationship

ssh admin@<SVC Cluster IP> svcinfo lsrcconsistgrp

Verify Synchronize or Both Select Verify.


Select Yes.


Select No (the default.)

Verify changes only? Select no to run all verification checks that apply to the current cluster configuration.



HACMP/XD for Metro Mirror with SVC Management Changing an HACMP/XD for Metro Mirror SVC Configuration

11

6

Synchronizing the HACMP Cluster Configuration

To propagate the new SVC PPRC configuration information (and possibly HACMP site information) across your HACMP cluster:

1. Enter smit hacmp

2. Select Extended Configuration > Extended Verification and Synchronization and press Enter.

Enter the field values as follows:

3. Press Enter. The cluster is synchronized. The output is displayed in the SMIT Command Status window.

Starting the HACMP/XD SVC PPRC ClusterYour HACMP cluster is now ready to be started. HACMP will manage the SVC PPRC replicated resources you have just defined. HACMP starts these SVC PPRC resources automatically, so there should be no need to manually manage the SVC resources from this point, under normal HACMP cluster operation.

Changing an HACMP/XD for Metro Mirror SVC Configuration

Using HACMP SMIT, you can change, show, or remove:• SVC Cluster configurations• SVC PPRC relationships• SVC PPRC Resources.

Be sure to update the HACMP resource group information and synchronize the cluster after making changes.

Verify Synchronize or Both Select Synchronize.


Select No (the default.).


Select No (the default.)

Verify changes only? Select no to run all verification checks that apply to the current cluster configuration.



HACMP/XD for Metro Mirror with SVC Management Changing an HACMP/XD for Metro Mirror SVC Configuration 6

Changing an SVC Cluster ConfigurationTo change the SVC cluster:

1. smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resources Configuration > Configure SVC-PPRC Replicated Resources > SVC Clusters Definition to HACMP > Change/Show an SVC Cluster and press Enter:

3. Select the SVC cluster to change from the picklist. and press Enter.


5. Press Enter after making your selections.

Removing an SVC Cluster ConfigurationTo remove the SVC cluster definition:

1. smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resources Configuration > Configure SVC-PPRC Replicated Resources > SVC Clusters Definition to HACMP > Remove an SVC Cluster and press Enter:

3. Select the SVC cluster to remove from the picklist. and press Enter.

4. Confirm that you want to remove this SVC cluster definition by pressing Enter again.

Changing SVC PPRC RelationshipsTo change the SVC PPRC relationships:

1. Enter smit hacmp

SVC Cluster Name The current name of the SVC cluster.

New SVC Cluster Name Enter the name of the SVC Cluster Use no more than 20 characters.

SVC Cluster Role Select Master or Auxiliary.

HACMP site Select the HACMP site associated with this SVC cluster. The site must already be defined in HACMP.

SVC Cluster IP Address IP address of this cluster.

Remote SVC Partner The name of the SVC Cluster that will be hosting vDisks from the other side of the SVC PPRC link.


HACMP/XD for Metro Mirror with SVC Management Changing an HACMP/XD for Metro Mirror SVC Configuration

11

6

2. Select Extended Configuration > Extended Resource Configuration > HACMP Extended Resources Configuration > Configure SVC-PPRC Replicated Resources > SVC PPRC Relationships Definition > Change/Show an SVC PPRC Relationship and press Enter.

3. Select the SVC PPRC relationship to change and press Enter.


5. Press Enter when you have finished the definition.

6. Create as many SVC PPRC Relationships as will be necessary to manage all the vpaths used for volume groups that will be managed by HAMCP.

Removing SVC PPRC RelationshipsTo remove the SVC PPRC relationships:

1. Enter smit hacmp

2. Select Extended Configuration > Extended Resource Configuration > HACMP Extended Resources Configuration > Configure SVC-PPRC Replicated Resources > SVC PPRC Relationships Definition > Remove an SVC PPRC Relationship and press Enter.

3. Select the SVC PPRC relationship to remove and press Enter.

4. Confirm that you want to remove this SVC PPRC relationship by pressing Enter again.

Relationship Name The current relationship name displays here.

New Relationship Name The new name to be used by both SVC and HACMP for configuration of SVC PPRC Relationships. Use no more than 20 characters.

Master VDisk Info The Master and Auxiliary VDisk names use the format:vdisk_name@svc_cluster_nameSee the section Setting up Volume Groups and Filesets on SVC PPRC-Protected Disks for steps on how to discover what virtual disks (VDisks) correspond to the vpaths used for your configuration.The Master VDisk is the disk that resides at the primary site for the resource group the SVC PPRC Relationship will be a part of.

Auxiliary VDisk Info The Auxiliary VDisk is the disk at the backup site for the resource group the SVC PPRC Relationship will be a part of.


HACMP/XD for Metro Mirror with SVC Management Changing an HACMP/XD for Metro Mirror SVC Configuration 6

Changing SVC PPRC ResourcesTo change the SVC PPRC resources:

1. Enter smit hacmp

2. Select Extended Configuration > Extended Resource Configuration > HACMP Extended Resources Configuration > Configure SVC-PPRC Replicated Resources > > SVC PPRC-Replicated Resource Configuration > Change/Show an SVC PPRC Resource, and press Enter.

3. Select the SVC PPRC resource to change and press Enter.


5. Press Enter.

Removing SVC PPRC ResourcesTo remove the SVC PPRC resources:

1. Enter smit hacmp

2. Select Extended Configuration > Extended Resource Configuration > HACMP Extended Resources Configuration > Configure SVC-PPRC Replicated Resources > > SVC PPRC-Replicated Resource Configuration > Change/Show an SVC PPRC Resource, and press Enter.

3. Select the SVC PPRC resource to remove and press Enter.

4. Press Enter again to confirm this action.

Modifying SVC PPRC Consistency Groups in HACMP Resource GroupsIf you made changes to SVC PPRC consistency groups, modify the SVC PPRC Consistency Groups currently contained in HACMP resource groups:

1. Enter smit hacmp

2. Select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Group Configuration > (*)

(*) Select Change/Show or Add a Resource Group depending on whether you are working with an existing resource group or creating a new resource group.

SVC PPRC Consistency Group Name

The current name to used by SVC and also used in the Resource Group Configuration.

New SVC PPRC Consistency Group Name

The new name to used by SVC and also to be used in the Resource Group Configuration.

Master SVC Cluster Name The name of the Master cluster is the SVC Cluster connected to the HACMP Primary Site.

Auxiliary SVC Cluster Name The name of the SVC Cluster connected to the HACMP Backup/Recovery Site.

List of Relationships A list of names of the SVC PPRC relationships.


HACMP/XD for Metro Mirror with SVC Management Troubleshooting HACMP/XD for Metro Mirror for SVC

11

6

The SVC PPRC Replicated Resources entry appears at the bottom of the page in SMIT. This entry is a picklist that displays the resource names created in the previous step. Make sure that the volume groups selected on the Resource Group configuration screen match the vDisks used for the SVC PPRC Replicated Resource.

3. Press Enter.

Troubleshooting HACMP/XD for Metro Mirror for SVCThis section provides information that may help with troubleshooting SVC PPRC clusters.

SVC PPRC StatesThe following PPRC Volume states are possible, for either Consistency Groups or PPRC Relationships:• inconsistent_stopped

In this state, the Primary is accessible for read and write I/O but the Secondary is not accessible for either. A copy process needs to be started to make the Secondary Consistent.

• inconsistent_copying

In this state, the Primary is accessible for read and write I/O but the Secondary is not accessible for either. This state is entered after a Start command is issued to an InconsistentStopped Relationship or Consistency Group. It is also entered when a Forced Start is issued to an Idling or ConsistentStopped Relationship or Consistency Group. A background copy process will run which will copy data from the Primary to the Secondary virtual disk.

• consistent_stopped

In this state, the Secondary contains a Consistent image but it might be out-of-date with respect to the Primary.

• consistent_synchronized

In this state, the Primary VDisk is accessible for read and write I/O. The Secondary VDisk is accessible for read-only I/O. Writes that are sent to the Primary VDisk are sent to both Primary and Secondary VDisks. Either good completion must be received for both writes, or the write must be failed to the host, or a state transition out of ConsistentSynchronized must take place before a write is completed to the host.

• idling

Both Master and Auxiliary disks are operating in the Primary role. Consequently both are accessible for write I/O. In this state the Relationship or Consistency Group will accept a 'Start' command. Remote Copy maintains a record of regions on each disk which have received write I/O while Idling. This will be used to determine what areas need to be copied following a 'Start' command.

• idling_disconnected

The Virtual Disk(s) in this half of the Relationship or Consistency Group are all in the Primary role and will accept read or write I/O. No configuration activity is possible (except for deletes or stops) until the Relationship becomes connected again. At that point, the Relationship will transition to a Connected state.


HACMP/XD for Metro Mirror with SVC Management Troubleshooting HACMP/XD for Metro Mirror for SVC 6

• inconsistent_disconnected

The Virtual Disks in this half of the Relationship or Consistency Group are all in the Secondary role and will not accept read or write I/O. No configuration activity except for deletes is permitted until the Relationship becomes connected again.

• consistent_disconnected

The VDisks in this half of the Relationship or Consistency Group are all in the Secondary role and will accept read I/O but not write I/O. This state is entered from ConsistentSynchronized or ConsistentStopped when the Secondary side of a Relationship becomes disconnected.

• empty

This state only applies to Consistency Groups. It is the state of a Consistency Group which has no Relationships and hence no other state information to show. It is entered when a Consistency Group is first created. It is exited when the first Relationship is added to the Consistency Group at which point the state of the Relationship becomes the state of the Consistency Group.

Viewing the State of Consistency Groups

To view the current state of the consistency groups HACMP is managing (via the configured resource groups) execute the following command on any of the HACMP cluster nodes:

ssh admin@<SVC cluster IP address> svcinfo lsrcconsistgrp

where <SVC Cluster IP Address> is for either SVC cluster being used by HACMP.

Viewing SVC ClustersYou can use the following commands to view information about SVC clusters. These commands are stored in the /usr/es/sbin/cluster/svcpprc/cmds directory.

cllssvc command

List SVC cluster information:cllssvc [-n < svccluster_name >] [-c]

The command lists information about all SVC clusters in the HACMP configuration or a specific SVC cluster. If no SVC is specified, all SVC clusters defined will be listed. If a specific SVC cluster is provided via the -n flag, information about this SVC only will be displayed. The -c flag displays information in a colon-delimited format.

Sample Output[/usr/es/sbin/cluster/svcpprc/cmds]> cllssvcsvc9Asvc78

[/usr/es/sbin/cluster/svcpprc/cmds]> cllssvc -n svc9Asvccluster_name svccluster_role sitename cluster_ip r_partnersvc9A Auxiliary Vancouver 9.114.230.93 svc78

[/usr/es/sbin/cluster/svcpprc/cmds]> cllssvc -n svc9A -c#SVCNAME:ROLE:SITENAME:IPADDR:RPARTNERsvc9A:Auxiliary:Vancouver:9.114.230.93:svc78


HACMP/XD for Metro Mirror with SVC Management Troubleshooting HACMP/XD for Metro Mirror for SVC

12

6

cllssvcpprc command

List information about all SVC PPRC resources or a specific SVC PPRC resource:

cllssvcpprc [-n < svcpprc_consistencygrp >] [-c] [-a] [-h]

If no resource name is specified, the names of all PPRC resources defined will be listed. If the -a flag is provided, full information about all PPRC resources is displayed. If a specific resource is provided via the -n flag, information about this resource only will be displayed. The -c flag displays information in a colon-delimited format. The -h flag turns off the display of column headers.

