VERITAS NetBackup™ 5.1 High Availability
System Administrator’s Guide
N12353C
Disclaimer The information contained in this publication is subject to change without notice. VERITAS Software Corporation makes no warranty of any kind with regard to this manual, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose. VERITAS Software Corporation shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this manual. VERITAS Legal Notice Copyright © 2002-2004 VERITAS Software Corporation. All rights reserved. VERITAS, VERITAS Software, the VERITAS logo, and all other VERITAS product names and slogans are trademarks or registered trademarks of VERITAS Software Corporation.VERITAS, VERITAS NetBackup, VERITAS Backup Exec, and the VERITAS Logo, Reg. U.S. Pat. & Tm. Off. Other product names and/or slogans mentioned herein may be trademarks or registered trademarks of their respective companies. Portions of this software are derived from the RSA Data Security, Inc. MD5 Message-Digest Algorithm. Copyright 1991-92, RSA Data Security, Inc. Created 1991. All rights reserved. VERITAS Software Corporation 350 Ellis Street Mountain View, CA 94043 USA Phone 650–527–8000 Fax 650–527–2901 www.veritas.com Third-Party Copyrights
ACE 5.2A: ACE(TM) is copyrighted by Douglas C.Schmidt and his research group at Washington University and University of California, Irvine, Copyright (c) 1993-2002, all rights reserved. IBM XML for C++ (XML4C) 3.5.1: Copyright (c) 1999,2000,2001 Compaq Computer Corporation; Copyright (c) 1999,2000,2001 Hewlett-Packard Company; Copyright (c) 1999,2000,2001 IBM Corporation; Copyright (c) 1999,2000,2001 Hummingbird Communications Ltd.; Copyright (c) 1999,2000,2001 Silicon Graphics, Inc.; Copyright (c) 1999,2000,2001 Sun Microsystems, Inc.; Copyright (c) 1999,2000,2001 The Open Group; All rights reserved. Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, provided that the above copyright notice(s) and this permission notice appear in all copies of the Software and that both the above copyright notice(s) and this permission notice appear in supporting documentation. This product includes software developed by the Apache Software Foundation (http://www.apache.org/). JacORB 1.4.1: The licensed software is covered by the GNU Library General Public License, Version 2, June 1991. Open SSL 0.9.6: This product includes software developed by the OpenSSL Project * for use in the OpenSSL Toolkit. (http://www.openssl.org/) TAO (ACE ORB) 1.2a: TAO(TM) is copyrighted by Douglas C. Schmidt and his research group at Washington University and University of California, Irvine, Copyright (c) 1993-2002, all rights reserved.
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix What Is In This Manual? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x Getting Help . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi NetBackup Manuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xii Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv Accessibility Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xv Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvi Chapter 1. Introduction to NetBackup High Availability . . . . . . . . . . . . . . . . . . . . . . . 1 How NetBackup Protects Data in a Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Supported Cluster Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Supported NetBackup Database Agents and Options in a Cluster . . . . . . . . . . . . . . 3 Accessing Compatibility Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Limitations of Using NetBackup with Clustering Solutions . . . . . . . . . . . . . . . . . . . . . . . 4 Media Manager Cluster Support Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Sample NetBackup Configuration in a Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Chapter 2. NetBackup in Microsoft Cluster Server . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Installation