Sample Output[/usr/es/sbin/cluster/svcpprc/cmds]> cllssvcpprcHASVC1

[/usr/es/sbin/cluster/svcpprc/cmds]> cllssvcpprc -n HASVC1svcpprc_consistencygrp MasterCluster AuxiliaryCluster relationshipsHASVC1 svc78 svc9A svcrel1

[/usr/es/sbin/cluster/svcpprc/cmds]> cllssvcpprc -n HASVC1 -ca#NAME:MASTER:AUXILIARY:RELATIONSHIPSHASVC1:svc78:svc9A:svcrel1

cllsrelationship command

List information about all SVC PPRC relationships or a specific PPRC relationship.

cllsrelationship [-n <relationship_name>] [-c] [-a] [-h]

If no resource name is specified, the names of all PPRC resources defined will be listed. If the -a flag is provided, full information about all PPRC relationship is displayed. If a specific relationship is provided via the -n flag, information about this relationship only will be displayed. The -c flag displays information in a colon-delimited format. The -h flag turns off the display of column headers.

Sample Output[/usr/es/sbin/cluster/svcpprc/cmds]> cllsrelationshipsvcrel1

[/usr/es/sbin/cluster/svcpprc/cmds]> cllsrelationship -n svcrel1relationship_name MasterVdisk_info AuxiliaryVdisk_infosvcrel1 c48f1rp06_075@svc78 c48f2rp08_095@svc9A

[/usr/es/sbin/cluster/svcpprc/cmds]> cllsrelationship -n svcrel1 -c#RELATIONSHIP:MASTERVDISK:AUXVDISKsvcrel1:c48f1rp06_075@svc78:c48f2rp08_095@svc9A

cl_verify_svcpprc_config command

The cl_verify_svcpprc_config command verifies the SVC definition in the HACMP configuration. After the successful verification of the SVC configurations, it establishes all the SVC relationships defined to HACMP on the SVC clusters and adds them to the corresponding consistency groups.


Chapter 7: HACMP/XD for Metro Mirror with eRCMF Management

This chapter describes the planning tasks for setting up HACMP/XD for MetroMirror with eRCMF Management, from here on referred to as eRCMF management. At the end of this chapter, your HACMP cluster will be configured to integrate eRCMF-managed PPRC replicated resources. eRCMF simplifies how you manage MetroMirror (PPRC) on ESS systems and how you can integrate PPRC replicated resources into an HACMP configuration.

This chapter contains the following sections:• Planning for eRCMF Management• Installing eRCMF Management Filesets• Configuring eRCMF Management• Monitoring eRCMF Status

Planning for eRCMF ManagementYou should be familiar with the planning tasks for HACMP. For information about planning HACMP clusters, see the HACMP for AIX: Planning Guide.

Planning OverviewTo plan for eRCMF-managed PPRC replicated resources in an HACMP cluster, complete the following planning tasks:• Plan connections to the PCM• Identify the eRCMF sites• Identify the eRCMF global attributes used to configure eRCMF• Plan the HACMP sites• Identify the volume sets, and the associated volume groups, to be included in specified

HACMP resource groups• Plan which resource groups will contain the eRCMF-managed PPRC replicated resources.

Limitations for eRCMF SupportHACMP/XD for Metro Mirror with eRCMF 5.4 has the following limitations:• Only non-concurrent volume groups can be used with PPRC. Disk heartbeat networks are

not supported in HACMP for sites using this configuration.• HACMP does not trap eRCMF SNMP notification events.• eRCMF version 2 does not support Fibre Channel connections between ESS systems.

ESCON links connect the ESS systems.


HACMP/XD for Metro Mirror with eRCMF Management Planning for eRCMF Management

12

7

• Cascading PPRC is not supported.

Sample ConfigurationYou can set up a mutual recovery configuration in which each site acts as a production site with the other site acting as an associated backup site. Implementing a mutual recovery configuration requires:• Two HACMP sites (the same as a single recovery configuration)• Two resource groups.

In HACMP, you configure two sites with the same names as the eRCMF sites and a resource group that includes nodes, eRCMF-managed PPRC replicated resources, and the volume groups associated with the replicated resources.

The following figure shows a mutual recovery configuration with two resource groups in which Site_2 is the production site with Site_1 as the recovery site for Resource Group 2, and Site_1 is the production site with Site_2 as the recovery site for Resource Group 1. In Resource Group 1, the nodes n1 and n2 have a higher priority, and in Resource Group 2 the nodes n3 and n4 have a higher priority. The volume groups in both resource groups are included in eRCMF-managed PPRC replicated resources.

Sample Configuration for Resource Groups for eRCMF-Managed PPRC Replicated Resources

The order of the nodes in each resource group’s nodelist indicates which site is considered to be the production site and which site is considered to be the backup site for that resource group. The site that includes the nodes with the highest priority is considered the production site for the resource group.


HACMP/XD for Metro Mirror with eRCMF Management Planning for eRCMF Management 7

In the preceding example, the configuration for Resource Group 1 would be:

In the preceding example, the configuration for Resource Group 2 would be:

Planning Worksheets for eRCMF EnvironmentUse the planning worksheets in Appendix E: Planning Worksheets for an eRCMF Environment to record the HACMP/XD for ESS PPRC configuration information. You can use this information when making HACMP configuration entries in SMIT. Sections in this chapter describe the information to provide on these worksheets.

Planning Integration with eRCMFBefore you configure HACMP/XD for Metro Mirror with the eRCMF option, you configure eRCMF to manage the mirrored volumes. Ensure that your eRCMF configuration:• Makes a consistent point-in-time copy of data available at all times• Has site split values supported by HACMP:

• (Recommended) Freeze & Go• (Not recommended) Freeze & Stop

• Has volume set types supported by HACMP:• XD volume sets:

• XDNOFCPY• XDALLFCPY

• Non-XD volume sets:• NOFCPY• ALLFCPY.

For information about configuring eRCMF, see the eRCMF User Guide.

With eRCMF managing PPRC pairs, whether a node can access data depends on the following:• Which site is the active production site• The site split behavior configured for eRCMF

For more information about the site split behaviors, see the eRCMF documentation.• The status of the volume set in eRCMF, such as in sync, split, or active recovery site.

Note: If the configuration contains shadow volumes, servers do not normally attach to them.







12

7

Working with eRCMF Site Split Behaviors

HACMP supports the following behaviors when an eRCMF site splits: • Freeze & Go splits the sites and frees the I/O as soon as possible.• Freeze & Stop splits the sites and blocks I/O until eRCMF receives instructions to release

the block and allow I/O to resume.

WARNING: Using Freeze & Stop with HACMP is not recommended. If you configure eRCMF to use Freeze & Stop, HACMP can do little to ensure availability because eRCMF is waiting for instructions to release the I/O. In this case, you need to run the eRCMF release command to release the I/O. HAMCP can then take control of the eRCMF-managed disks.

For information about how eRCMF manages site splits, see the eRCMF User Guide.

Planning Connections to the Productivity Center Machines

For HACMP to communicate with eRCMF, it requires that each cluster node connect to each PCM through the ESSNet administrative network. The HACMP cluster nodes can either connect directly to the ESSNet or connect over a router or bridge. The interfaces that communicate with the ESSNet can be defined to HACMP, but IP address takeover (IPAT) should not be used for those interfaces. If the interfaces are defined to HACMP, ensure that the nodes can gain access to the PCM at any time.

Note: The reliability of this connection directly affects the performance of HACMP/XD for ESS PPRC. A slow or unreliable network connection results in a fallover being processed slowly or unreliably.

Understanding eRCMF Sites

HACMP supports two sites, each in a different physical location. These sites are also defined in eRCMF: • Production. The production site is the active production site—the site on which

applications are running. After a fallover, the production site can become the backup site.• Backup. The backup site has the applications set up, but they are not active. After a fallover,

the backup site becomes the production site.

At any point in time, eRCMF considers one site to be the production site, and the other site to be the backup site.

An instance of eRCMF runs on each PCM in the two-site configuration. Each PCM also acts like an active Copy Services Server. The active eRCMF instance resides on the system located at the initial production site. The backup eRCMF instance resides on the system located at the initial backup site. If the active PCM fails, the backup PCM takes over as the active eRCMF instance.

Note: Each site may have FlashCopy enabled. FlashCopy is an optional feature of the ESS.



Identifying Global Attributes

HACMP/XD for Metro Mirror requires information about the global attributes configured for eRCMF. The global attributes provide general information that eRCMF uses to manage PPRC.

Completing the eRCMF Global Attributes Worksheet

Complete an eRCMF Global Attributes Worksheet to identify the configuration parameters for eRCMF. The enterprise configuration file for eRCMF provides the information to be supplied for step 3 through step 7 of the following procedure. This file is typically named enterprise.dat and located in the home directory for eRCMF (/usr/ercmf by default) on the PCMs.

To complete the eRCMF Global Attributes Worksheet:

1. Record the full path and file name of the enterprise configuration file. This information is for your reference, should you need to verify configuration information at another time.

2. List the names of the eRCMF sites.The default site names are SITE1 and SITE2.

3. Record the site Split Behavior configured for eRCMF. When eRCMF splits sites, it suspends PPRC mirroring. This behavior defines the operation mode for eRCMF when an event occurs that suspends mirroring between members of a PPRC pair:

4. Record the PPRC-XD Recovery Policy:

5. Record the PPRC Link Types to identify the direction in which PPRC links are established:

Freeze&Go eRCMF splits the sites and frees I/O as soon as possible. Freeze&Go is the recommended setting and is the default.

Freeze&Stop (Not recommended) eRCMF splits the sites and blocks I/O until it receives instructions to release the block and allow I/O to resume.

Automatic XD volume sets fall over to the recovery site without manual intervention. In this situation, a point-in-time copy may be inconsistent with the copy that was active on the production site prior to the fallover.

Manual XD volume sets fall over to the recovery site after you provide input to eRCMF. This lets you evaluate whether or not you have a consistent point-in-time copy, and decide what actions should be taken for your application.

TwoWay Links are established in both directions at the same time.TwoWay is the default.

OneWay All links are established in the same direction (one direction at a time).



12

7

6. Record the name of the Primary Copy Services Server, the Copy Services Server that is typically active until a failure occurs. The Primary Copy Services Server is the Copy Services Server on the PCM that is initially active.

7. Record the name of the Backup Copy Services Server, the secondary Copy Services Server that acts as a backup should the primary Copy Services Server be unavailable. The BAckup Copy Services Server is the Copy Services Server on the PCM that is initially the backup PCM.

8. Record the eRCMF User ID, the administrative user ID that is used to access eRCMF.

9. Record the eRCMF User Password, the password that is associated with the eRCMF user ID.If you are using Secure Socket Layer (SSL) for security in eRCMF, this password also logs you into SSL.

Note: In HACMP, you can specify only one set of global attributes.

Identifying Volume Sets

A volume set is a group of volumes that eRCMF manages as a unit. A volume set, when defined to HACMP, is an eRCMF-managed PPRC replicated resource. To add volume sets to an HACMP configuration:• Define the volume set to HACMP as an eRCMF-managed replicated resource.• Add eRCMF-managed PPRC replicated resources to HACMP resource groups.• Define volume groups to HACMP that are associated with the volume sets that are defined

to HACMP.

Volume sets have attributes that identify whether or not FlashCopy is enabled for the volumes in the group, and if it is enabled, how it is configured. eRCMF manages the fallover and fallback of PPRC volumes based on the volume set that includes the volumes.

Completing the Volume Sets Worksheet

Complete a Volume Sets Worksheet for each volume set to be included in the cluster. You can locate a list of volume sets configured for eRCMF in the enterprise configuration file for RCMF. This file is typically named enterprise.dat and is located in the home directory for eRCMF (/usr/ercmf by default) on the PCMs.

You can locate the remainder of the information about a specified volume set in the configuration file for the volume set. This file is typically named Volume_Set_Name.dat, where Volume_Set_Name is the name of the volume set. The file is located in the home directory for eRCMF (/usr/ercmf by default) on the PCMs.

To complete the Volume Sets Worksheet:

1. Record the full path and file name of the volume set configuration file (Volume_Set_Name.dat). This information is for your reference, should you need to verify configuration information at another time.

2. Record the name of the eRCMF Production Site and the eRCMF Backup Site for this volume set.



3. Record the Volume Set Name of the volume set to be included in the HACMP configuration. This name may contain up to 20 characters.

4. Record the Volume Set Type. The type of a volume set depends on whether the set is in an extended distance environment or not. The type indicates whether or not FlashCopy is enabled and if it is, how it is configured. Type is one of the following values:

5. Record the names of the Volume Groups in the volume set that will be part of the HACMP configuration. These volume groups must already be defined to AIX before they can be defined to HACMP.

Planning HACMP Configuration with eRCMF ManagementHACMP manages resource groups and HACMP sites, and it communicates site events to eRCMF. eRCMF manages the PPRC mirrored volumes.

Planning HACMP Sites

An HACMP configuration to support HACMP/XD for ESS PPRC requires two HACMP sites. Each HACMP site is associated with an eRCMF site and has the same name as the eRCMF site.

For information about planning for HACMP sites in general, see the HACMP Planning and Installation Guide.

Completing the eRCMF-Managed Resources Site Worksheet

Complete an eRCMF-Managed Resources Site Worksheet to identify eRCMF and HACMP sites for use by HACMP/XD for ESS PPRC.

To complete the eRCMF-Managed Resources Site Worksheet:

1. Record the Site Name for each site. The site name can contain up to 32 alphanumeric characters, including underscores.This name is the same as the eRCMF site name. For the eRCMF site name, see the eRCMF Global Attributes Worksheet.

2. List the nodes associated with each site. A node can belong to only one site.

3. Record the site Dominance. Set the dominance for one site to yes, and for the other site to no. It does not matter which value you assign to which site.

4. Record the site Backup Communication Method as None.

For XD Volume Sets

For Non-XD Volume Sets

XDNOFCPY NOFCPY Neither site has FlashCopy enabled.

XDALLFCPY ALLFCPY FlashCopy is enabled on both sites.



12

7

Planning for Resource GroupsPlan your resource groups using the guidelines in Chapter 6: Planning Resource Groups in the Planning Guide. In addition to basic resource group planning, plan for the following resource group attributes:• Sites• eRCMF-managed PPRC replicated resources.

Note: If a resource group contains PPRC-mirrored volume groups, the policy Online on All Nodes is not supported.

Site Policy for eRCMF Management

When configuring a resource group to include eRCMF-managed PPRC replicated resources, the resource group includes nodes in both cluster sites. Each volume set will be included in a resource group whose home node is located at the production site of the volume set. In order for HACMP to manage resources, you assign an intersite management policy to each resource group. You can use one of the following site policies for resource groups that include eRCMF-managed PPRC replicated resources:• Prefer Primary Site. (Recommended) In a two-site configuration, replicated resources at

startup are on the site with the higher priority, fall over to the other site, and then fall back to the site with the higher priority.

• Online on Either Site. Replicated resources are on either site at startup, fall over to the other site, and remain on that site after fallover.

eRCMF-Managed PPRC Replicated Resources

The definition for an eRCMF-managed PPRC replicated resource contains the name of the volume set and the mode of mirroring for PPRC associated with the volume set. eRCMF uses information in the Volume_Set_Name.dat file (where Volume_Set_Name is the name of the volume set) to determine which volumes are mirrored pairs for each PPRC replicated resource. PPRC copies the volume information, including the PVID, from one volume in a PPRC pair to the volume at the other site. The volumes at both sites contain the same logical volumes and must therefore be imported with the same volume group name.

Completing the Resource Groups Worksheet

Complete a Resource Groups Worksheet for each resource group that includes eRCMF-managed PPRC replicated resources.

To identify information required to include eRCMF-managed PPRC replicated resources in a resource group, enter the following information on the Resource Groups Worksheet:

1. Record the Resource Group Name for the resource group that will contain eRCMF-managed PPRC replicated resources. The name can contain up to 32 characters and underscores, but without a leading numeric character.



HACMP/XD for Metro Mirror with eRCMF Management Installing eRCMF Management Filesets 7

3. Record the Site Management Policy: Prefer Primary Site or Online on Either Site.Prefer Primary Site is recommended for use with PPRC.

4. Record the names of the nodes in the resource group, listing the nodes in the prioritized order in which they will appear in the nodelist for the resource group.

5. Record the name of each eRCMF-managed PPRC replicated resource (volume set) to be included in the resource group.


Where You Go from HereAfter you plan the configuration for eRCMF-managed PPRC replicated resources in an HACMP cluster:

1. Install the HACMP/XD for Metro Mirror software on HACMP cluster nodes.See Chapter 3: Installing HACMP/XD for Metro Mirror and the next section in this chapter.

2. Configure HACMP sites and HACMP/XD for eRCMF Management.

Installing eRCMF Management FilesetsRefer to Chapter 3: Installing HACMP/XD for Metro Mirrorfor more information on how to install the filesets described in this section.

Software PrerequisitesThe following software is needed for the HACMP/XD solution using eRCMF:• eRCMF version 2.0.5 installed on the PCMs• ESS microcode level vrmf 2.2.x.x. This version provides support for a dual active Copy

Services Server.

eRCMF Management FilesetsThe following filesets are included on the HACMP media, that need to be installed for eRCMF Management.

cluster.es.ercmf

cluster.es.ercmf.rte HACMP/XD for Metro Mirror runtime commands for eRCMF support

cluster.es.ercmf.cmds HACMP/XD for Metro Mirror commands for eRCMF support


HACMP/XD for Metro Mirror with eRCMF Management Installing eRCMF Management Filesets

13

7

Select the appropriate message fileset for your locale(s):

cluster.msg.ja_JP.ercmf


cluster.msg.Ja_JP.ercmf

cluster.msg.En_US.ercmf

Migrating from Direct Management to eRCMF Management SupportIf you want to replace your current configuration that includes PPRC replicated resources with eRCMF support that includes eRCMF-managed PPRC replicated resources, you should have completed the planning steps outlined in the Planning for eRCMF Management of this chapter.

Coexistence of Two Solutions

You can install HACMP/XD for ESS PPRC with eRCMF if you already have HACMP Remote Copy 5.1 or higher installed with HACMP directly managing PPRC pairs. You can also configure HACMP/XD for Metro Mirror to directly manage some of the PPRC pairs and for eRCMF to manage other PPRC pairs.

For the two implementations to coexist, volumes and PPRC pairs can be managed by HACMP with Copy Services Server or by eRCMF, but not by both. If both management mechanisms try to manage the same volumes, the Freeze functionality (at the PPRC pair level) for one solution impacts the other.

Moving to eRCMF Support

To move to eRCMF support:

1. Configure your eRCMF environment.

2. Stop the HACMP cluster.

3. Ensure that the cluster.es.ercmf image was installed.

4. Configure HACMP for HACMP/XD for Metro Mirror with eRCMF.For information about configuring HACMP, see Configuring eRCMF Management in this chapter

cluster.msg.ja_JP.pprc HACMP/XD for MetroMirror for eRCMF support-Japanese IBM-eucJP

cluster.msg.en_US.ercmf HACMP/XD for MetroMirror messages for eRCMF support-U.S. English

cluster.msg.Ja_JP.ercmf HACMP/XD for MetroMirror messages for eRCMF support-Japanese

cluster.msg.En_US.ercmf HACMP/XD for MetroMirror messages for eRCMF support-U.S. English


HACMP/XD for Metro Mirror with eRCMF Management Configuring eRCMF Management 7

5. Remove the existing configuration for PPRC replicated resources.

For information about how to remove the existing configuration, see the section Removing the Existing Configuration.

Removing the Existing Configuration

For each resource group, remove the existing PPRC replicated resource from the HACMP configuration.

If you do not remove the existing configuration, HACMP/XD runs PPRC methods during cluster startup.


2. In the Change/Show Resources and Attributes for a Resource Group panel, remove the PPRC replicated resources listed in the PPRC Replicated Resources field.

3. Select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Configuration > Define PPRC Replicated Resources and press Enter.From this panel, you remove the configuration for PPRC replicated resources by selecting each of the following options:• Define ESS Disk Subsystem• Define PPRC Replicated Resource• Define PPRC Tasks

4. Synchronize and verify the HACMP configuration.

For information about synchronizing a cluster, see the chapter on Verifying and Synchronizing a Cluster Configuration in the HACMP for AIX: Administration Guide.

Where You Go from HereAfter you successfully install the software on each cluster node, you are ready to configure eRCMF Management support.

Configuring eRCMF ManagementBefore configuring HACMP/XD for Metro Mirror for eRCMF Management, ensure that:• eRCMF version 2.0.5 is configured and running on the PCMs at the two eRCMF sites.• You have a good understanding of eRCMF sites and HACMP sites.

For information about eRCMF sites, see the eRCMF User Guide.For information about HACMP sites, see the HACMP Planning Guide.

• HACMP/XD for Metro Mirror software is installed on each cluster node that will support eRCMF-managed PPRC replicated resources.For information about installing HACMP/XD for Metro Mirror, see Chapter 3: Installing HACMP/XD for Metro Mirror.


HACMP/XD for Metro Mirror with eRCMF Management Configuring eRCMF Management

13

7

• The volume groups in the volume sets to be included in the HACMP/XD for Metro Mirror configuration are already defined to AIX.

• The HACMP cluster is configured for:• Nodes• Networks and network interfaces• Service labels.

Overview of Configuration for eRCMF managementTo configure HACMP for eRCMF support:

1. Define HACMP sites for eRCMF support.

2. Define eRCMF global attributes to HACMP.

3. Define eRCMF-managed PPRC replicated resources.

4. Configure resource groups to include specified eRCMF-managed PPRC replicated resources.

Defining HACMP Sites for eRCMF SupportHACMP site events control the HACMP/XD for Metro Mirror support for eRCMF. Therefore, you define two sites as part of the HACMP cluster topology.

Consult your completed HACMP Site Worksheet for the values planned for your cluster.

To define sites in HACMP:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Topology Configuration > Configure HACMP Sites > Add a Site and press Enter.

3. In the Add a Site panel, enter field values as follows and press Enter:

4. Repeat steps 2 and 3 for the second site.

5. Verify and synchronize the cluster.For information about synchronizing a cluster, see Chapter 7: Verifying and Synchronizing an HACMP Cluster in the Administration Guide.

Site Name Enter a name for this site using up to 32 alphanumeric characters and underscores.The site name should be the same as the site name in eRCMF.

Site Nodes Enter the names of the cluster nodes that belong to the site. Leave a space between names. A node can belong to only one site.

Dominance Set one site to yes, and the other to no. It does not matter which value you assign to which site.

Backup Communications

Select None.



13

7

Defining eRCMF Global AttributesFor HACMP to integrate with eRCMF, HACMP requires the values for the global attributes set for eRCMF.

Consult the completed HACMP Site Worksheet for the values planned for your cluster. You can also locate this information in the enterprise configuration file for eRCMF, typically named enterprise.dat and located in the home directory for eRCMF (/usr/ercmf by default) on the PCMs.

To define eRCMF global attributes:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Configuration > Define eRCMF Replicated Resources > Define eRCMF-Managed PPRC Replicated Resources > eRCMF Global Attributes > Add eRCMF Global Attributes and press Enter.

3. In the Add eRCMF Global Attributes panel, enter field values as follows and press Enter:

Note: In HACMP, you can specify only one set of global attributes.

Split Policy Operation mode for eRCMF when a suspending event occurs. Select one of the following:• Freeze&Go. eRCMF splits the sites and frees I/O as soon

as possible. Freeze&Go is the recommended setting and is the default.

• Freeze&Stop. eRCMF splits the sites and blocks I/O until it receives instructions to release the block and allow I/O to resume.

PPRC-XD Recovery Policy

Way in which volume sets fall over to the recovery site when the sites split.Automatic indicates that volume sets automatically fall over to the recovery site.Manual requires that you intervene to have volume groups fall over to the recovery site.

PPRC Links Type Direction in which PPRC links are established. Select one of the following:• TwoWay. Links are established in both directions at the

same time. TwoWay is the default.• OneWay. All links are established in the same direction.

Primary CSS Copy Services Server that is typically active until a failure occurs.This is one of the PCMs.



Defining eRCMF-Managed PPRC Replicated ResourcesDefine eRCMF-managed PPRC replicated resources before adding them to resource groups.

Consult the completed HACMP Site Worksheets for the values planned for your cluster. You can also locate this information in the volume set configuration file for eRCMF. The file is named Volume_Set_Name.dat, where Volume_Set_Name is the name of the volume set. This file is located in the home directory for eRCMF (/usr/ercmf by default) on the PCMs.

To define an eRCMF-managed PPRC replicated resource:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > Define PPRC Replicated Resources > Define eRCMF-Managed PPRC Replicated Resources > eRCMF-Managed PPRC Replicated Resources > Add an eRCMF-Managed PPRC Replicated Resource and press Enter.

3. In the Add an eRCMF-Managed PPRC Replicated Resource panel, enter field values as follows:

4. Verify and synchronize the configuration.For information about synchronizing a cluster, see Chapter 7: Verifying and Synchronizing a Cluster Configuration in the Administration Guide.

Backup CSS Copy Services Server that typically acts as a backup. If the primary Copy Services Server fails, the backup Copy Services Server becomes the active Copy Services Server.This is other PCM.

eRCMF User Id Administrative user ID used to access eRCMF.

eRCMF User Password Password associated with the eRCMF user ID.

Volume Set Name Name of the volume set to be included in the HACMP configuration. This name may contain up to 20 characters.

Volume Set Type Type of the volume set based on the configuration for FlashCopy.For XD volume sets, select one of the following:XDNOFCPY. Neither site has FlashCopy enabled.XDALLFCPY. FlashCopy is enabled on both sites.For non-XD synchronous volume sets, select one of the following:NOFCPY. Neither site has FlashCopy enabled.ALLFCPY. FlashCopy is enabled on both sites.

Production Site Name Site that initially assumes the role of production site for the volume sets. If a failure occurs, the other site becomes the active production site. The picklist shows the HACMP sites configured.



13

7

Configuring Resource GroupsAfter defining eRCMF-managed PPRC replicated resources, you can add them to a resource group.

Configure resource groups following the procedures in Chapter 4: Configuring HACMP Resource Groups (Extended) in the Administration Guide. Make sure that you understand site support for HACMP resource groups; see Chapter 6: Planning Resource Groups in the Planning Guide.

When you configure a resource group ensure that:• The site policy is set to Prefer Primary Site or Online on Either Site.• A Startup policy other than Online on All Available Nodes is specified.• The Resource Group Processing Ordering is set to serial.


To add eRCMF-managed PPRC Replicated resources to your resource group:

1. Enter smit hacmp


3. In the Change/Show Resources and Attributes for a Resource Group panel specify:• The e-RCMF-managed PPRC Replicated Resources to add• The volume groups associated with the eRCMF-managed PPRC replicated resources.

4. Verify and synchronize the cluster.For information about synchronizing a cluster, see Chapter 7: Verifying and Synchronizing a Cluster Configuration in the Administration Guide.

Starting the ClusterAfter you verify and synchronize the HACMP/XD for Metro Mirror configuration changes, start the HACMP cluster.

At cluster startup, the state of an eRCMF-managed PPRC replicated resource is consistent with the state of the volume sets, which is determined by: • The location of the active production site• The state of the PPRC pairs.

The valid, initial state of the PPRC pairs can be one of the following:

This state Indicates

InSync All PPRC pairs are synchronized. InSync is the preferred state for PPRC pairs.



For more information about the status of volume sets, see the eRCMF User Guide.

Changing the ConfigurationUse SMIT to change the configuration for eRCMF-managed PPRC replicated resources.


Any configuration changes you make to any of the following components has an effect on the others listed:• Sites• eRCMF-managed PPRC replicated resources• Volumes• Resource groups.

To change the configuration for eRCMF-managed PPRC replicated resources:

1. Enter smit hacmp

2. In SMIT, select Extended Configuration > Extended Resource Configuration > HACMP Extended Resource Configuration > Define eRCMF Replicated Resources > Define eRCMF-Managed PPRC Replicated Resources and press Enter.From this panel select:• eRCMF-Managed PPRC Replicated Resource• eRCMF Global Attributes

3. After selecting a configuration option, select the Change/Show option for the value you want to change.For information about field values, see Defining eRCMF-Managed PPRC Replicated Resources or Defining eRCMF Global Attributes.

4. Verify and synchronize the configuration.

XD/Mode Extended distance copy (PPRC-XD) mode is active for PPRC pairs. Because PPRC-XD is not synchronous, pairs are in duplex pending XD mode while an application writes updates to a primary volume.

OutofSync The sites are not consistent with each other. The backup site may not be internally consistent. A site resynchronization may be in progress.

RecoverySiteActive The sites are not consistent with each other. A recovery operation is underway, but the site resynchronization has not started. This state initiates a node startup at a recovery site.

This state Indicates


HACMP/XD for Metro Mirror with eRCMF Management Monitoring eRCMF Status

13

7

Note: The verification process copies the current eRCMF configuration files (enterprise.dat, Hosts.dat, and Volume_Set_Name.dat) to each node and verifies the HACMP configuration.

Monitoring eRCMF StatusWhen eRCMF is installed to manage PPRC pairs, HACMP integrates with eRCMF to communicate site events to eRCMF. HACMP calls a number of the eRCMF utilities including the ones in the following list. You can run these utilities independently to get information about eRCMF volume sets.

For more information about these utilities, see the eRCMF User Guide.

eRCMF Utility Purpose

check Checks the consistency of the volume set.

display Displays the volumes in the volume set and their current status.

states Displays the states of all of the volume sets.


Chapter 8: Maintaining and Troubleshooting HACMP/XD for Metro Mirror

This chapter presents general information for maintaining and troubleshooting an HACMP/XD for Metro Mirror configuration. Issues specific to direct management (ESS CLI), to eRCMF, or to DSCLI Management support are noted as such.

For SVC-PPRC Management troubleshooting information, see the Chapter 6: HACMP/XD for Metro Mirror with SVC Management.

This chapter contains the following section:• Logging Messages• Maintaining HACMP/XD for Metro Mirror• Maintaining HACMP/XD for Metro Mirror with eRCMF Support• Troubleshooting DSCLI-Managed Clusters.

Logging MessagesHACMP/XD for Metro Mirror uses the standard logging facilities for HACMP. For information about logging in HACMP, see the HACMP Administration Guide.

In an eRCMF environment, errors are also logged to the eRCMF utilities. For information about logging in eRCMF, see the eRCMF User Guide.

Maintaining HACMP/XD for Metro MirrorThis section describes situations you might encounter in a running cluster after synchronizing your HACMP/XD for Metro Mirror configuration. This information applies to direct and eRCMF management.

Using C-SPOC for Configuration ChangesYou can use C-SPOC to add, change, and delete logical volumes dynamically on nodes connected to the source volumes for a PPRC pair, as long as these operations do not involve allocating additional storage that requires modification of the PPRC pairs. If you make configuration changes that allocate new storage or create logical units, first stop the associated PPRC volume pairs.From C-SPOC, you cannot modify the configuration on nodes connected to target volumes for a PPRC pair.


Maintaining and Troubleshooting HACMP/XD for Metro Mirror Maintaining HACMP/XD for Metro Mirror

14

8

Restarting ESS Copy Services Server on the Backup CSSAn ESS system with a microcode level less then vrmf 2.2.x.x does not support dual active Copy Services Servers. If the primary ESS Copy Services Server fails, you manually start the backup Copy Services Server as the active Copy Services Server for the ESS. You do this through the using the ESS Web Interface. For information about starting Copy Services Server, see the ESS Web Interface User’s Guide.

For information about setting up a Copy Services Server for an HACMP/XD for Metro Mirror configuration, see Planning Copy Services Server on ESS in Chapter 4: HACMP/XD for Metro Mirror with Direct Management.

Detecting ESS FailuresThere is no simple way to detect the failure of an ESS subsystem in a direct management environment. You can check for the loss of quorum for all the volume groups indicated by a LVM_SA_QUORCLOSE error. These errors are written to the AIX error log. HACMP checks whether the LVM_SA_QUORCLOSE error appeared in the AIX 5L error log file and if so, informs the Cluster Manager to selectively move the affected resource group. This HACMP function is called selective fallover for volume group loss. For more information about selective fallover for volume group loss, see the Administration Guide.

Note: If fallover does not occur, check that the LVM_SA_QUORCLOSE error appeared in the AIX 5L error log. When the AIX 5L error log buffer is full, new entries are discarded until space becomes available in the buffer and an error log entry informs you of this problem. For information about increasing the size of the error log device driver internal buffer, see the AIX 5L documentation section in About This Guide.

Using FlashCopy User Exit to Recover DataA rare data exposure may arise if a failure occurs during the resynchronization of PPRC pairs. This condition arises under the following conditions:

1. The links between the ESS systems fail. This causes the PRPC pair to suspend.

2. Changes are logged only on the primary copy only. The second copy is invalid.

3. When the links are restored and the changed data is synchronizing, a failure of the primary volume occurs.

At this point, the primary and secondary volumes are inconsistent. Recovery from this situation is difficult. A user exit provides some protection from this problem. This provides a FlashCopy of the volumes before starting the synchronization. If a failure occurs, you can use FlashCopy to recover data manually.

Note: You can use this user exit if your ESS configuration includes the FlashCopy feature and if the feature is enabled.

To invoke this user exit:

1. Uncomment the following lines near the top of the event handler script:


Maintaining and Troubleshooting HACMP/XD for Metro Mirror Maintaining HACMP/XD for Metro Mirror with eRCMF Support 8

/usr/es/sbin/cluster/pprc/utils/pprc_postvg_offline

/usr/es/sbin/cluster/pprc/utils/pprc_predisk_available

# export FLASHCOPY_USER_EXIT=true

2. Create the appropriate task to FlashCopy the pairs for the volume group using the ESS Copy Services Web Interface. Name the task $vgnameFlashUE, where $vgname is the name of the volume group.

3. Repeat step 2 for any remaining volume groups.

Maintaining HACMP/XD for Metro Mirror with eRCMF Support

eRCMF manages the states of volumes that it controls. You can run the MASTER command on the PCMs to access a number of enterprise and volumeset level commands to get status information. For further information, see the eRCMF User Guide.

Security IssuesHACMP/XD for Metro Mirror with eRCMF support requires that you have root permissions to configure this feature.

Communications between HACMP and eRCMF can be secured at two different levels to protect HACMP processing of eRCMF commands:• Using secure socket layer (SSL) between the HACMP server and the eRCMF server• Requiring a user ID and password be sent with each command.

Doing this requires both of the following:• A user ID and a password is created on an eRCMF server• Informing WebSphere that the specified user is allowed to access eRCMF.

Recovery SupportThe following sections summarize recovery support for HACMP/XD for Metro Mirror with eRCMF support.

Failure of the ESS at the Production Site

If the primary ESS fails, the following occurs:• All disk access at the primary site stops• Loss of a volume group is detected by AIX through an LVM_SA_QUORCLOSE error

which causes a fallover to the other ESS • Notifications are also sent via SNMP messages by eRCMF.

Note: An SNMP agent can be configured to give appropriate notification of events.


Maintaining and Troubleshooting HACMP/XD for Metro Mirror Troubleshooting DSCLI-Managed Clusters

14

8

Failure of the ESS at the Backup Site

If the secondary ESS fails, the primary ESS suspends the PPRC pairs. HACMP does not take any action and eRCMF splits the sites, After the problem is fixed, HACMP initiates a resynchronization.

Failure of ESCON Links between ESS Systems

If some of the ESCON links between the ESS systems fail, PPRC continues to operate. The decrease in bandwidth might affect performance.

If all of the ESCON links between the ESS systems fail, the ESS systems place the PPRC pairs in a suspended state. HACMP does not take additional recovery action. This type of failure initiates sending of an SNMP trap. eRCMF detects this trap, splits the sites, and sends a notification.

Failures of the Active PCM Running eRCMF

If the primary PCM is unavailable, the slave process running on the backup PCM can take over the eRCMF management and continue operations at the backup site until the primary site and the primary PCM is available again. The recover.sh utility script manages this issue.

Troubleshooting DSCLI-Managed ClustersThis section provides information that may help with troubleshooting DSCLI Management clusters. When a DSCLI Management HACMP Cluster is in a less than optimal functioning state, check the /tmp/hacmp.out file for any obvious issues, and then, if necessary, check the states of the PPRC instances and paths that underlay the volume groups to make sure they are write-accessible. Use the commands listed below to discover what the PPRC instances states are, as this will be a major indicator of the health of your volume groups.

If you encounter problems with your cluster that are not covered in this section or are not easily remedied, contact your IBM Service Representative.

Common cluster startup problems1. If a resource group will not stabilize on its primary node and site, it is probable that the

PPRC Replicated Resource and Resource Group are defined to be primary on different sites. Refer to the Planning Primary and Secondary Site Layout for DSCLI-Managed PPRC Replicated Resources in HACMP Resource Groups. Make sure all the appropriate entries in the PPRC replicated resource are aligned in the correct direction, as described in this and the Sample Configuration section.

2. Another reason a resource group may not come ONLINE, may be unstable, or go into ERROR state is if there is trouble with the PPRC instance or path that the volume group and PPRC replicated resource being managed requires. See the descriptions below for commands to use to display PPRC instance states and a description of nominal working states.

3. Keep in mind the limitation that a volume group needs to have the same volume major number across all cluster nodes. This has been known to cause instability in a resource group when it initially comes online.


Maintaining and Troubleshooting HACMP/XD for Metro Mirror Troubleshooting DSCLI-Managed Clusters 8

4. If the Pri-Sec and Sec-Pri port pairs are not declared correctly, this will lead to errors in creating the initial PPRC paths, which will keep the cluster from coming up correctly. In the /tmp/hacmp.out file, there will be an error message returned from the mkpprcpath call for the PPRC replicated resource associated with the resource group that will not come ONLINE.

5. In certain instances, disk reserves can be left on vpaths or hdisks, that DSCLI Management is not able to break. These will keep a resource group from being able to come ONLINE because the disk that has a reserve on it will never be write accessible. See Other Handy AIX Commands below for ideas on how to break disk reserves.

Handy DSCLI Commands to KnowRefer to the DSCLI documentation to become familiar with these management commands. Ideally you will not have to use the DSCLI interface except in troubleshooting situations. In that case, however, the following commands are useful (all run through the /opt/ibm/dscli/dscli command):• lspprc: list the current state of the PPRC instances (see below for valid and common

operational states).• lspprcpath: list the current state of the PPRC paths configured.• lsavailpprcport: list the available PPRC port pairs between two storage systems.• failoverpprc, failbackpprc: switch the direction of PPRC instances.• rmpprc: remove an existing PPRC instance.• rmpmkpprc: create a new PPRC instances.• mkpprcpath: create a new PPRC path.

Refer to the DSCLI documentation for detailed descriptions of these commands, and of others that will allow you to view details of the connected storage systems.

PPRC Instance StatesThe following PPRC instance states exist during normal operation. The states are described in pairs because if you poll the DSCLI at the primary site and the secondary site you will get different state information which reflects the I/O accessibility of the disk at that site, as well as the state of the PPRC instance.

Normal Working State

Primary state: Full Duplex

Secondary state: Target Full Duplex

This is the nominal working state for a PPRC instance. In this state, the Primary is accessible for read and write I/O but the Secondary is not accessible for either.

Suspended State

Primary state: Full Duplex

Secondary state: Suspended


Maintaining and Troubleshooting HACMP/XD for Metro Mirror Troubleshooting DSCLI-Managed Clusters

14

8

In this state, the Primary is accessible for read and write I/O and the Secondary is as well. Data integrity between the two sides of the PPRC instance is not ensured in this state.

In the context of HACMP clusters, this PPRC instance state occurs when there has been a failure on the HACMP site connected to the primary site, causing a failover to the secondary state.

Copy State

Primary and secondary state: Copy Pending

In this state, there is a full disk copy in process between one side of the PPRC instance and the other.

Handy AIX CommandsRefer to the AIX documentation for more details on the following commands:

To discover whether a disk reserve exists on a given disk:lquerypr -Vh /dev/<disk name>

To clear a disk reserve that exists on a given disk:lquerypr -ch /dev/<disk name>

To list the existing vpath to storage volume ID (LUN + storage ID) mapping:lsvpcfg

To list the existing vpaths on your system, which will display the volume group mapping:lspv | grep vpath

To list the existing volume groups on your system:lsvg

To list the existing file systems on your system:lsfs

To list the available volume major numbers on your system:lvlstmajor


Appendix A: General HACMP Planning Worksheets

You should print and use the paper planning worksheets from the PDF version of this guide instead of the HTML version. The PDF version has each worksheet aligned to start at the top of a page. You may need more than one copy of some of the worksheets.

This appendix contains the following worksheets:

Worksheet Use the worksheet to record information about ...

HACMP Site Worksheet HACMP sites to support PPRC replicated resources

Resource Groups for PPRC Replicated Resources Worksheet

Resource groups that contain PPRC replicated resources


General HACMP Planning Worksheets HACMP Site Worksheet

14

A

HACMP Site Worksheet

Site 1 Site 2

Site Name ________________________ ________________________

Site Nodes ________________________ ________________________

________________________ ________________________

________________________ ________________________

________________________ ________________________

________________________ ________________________

________________________ ________________________

Site Dominance _____ _____

Site Backup Communication Method

None None


General HACMP Planning Worksheets Sample HACMP Site Worksheet A

Sample HACMP Site Worksheet

Production Site Backup Site

Site Name SiteA SiteB

Site Nodes n3 n1

n4 n2

Site Dominance yes no


None None


General HACMP Planning Worksheets Resource Groups for PPRC Replicated Resources Worksheet

14

A

Resource Groups for PPRC Replicated Resources Worksheet

* Circle one: P-primary, or S-secondary

Resource Group Name _______________________________

Management Policies:

Startup _______________________________

Fallover _______________________________

Fallback _______________________________

Intersite Management Policy _______________________________

Nodes (in prioritized order as they will appear in the nodelist)

_______________________________

_______________________________

_______________________________

_______________________________

_______________________________

_______________________________

_______________________________

Name of PPRC Replicated Resource

____________________________ P or S

____________________________ P or S

____________________________ P or S

____________________________ P or S

____________________________ P or S

____________________________ P or S


General HACMP Planning Worksheets Sample Resource Groups for PPRC Replicated Resources Worksheet A

Sample Resource Groups for PPRC Replicated Resources Worksheet

* Circle one: P-primary, or S-secondary

Resource Group Name rg_pprc_1


Startup Online on home node only

Fallover Fallover to next priority node in the list

Fallback Fallback to higher priority node in the list

Intersite Management Policy Prefer primary site


n3

n4

n1

n2

Name of PPRC Replicated Resource

pprc4_1 (P) or S

pprc4_2 (P) or S

pprc5_1 (P) or S

pprc5_2 (P) or S


General HACMP Planning Worksheets Sample Resource Groups for PPRC Replicated Resources Worksheet

15

A


Appendix B: Planning Worksheets for Direct Management (ESS CLI)




PPRC-Mirrored Volumes Worksheet

PPRC pairs and the associated PPRC replicated resources

User-Specific Tasks Worksheets

User-specific task names (not needed for most configurations)

ESS Disk Subsystem Worksheet

ESS disk subsystems that contains PPRC pairs

Resource Groups for PPRC replicated resources

See Appendix A: General HACMP Planning Worksheets.


Planning Worksheets for Direct Management (ESS CLI) PPRC-Mirrored Volumes Worksheet

15

B


*P/S—Specify primary or secondary

Note: Enter the volume information so that the information for each volume in a PPRC pair appears on the same line.

Site Name ________________ Site Name ________________

ESS Serial Number_______________


Volume Group Name

PPRC Replicated Resource Name

P/S*VolumesVolume ID Logical

Subsystem IDP/S*

VolumesVolume ID Logical

Subsystem ID

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ ______ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________


Planning Worksheets for Direct Management (ESS CLI) Sample PPRC-Mirrored Volumes Worksheet B

Sample PPRC-Mirrored Volumes Worksheet



Site Name Site_1 Site Name Site_2

ESS Serial Number 22011 ESS Serial Number 22012

Volume Group Name



Subsystem IDP/S*


Subsystem ID

vg4 pprc4.1 P 70D22011 17 S 60D22012 16

vg4 pprc4.2 P 70E22011 17 S 60E22012 16

vg5 pprc5.1 P 70F22011 17 S 60F22012 16

vg5 pprc5.2 P 71022011 17 S 61022012 16


Planning Worksheets for Direct Management (ESS CLI) User-Specific Tasks Worksheets

15

B

User-Specific Tasks WorksheetsUse this worksheet only if you use task names which are different from the recommended ones.

Task Name Description

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________

________________ __________________________________________________


Planning Worksheets for Direct Management (ESS CLI) Sample User-Specific Tasks Worksheet B

Sample User-Specific Tasks WorksheetUse this worksheet only if you use task names which are different from the recommended ones.

Task Name Description

vg4EstPtPS This name is the same as the recommended name.

vg4DelPtPSF This name is the same as the recommended name.

vg4DelPtPS This name is the same as the recommended name.

vg4EstPtSP This name is the same as the recommended name.

vg4DelPtSP This name is the same as the recommended name.

vg4DelPtSPF This name is the same as the recommended name.

vg4EstPrPSFCex Example task to establish a PPRC pair from the source volume on ESS 22011 to the target volume on ESS 22012.

vg4EstPrPSNC This name is the same as the recommended name.

vg4EstPrPSSC This name is the same as the recommended name.

vg4EstPrPSFO This name is the same as the recommended name.

vg4EstPrPSFB This name is the same as the recommended name.

vg4SusPrPSP This name is the same as the recommended name.

vg4SusPrPSS This name is the same as the recommended name.

vg4TerPrPSP This name is the same as the recommended name.

vg4TerPrPSS This name is the same as the recommended name.

vg4EstPrSPFC This name is the same as the recommended name.

vg4EstPrSPNC This name is the same as the recommended name.

vg4EstPrSPSC This name is the same as the recommended name.

vg4EstPrSPFO This name is the same as the recommended name.

vg4EstPrSPFB This name is the same as the recommended name.

vg4TerPrSPS This name is the same as the recommended name.

vg4TerPrSPP This name is the same as the recommended name.

vg4SusPrSPS This name is the same as the recommended name.

vg4SusPrSPP This name is the same as the recommended name.


Planning Worksheets for Direct Management (ESS CLI) ESS Disk Subsystem Worksheet

15

B

ESS Disk Subsystem Worksheet

Site Name ______________ Site Name ______________

ESS Disk Subsystem Name _______________________ _______________________

ESS IP Address _______________________ _______________________

ESS User ID _______________________ _______________________

ESS Password _______________________ _______________________


Planning Worksheets for Direct Management (ESS CLI) Sample ESS Disk Subsystem Worksheet B

Sample ESS Disk Subsystem Worksheet

Site Site_1 Site Site_2

ESS Disk Subsystem Name ESS_1 ESS_2

ESS IP Address 192.0.2.12 192.0.24.12

ESS User ID s1_root_user s2_root_user

ESS Password s1_root_user_passw s2_root_user_passw


Planning Worksheets for Direct Management (ESS CLI) Sample ESS Disk Subsystem Worksheet

15

B


Appendix C: Planning Worksheets for DSCLI Management




Copy Services Server Worksheet

Copy Services Server to support PPRC DCSLI Management


PPRC pairs and the associated PPRC replicated resources

DS ESS Disk Subsystem Worksheet

ESS disk subsystems that contain PPRC pairs

PPRC Replicated Resources Worksheet

PPRC replicated resources that go in HACMP resource groups


Planning Worksheets for DSCLI Management Copy Services Server Worksheet

16

C

Copy Services Server Worksheet

CSS (HACMP) Site Name CSS (HACMP) Site Name

CSS Subsystem name

CSS Type DSCLI DSCLI

CLI IP Address

CSS User ID .

CSS password


Planning Worksheets for DSCLI Management Sample Copy Services Server Worksheet C

Sample Copy Services Server WorksheetSite A Site B

CSS subsystem name m222h m555h

CLI Type DSCLI DSCLI

CSS IP Address 9.112.112.2 9.115.115.2

CSS User ID .xyz xyz

CSS password 123xyz 123xyz


Planning Worksheets for DSCLI Management PPRC-Mirrored Volumes Worksheet

16

C




Site Name ________________ Site Name ________________



Volume Group Name



Subsystem IDP/S*


Subsystem ID

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ ______ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________

________ __________ _____ ___________ ___________ _____ ___________ ___________


Planning Worksheets for DSCLI Management Sample PPRC-Mirrored Volumes Worksheet C

Sample PPRC-Mirrored Volumes Worksheet



Site Name Site_1 Site Name Site_2

ESS Serial Number 22011 ESS Serial Number 22012

Volume Group Name



Subsystem IDP/S*


Subsystem ID

vg4 pprc4.1 P 70D22011 17 S 60D22012 16

vg4 pprc4.2 P 70E22011 17 S 60E22012 16

vg5 pprc5.1 P 70F22011 17 S 60F22012 16

vg5 pprc5.2 P 71022011 17 S 61022012 16


Planning Worksheets for DSCLI Management DS ESS Disk Subsystem Worksheet

16

C

DS ESS Disk Subsystem Worksheet

Site Name ______________

Site Name ______________

ESS Disk Subsystem Name ______________________ ______________________

ESS Cluster 1 IP Address ______________________ ______________________

ESS Cluster 2 IP Address ______________________ ______________________

ESS User ID ______________________ ______________________

ESS Password ______________________ ______________________

ESS Storage ID ______________________ ______________________

Associated CSS ______________________ ______________________


Planning Worksheets for DSCLI Management Sample DS ESS Disk Subsystem Worksheet C

Sample DS ESS Disk Subsystem Worksheet

Site Name _Site_A_____________

Site Name _Site_B_____________

ESS Disk Subsystem Name __m222____________ __m555____________

ESS Cluster 1 IP Address 9.22.22.22 9.55.55.55

ESS Cluster 2 IP Address 9.44.44.44 9.77.77.77

ESS User ID root_xyz___________ root_zyx____________

ESS Password 123xyz_____________ 123zyx______________

ESS Storage ID IBM.2107-2222222 IBM.2107-5555555

Associated CSS m222h____________ m555h______________


Planning Worksheets for DSCLI Management PPRC Replicated Resources Worksheet

16

C

PPRC Replicated Resources WorksheetFill in one table per PPRC resource to be configured. Note that the information in all rows except the PPRC Volume Pairs (last row) will be the same for each replicated resource.

Primary Secondary

Storage ID

HACMP Sites

ESS Pair

LSS Pair

PPRC Link Type This information will be the same for both sites.

PPRC Critical Mode


Volume Group

Port Pair ID-1

Port Pair ID-2

PPRC Volume Pairs

Continue adding pairs as needed...


Planning Worksheets for DSCLI Management Sample PPRC Replicated Resource Worksheet C

Sample PPRC Replicated Resource WorksheetThis sample shows the information for several PPRC volume pairs (replicated resources).

Note the information is the same except for the PPRC Volume Pair numbers. In SMIT, you will need to add all the information for each replicated resource.

Primary Secondary

Storage ID IBM.2107-111111 IBM.1750-222222

HACMP Sites Site A Site B

ESS Pair 10 10

LSS Pair DS8000 DS6000

PPRC Link Type FCP

on

Automatic

VG1

PPRC Critical Mode


Volume Group

Port Pair ID -1 10020 10101

Port Pair ID-2 11024 10001

PPRC Volume Pairs 1000 3033

1001 3034

1002 3035

1003 3036

1004 3037

1005 3038

1006 3039


Planning Worksheets for DSCLI Management Sample PPRC Replicated Resource Worksheet

16

C


Appendix D: Planning Worksheets for SVC PPRCYou should print and use the paper planning worksheets from the PDF version of this guide instead of the HTML version. The PDF version has each worksheet aligned to start at the top of a page. You may need more than one copy of some of the worksheets.


Worksheet Use the worksheet to record information about...

SVC Cluster Worksheet Information for defining SVC clusters to be accessed by HACMP

SVC PPRC Mirrored Volumes

Cataloging information for building SVC PPRC relationships, including how the SVC vdisks relate to AIX disks and volume groups.

SVC PPRC Replicated Resource

Listing the SVC Consistency Group names (and their related SVC Clusters) that will be used as the PPRC Replicated Resource names when defining HACMP Resource Group components.

HACMP Resource Group Worksheet

See Appendix A: General HACMP Planning Worksheets.


Planning Worksheets for SVC PPRC SVC Cluster Worksheet

17

D

SVC Cluster Worksheet

HACMP Site HACMP Site

SVC Cluster Name

SVC Cluster Role(Master or Auxiliary)

SVC Cluster IP Address

Remote SVC Partner .


Planning Worksheets for SVC PPRC Sample SVC Cluster Worksheet D

Sample SVC Cluster Worksheet

Site A Site B

SVC Cluster Name svc_Cl1 svc_Cl2

SVC Cluster Role(Master or Auxiliary)

Master Auxiliary

SVC Cluster IP Address 10.10.10 11 10.10.10.12

Remote SVC Partner .svc_c12 svc_11


Planning Worksheets for SVC PPRC SVC PPRC Relationships Worksheet

17

D

SVC PPRC Relationships WorksheetFill in the worksheet so that the information for each volume group appears on the same line.

Site name Site name

SVC Cluster Name SVC Cluster name

VG name

VG Major Num

AIX vpath name

CVC PPRC Rel. name

SVC Disk Name & Role SVC Disk name & Role


Planning Worksheets for SVC PPRC Sample PPRC Relationships Worksheet D

Sample PPRC Relationships Worksheet

Site A Site B

SVCCluster 1 SVCCluster 2

VG name

VG Major Num

AIX vpath name

CVC PPRC Rel. name

SVC Disk Name & Role

SVC Disk name & Role

VG1 10 vpath1 svcrel1 vDisk1_1 Master vDisk1_1 Auxiliary

VG2 11 vpath2 svcrel2 vDisk2_2 Master vDisk2_2 Auxiliary

VG3 12 vpath3 svcrel3 vDisk2_3 Auxiliary vDisk2_3 Master

VG4 13 vpath4 svcrel4 vDisk2_4 Auxiliary vDisk2_4 Master


Planning Worksheets for SVC PPRC SVC PPRC Replicated Resources Worksheet

17

D

SVC PPRC Replicated Resources Worksheet

SVC PPRC Consistency Group Name(rep. resource name)

HACMP Resource Group


Assoc. HACMP Site


Assoc. HACMP Site

List of SVC relationships


Planning Worksheets for SVC PPRC Sample SVC PPRC Resources Worksheet D

Sample SVC PPRC Resources Worksheet

SVC PPRC Consistency Group Name(rep. resource name)

HACMP Resource Group


Assoc. HACMP Site


Assoc. HACMP Site

List of SVC relationships

SVCcg1 HArg1 svcCL1 Site A svcCL2 Site B svcRel1 svcRel2

SVCcg2 HArg2 svcCL2 Site B svcCL1 Site A svcRel3 svcRel4


Planning Worksheets for SVC PPRC Sample SVC PPRC Resources Worksheet

17

D


Appendix E: Planning Worksheets for an eRCMF Environment




eRCMF Global Attributes Worksheet

General attributes configured for eRCMF

Volume Sets Worksheet Volume sets configured for eRCMF

eRCMF-Managed Resources Site Worksheet

HACMP sites to support eRCMF-managed PPRC replicated resources

Resource Groups Worksheet

Resource groups that contain eRCMF-managed PPRC replicated resources


Planning Worksheets for an eRCMF Environment eRCMF Global Attributes Worksheet

17

E

eRCMF Global Attributes Worksheet

Full pathname of the Enterprise Configuration File

__________________________________

Name of the Primary eRCMF site __________________________________

Name of the Secondary eRCMF site __________________________________

Site Split Behavior __________________________________

PPRC-XD Recovery Policy __________________________________

PPRC Link Types __________________________________

Primary CSSThe PCM that is initially active

__________________________________

Backup (secondary) CSSThe PCM that is initially the backup

__________________________________

eRCMF User Id __________________________________

eRCMF User Password __________________________________


Planning Worksheets for an eRCMF Environment Sample eRCMF Global Attributes Worksheet E

Sample eRCMF Global Attributes Worksheet

Full pathname of the Enterprise Configuration File

/usr/ercmf/enterprise.dat

Name of the Primary eRCMF site Site1

Name of the Secondary eRCMF site Site2

Site Split Behavior Freeze&Go

PPRC-XD Recovery Policy Automatic

PPRC Link Types TwoWay

Primary CSSThe PCM that is initially active

PCM_1

Backup (secondary) CSSThe PCM that is initially the backup

PCM_2

eRCMF User Id User1

eRCMF User Password xxxx


Planning Worksheets for an eRCMF Environment Volume Sets Worksheet

18

E

Volume Sets Worksheet

eRCMF Production Site ________________ eRCMF Backup Site ________________

Volume Set Name ________________

Volume Set Configuration File

________________

Volume Set Type ________________

Volume Groups ________________

________________

________________

________________



________________



________________

________________

________________



________________



________________

________________

________________


Planning Worksheets for an eRCMF Environment Sample Volume Sets Worksheet E

Sample Volume Sets Worksheet

eRCMF Production Site Site1 eRCMF Backup Site Site2

Volume Set Name vset1


/usr/ercmf/vset1.dat

Volume Set Type XDALLFCPY

Volume Groups vg5

vg10

________________

________________

Volume Set Name vset2


/usr/ercmf/vset2.dat

Volume Set Type XDALLFCPY

Volume Groups vg50

________________

________________

________________



________________



________________

________________

________________


Planning Worksheets for an eRCMF Environment eRCMF-Managed Resources Site Worksheet

18

E

eRCMF-Managed Resources Site Worksheet

eRCMF Production Site eRCMF Backup Site

Site Name ________________________ ________________________

Site Nodes ________________________ ________________________

________________________ ________________________

________________________ ________________________

________________________ ________________________

________________________ ________________________

________________________ ________________________

Site Dominance _____ _____


None None


Planning Worksheets for an eRCMF Environment Sample eRCMF-Managed Resources Site Worksheet E

Sample eRCMF-Managed Resources Site Worksheet

eRCMF Production Site eRCMF Backup Site

Site Name Site1 Site2

Site Nodes n3 n1

n4 n2

Site Dominance yes no


None None


Planning Worksheets for an eRCMF Environment Resource Groups Worksheet

18

E

Resource Groups Worksheet

Resource Group Name _______________________________


Startup _______________________________

Fallover _______________________________

Fallback _______________________________

Intersite Management Policy _______________________________


_______________________________

_______________________________

_______________________________

_______________________________

_______________________________

_______________________________

_______________________________

Name of eRCMF-Managed PPRC Replicated Resources Included

_______________________________

_______________________________

_______________________________

_______________________________

_______________________________

_______________________________

Volume Groups _______________________________

_______________________________

_______________________________

_______________________________


Planning Worksheets for an eRCMF Environment Sample Resource Groups Worksheet E

Sample Resource Groups Worksheet

Resource Group Name rg_pprc_1


Startup Online on home node only

Fallover Fallover to next priority node in the list

Fallback Fallback to higher priority node in the list

Intersite Management Policy Prefer primary site


n3

n4

n1

n2

Name of eRCMF-managed PPRC Replicated Resources Included

ercmf_1

ercmf_2

Volume Groups vg5

vg10

vg50


Planning Worksheets for an eRCMF Environment Sample Resource Groups Worksheet

18

E


GlossaryThis Glossary contains definitions of terms used in the HACMP/XD for Metro Mirror: Planning and Administration Guide. The definitions assume a general familiarity with terminology for HACMP and for the IBM TotalStorage Enterprise Storage Server and ESS Copy Services.Also see the Master Glossary.

CCopy Services

See ESS Copy Services.

DDirect management of PPRC pairs

HACMP manages the fallover and resynchronization of the PPRC pairs by issuing ESS CLI commands directly to the ESS systems.

DSCLIInterface for direct management system of PPRC pairs on IBM Models DS6000, DS8000 and some others. (See ESSNI Server.) Using this configuration interface eliminates the need to define static PPRC tasks.

EEnterprise Remote Copy Management Facility (eRCMF)

A multisite disaster recovery solution for open systems that manages and monitors PPRC pairs.

Enterprise System Connection (ESCON)An Enterprise Systems Architecture/390 and zSeries computer peripheral interface. The I/O interface uses ESA/390 logical protocols over a serial interface that configures attached units to a communication fabric.

ESCON links provide serial connections between two ESS systems to mirror data. ESCON links carry many paths, with each path transmitting data in one direction.

eRCMFSee Enterprise Remote Copy Management Facility.

eRCMF-Managed PPRC Replicated Resources

PPRC replicated resources managed by eRCMF.

ESSSee IBM TotalStorage Enterprise Storage Server.

ESS Copy ServicesIn the ESS, a collection of optional software features, with a Web browser interface, used for configuring, managing, and monitoring data-copy functions such as PPRC.

ESSNetA private management network that also provides Web browser access to an IBM TotalStorage Enterprise Storage Server.

ESSNI ServerThe interface for managing DS6000 and DS8000 disk systems. ESSNI servers also enable the management of ESS 2105 or ESS 1750. For Model 2107, the ESSNI server runs on the HMC. For Model 1750, a management node runs on a customer machine. In all cases, the DSCLI is invoked with the ESSNI addresses to connect to a particular domain. For Model 2105, there can be two ESSNI servers for a CS domain, which are run on clusters of the ESSs. Though Model 2105 can have two ESSNI servers, there is no added benefit of connecting to both. The servers are peers, but do not back each other up (a request sent to one won't be handled by the second if the first fails).


18

F

FFlashCopy

An optional feature for the ESS that can make an instant copy of data, that is, a point-in-time copy of a volume.

GGlobal Copy (eXtended Distance PPRC)

Copies data non-synchronously and over longer distances than is possible with Metro Mirror. When operating in Global Copy mode, the source volume sends a periodic, incremental copy of updated tracks to the target volume instead of a constant stream of updates. This causes less impact to application writes for source volumes and less demand for bandwidth resources, while allowing a more flexible use of the available bandwidth.

Global Mirror (Asynchronous PPRC) Provides a long-distance remote copy feature across two sites using asynchronous technology. It supports data consistency across all volumes that are defined for mirroring. The volumes can reside on several different storage units and can mirror the volumes over a thousand kilometers from the source site to the target site.

IIBM TotalStorage Enterprise Storage Server (ESS)

A high-performance, fault-tolerant storage and management system for enterprise data, providing access through multiple concurrent operating systems and communication protocols. High performance is provided by multiple symmetrical multiprocessors, integrated caching, RAID support for the disk drive modules, and disk access through a high-speed serial storage architecture (SSA) interface.

LLogical units (LUNs)

A unit of storage within a volume group on an IBM TotalStorage Enterprise Storage Server.

MMetro Mirror (Synchronous PPRC)

Provides real-time mirroring of logical volumes between two ESS 800, DS6000s or DS8000s that can be located up to 300km from each other. It guarantees that the secondary copy is up-to-date by ensuring that the primary copy will be written only if the primary system receives acknowledgment that the secondary copy has been written. If the acknowledgment of the remote write is not received within a fixed period of time, the write is considered to have failed, and is rendered ineligible for storage to disk. At this point, the application receives an I/O error, and in due course, the failed write I/O is “aged-out” of each NVS.

PPCM

See Productivity Center Machine.

Peer-to-Peer Remote Copy (PPRC)A synchronous protocol that allows real-time mirroring of data from one Logical Unit (LUN) in an IBM TotalStorage Enterprise Storage Server (ESS) to another LUN in a second ESS system. The second ESS can be located at another site at some distance away. PPRC is a data-copy service on an ESS system.

Primary volumeOne volume in a pair of PPRC-mirrored volumes. Typically, a primary volume is the source volume mirrored to the target volume. However, after a fallover, the primary volume may be the target volume.

Productivity Center MachineThe system that runs the eRCMF server. The machine has a communication path to ESSNet.

PPRCSee Peer-to-Peer Remote Copy.


R

PPRC instanceRefers to the PPRC relationship that exists between a pair of disk volumes.

PPRC pairTwo PPRC-mirrored volumes, residing on different IBM TotalStorage Enterprise Storage Servers.

PPRC pathThe communication path between two PPRC-mirrored volumes over an ESCON or Fibre Channel link.

PPRC replicated resourceA PPRC pair associated with an HACMP site and defined to HACMP. A PPRC replicated resource can be added as a resource to an HACMP resource group.

RRemote Mirror and Copy

Newer term for PPRC.

SSAN Volume Controller (SVC)

The IBM TotalStorage SAN Volume Controller (SVC) is a virtualization appliance solution that maps virtualized volumes visible to hosts and applications to physical volumes on storage devices. It is an in-band implementation that minimizes the dependency on unique hardware and software, telecopying the storage functions expected in a SAN environment from the storage subsystems and managing storage resources. In addition, it provides advanced copy services for data migration and business continuity similar to the Peer to Peer Remote Copy function in the ESS and DS4000 (formerly named the FAStT) series Storage subsystems. Since the copy services operate on the virtual volumes, dramatically simpler replication configurations can be created using the SAN Volume Controller, rather than replicating each physical volume in the managed storage pool.

SiteIn HACMP, a data center that contains volume groups on an IBM TotalStorage Enterprise Storage Server (ESS) that will be mirrored over

some distance to another ESS that is defined as a second site. Nodes are defined as belonging to sites; the site becomes a component, like a node or a network, that is known to the HACMP software.eRCMF also has a concept of sites. An HACMP site has some of the same components, and uses the same name as an eRCMF site.

Secondary volumeOne volume in a pair of PPRC-mirrored volumes. Typically, a secondary volume is the target volume mirrored from the source volume. However, after a fallover, the secondary volume may be the source volume.

Source volumeThe volume written to by an application.

SVC PPRCVersion of HACMP/XD for Metro Mirror software for use with the SAN Volume Controller.

TTarget volume

The volume to which data is copied from the source volume.

TaskA set of actions to be executed on an IBM TotalStorage Enterprise Storage Server. Tasks let you automate a series of steps that otherwise would be performed in the ESS Copy Services Web Interface.

VVolume set

A group of volumes to be managed as a unit by eRCMF.


19

V


Notices for HACMP/XD for Metro Mirror: Planning and Administration Guide

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

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

IBM may have patents or pending patent applications covering subject matter described in this document. The furnishing of this document does not give you any license to these patents. You can send license inquiries, in writing, to:

IBM Director of Licensing IBM Corporation North Castle Drive Armonk, NY 10504-1785 U.S.A.

For license inquiries regarding double-byte (DBCS) information, contact the IBM Intellectual Property Department in your country or send inquiries, in writing, to:

IBM World Trade Asia Corporation Licensing 2-31 Roppongi 3-chome, Minato-ku Tokyo 106, Japan

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

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

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


Notices for HACMP/XD for Metro Mirror: Planning and Administration Guide

19

Licensees of this program who wish to have information about it for the purpose of enabling: (i) the exchange of information between independently created programs and other programs (including this one) and (ii) the mutual use of the information which has been exchanged, should contact:

IBM Corporation Dept. LRAS / Bldg. 003 11400 Burnet Road Austin, TX 78758-3493 U.S.A.

Such information may be available, subject to appropriate terms and conditions, including in some cases, payment of a fee.

The licensed program described in this document and all licensed material available for it are provided by IBM under terms of the IBM Customer Agreement, IBM International Program License Agreement or any equivalent agreement between us.

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

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


Index +-*/

/usr/es/sbin/cluster/svcpprc 109/usr/es/sbin/cluster/svcpprc/utils/cl_verify_svcpprc_co

nfigSVC PPRC verification command 113

/usr/sbin/lvlstmajor command 78

Aactive site 22adding

DSCLI-managed PPRC replicated resources 89SVC cluster definition 110

AIX commands 144auxiliary

SVC terminology 98

Bbackup site 22

Cchanging

configuration with C-SPOC 139definition of SVC PPRC relationship 115DSCLI configuration 95DSCLI replicated resource configuration 94PPRC replicated resource configuration 69, 137site configuration 70, 95SVC cluster definition 115SVC PPRC resources 117

cl_convert command 40cl_sync_pprc_config command 68cl_verify_pprc_config command 68cl_verify_svcpprc_config command 120clconvert_snapshot command 40CLI Type

DSCLI for ESS 2107 or ESSCLI for 2105 88DSCLI or ESSCLI 79

cllsrelationship command 120cllssvc command 119cllssvcpprc command 120cluster startup problems

troubleshooting 142cluster, starting 69cluster.es.spprc filesets 86cluster.license fileset 31cluster.xd.license fileset 31

commandscl_sync_pprc_config 68cl_verify_pprc_config 68cl_verify_svcpprc_config 120cllsrelationship 120cllssvc 119cllssvcpprc 120cllsvcpprc 120lscfg 48lspv 48rsExecuteTask.sh 60rsList2105.sh 48rsPrimeServer 48

configurationchanging for PPRC replicated resources 69, 94,

137overview of 55PPRC replicated resources 61, 87PPRC tasks 56, 69prerequisites 55prerequisites for eRCMF 131resource groups with DSCLI 91sites 132

configuration database. See HACMP configuration database.

configuringCopy Services Server (spprc) 88PPRC for Direct Management 56SVC PPRC 109

connections to ESSNet 46Consistency Group Name

SVC replicated resource 99consistency groups

PPRC volume states 118SVC PPRC resources 112, 117viewing current state 119

Copy Services Server Worksheet 160, 161Copy Services Servers

adding to spprc configuration 88planning for DSCLI Management 79

C-SPOC 139limitations for DSCLI Management 73

CSS IP Address 79


Ddefining

eRCMF-managed PPRC replicated resources 135ESS disk subsystem for DSCLI 88ESS disk subsystem to HACMP 62PPRC pairs (Direct mnagement) 62PPRC replicated resources 61PPRC tasks (Direct management) 64SPPRC replicated resources 87SVC PPRC relationship 111SVC PPRC resources 111

Direct Managementconfiguring PPRC 56definition of PPRC replicated resources 21filesets 32planning worksheets 45prerequisites for planning 44sample configuration 44software requirements 53

disaster recoveryfeatures of HACMP/XD for Metro Mirror 16

disk subsystemsplanning for DSCLI Management 79

DS ESS Disk Subsystem Worksheet 164, 165DS4000 series 97, 189DSCLI

filesets 32overview 19

DSCLI commands 143mkpprc 78mkpprcpath 78rmpprc 78

DSCLI Managementadding replicated resource 89changing configuration of resources 95configuration prerequisites 87configuring replicated resources 87C-SPOC limitations 73defining ESS disk subsystem to HACMP 88installing filesets 85overview 71planning CSS systems 79planning disk susbsystems 79planning overview 72planning sites for replicated resources 80removing configuration 95resource groups 91resource groups worksheet 84restrictions 72setting up volume groups and filesystems 77steps for configuring 87troubleshooting tips 142

DSCLI-managed PPRC replicated resourcesdefinition 21

EeRCMF

configuring resource groups 136definiing sites 132filesets 32global attributes 134overview 20prerequisites for configuration 131site splits behaviors 124worksheets 145, 177

eRCMF managementchanging the configuration 137initial state of PPRC pairs 136limitations 121monitoring site events 138planning integration with HACMP 123planning resource groups 128planning sites 124sample configuration 122security 141starting the cluster 136

eRCMF-managed PPRC replicated resourcesdefinition 21

ESCON links 46defining for Direct management 57definition 187

ESS 62, 188defining to HACMP for DSCLI 88detecting failures 140ESS Copy Services 17, 49, 187

planning 46ESSNet 187

node connections 46identifying system 51planning for sites 52

ESS Copy Services Serverdefining tasks 57failure handling 140

ESS Copy Services Web Interfacefor Direct Management 44

ESS Disk Subsystem Worksheet 156, 157, 164, 175ESS Disk Subsystem Worksheet (Direct Management)

51ESSNet 17example of HACMP Remote Copy configuration 122

Ffallback

with HACMP/XD for Metro Mirror 22fallover

with HACMP/XD for Metro Mirror 22


filesetsDirect Management 54for each type of PPRC management 32

filesystemssetting up for DSCLI management 77

FlashCopy 188User Exit for handling ESS failures 140

Freeze & Go 124Freeze & Stop 124

Gglobal attributes

configuring for eRCMF management 125eRCMF 134

Global Attributes Worksheetfor eRCMF management 125

Global Copytype of mirroring 17

Global Mirrortype of mirroring 17

HHACMP

fileset prerequisites for HACMP/XD 31planning sites 27sites

for HACMP/XD for Metro Mirror 22upgrade paths 38upgrading

overview 38with PPRC mirrored volumes 20

HACMP configuration databaseconverting for upgrade 40

HACMP Remote Copyexample configuration 122

HACMP/XD for Metro Mirrorfeatures 16installing 31installing for Direct Management. 52overview 15overview of management options 26

HACMP/XD SVC PPRCoverview 97

hard disk, installing from 35

IIBM TotalStorage Disk subsystems

working with Metro Mirror 17IBM TotalStorage SAN Volume Controller 97identifying ESS system 51images for installation 33

installation mediafilesets 33

installingDirect Management 52eRCMF filesets 129from

CD-ROM 36from a hard disk 35from installation media 36from installation server 35HACMP/XD for Metro Mirror 31installation choices 35on server node 36optional filesets 32prerequisites 31SCV PPRC cluster 109spprc filesets 85

integratingeRCMF management and HACMP 123

intersite management policy 21, 29

Llimitations

DSCLI Management 72for eRCMF management 121HACMP/XD for Metro Mirror

SVC Management 99resource groups with PPRC 29

logical subsystem identifier 49logical unit number 17logical units 188lscfg command 48lspv command 48LUN 17

corresponding vpath and hdisk 77

Mmaster

SVC terminology 98Metro Mirror

supported for DSCLI v.5.3 and 5.4 73type of mirroring 17

migratingfrom direct management to eRCMF 130

mkpprc command 78mkpprcpath command 78mutual recovery configuration

DSCLI management 74

Nnames for PPRC tasks 56


Ooverview

configuration 55planning for Direct management 43planning for eRCMF management 121planning for HACMP/XD for Metro Mirror 27

Pplanning

connections to ESS(Direct management) 46ESS disk subsystems 51for DSCLI Management 72overview 27overview of Direct management 43overview of eRCMF management 121PPRC paths 47PPRC tasks 49PPRC-SVC Management support 103prerequisities for Direct Management 44resource groups 28SVC management 98SVC PPRC replicated resources 107worksheets 27

Direct Management 45worksheets for eRCMF management 123See also worksheets.

planning ESS Copy Services 46planning worksheets

for Direct Management 151for DSCLI management 159for eRCMF management 177for SVC PPRC 169

policiesintersite management policy 21, 29

PPRCconfiguring for Direct Management 56definition 17, 188DSCLI replicated resources

changing configuration 94PPRC instance 17PPRC pairs 17, 46, 61, 69, 189

defining 62PPRC paths 46, 189

planning 47tasks 64

PPRC replicated resources 21, 61, 69, 189adding to resource group 68changing configuration 69, 94, 137

PPRC tasks 48, 49, 189configuring 56, 69running 61

PPRC tasks for Direct management

defining 64sample configuration 18

PPRC instance statesDSCLI management 143

PPRC pairsdefining to HACMP (Direct management) 65initial state for eRCMF management 136

PPRC pathsdefining tasks for Direct management 57

PPRC Relationshipsvolume states 118

PPRC Relationships Worksheet 172, 173PPRC replicated resources

definition 20planning sites for DSCLI 80

PPRC tasksnames 56understanding names 50

PPRC-Mirrored Volumes Worksheet 152, 153, 162, 163

PPRC-Mirrored Volumes Worksheet (Direct Management) worksheets

PPRC-Mirrored Volumes Worksheet (Direct Management) 48

prerequisitesconfiguration (Direct management) 55DSCLI management configuration 87

primary site 22primary site ESS information 63primary volume 47, 49

Rremoving

definition of SVC PPRC relationship 116DSCLI configuration 95SVC cluster definition 115SVC PPRC resources 117

replacingdirect management wioth eRCMF 130

replicated resourcesfor eRCMF management 135types 20

replicated resources. See PPRC, PPRC replicated resources.

Resource Group for HACMP/XD for Metro Mirror Worksheet 29


Index S – S

resource groupsconfiguring for eRCMF resources 136HACMP definition 21intersite management policy 29management policies 21planning 28planning for eRCMF management 128planning for SVC PPRC 108PPRC limitations 29PPRC replicated resources 68, 69restrictions for PPRC replicated resources 21site policy for eRCMF 128with PPRC replicated resources 21worksheet for DSCLI 84

Resource Groups Worksheet 148, 184rmpprc command 78rsExecuteTask.sh command 60rsList2105.sh command 48rsPrimeServer command 48running tasks 61

Ssample configuration

DCSLI management 74Direct Management 44eRCMF management 122SVC Management 100SVC PPRC 102

sample hardware configuration 18SAN Volume Controller

definition 97secondary site 22secondary site ESS information 64secondary volume 47, 49security

eRCMF management 141serial number for volume 48site splits behaviors

eRCMF 124Site Worksheet 28sites 63, 189

changing configuration 70, 95defining 132defining for eRCMF 132ESS 62for HACMP/XD for Metro Mirror 22importing volume groups to secondary 60intersite management policy 21, 29planning ESS site 52planning for eRCMF management 124planning for HACMP/XD for Metro Mirror 27primary site ESS information 63secondary site ESS information 64site backup communication method 28site dominance 28

software requirementsfor isntalling Direct management 53

source volume 47, 189ssh command

needed for SVC PPRC 110starting

HACMP clustereRCMF management 136

SVC PPRC cluster 114starting cluster 69state at cluster startup 69storage servers 97, 189SVC

filesets 32SVC CLI commands

listing host to vDisk maps 110SVC cluster

adding definition to HACMP 110changing definition to HACMP 115removing definition to HACMP 115

SVC Cluster Worksheet 170, 171SVC Management

planning 98sample configuration 100

SVC PPRCadding Consistency Groups to HACMP resource

groups 112changing definition of relationships 115configuring a cluster 109Consistency Group Name 112defining relationships 111filesets for cluster 109installing 109limitations and restrictions 99modifying Consistency Groups to HACMP resource

groups 117overview 19removing definition of relationships 116resources

changing definition in HACMP 117defining to HACMP 111removing definition in HACMP 117

sample cluster configuration 102setting up volume groups for cluster 103starting the cluster 114synchronizing configuration 114troubleshooting clusters 118verifying cluster configuration 112

SVC PPRC replicated resourcesplanning 107

SVC PPRC Resources Worksheet 174, 175SVC PPRC Volume states 118SVC-managed PPRC replicated resources

definition 21synchronizing

DSCLI-managed cluster 94SVC PPRC cluster 114


Index T – W

19

synchronizing Direct management configuration 68

Ttarget volume 47, 189Tasks Worksheet 49terminology

auxiliary (SVC) 98master (SVC) 98

troubleshootingDSCLI management 142SVC PPRC clusters 118

Uupgrading

to HACMP/XD for Metro Mirror v.5.4spprc 86

verification 39upgrading to HACMP 5.4 38

upgrade paths 38User-specific Tasks Worksheet 154, 155

Vverifying

cluster with SVC PPRC configuration 112Direct Management configuration 68DSCLI-managed PPRC configuration 92upgrade 39

viewingcurrent state of consistency groups 119

volume groupsconfiguring tasks for Direct management 56identifying by name 48importing at secondary site 60logical subsystem identifier 49secondary 61setting up for DSCLI management 77source volumes 47target volumes 47volume number 48

volume groups and filesetssetup for SVC PPRC cluster 103

volume number 48volumes

primary 47, 49, 188secondary 47, 49, 189target 189

Wworksheets 27

Copy Services Server Worksheet 160, 161DS ESS Disk Subsystem Worksheet 164, 165ESS Disk Subsystem Worksheet 156, 157, 164,

175ESS Disk Subsystem Worksheet (Direct

Management) 51HACMP resource groups 148PPRC Relationships Worksheet 172, 173PPRC-Mirrored Volumes Worksheet 152, 153,

162, 163Resource Group for HACMP/XD for Metro Mirror

Worksheet 29Resource Groups Worksheet 148, 184Site Worksheet 28summary 145, 151, 159, 177SVC Cluster Worksheet 170, 171SVC PPRC

summary 169SVC PPRC Resources Worksheet 174, 175User-specific Tasks Worksheet 49, 154, 155


hacmp aix

Documents

hacmpxd metro mirror

metro mirror general

overview of hacmpxd

metro mirror features

metro mirror management

hacmpxd software39upgrading

planning prerequisites

planning overview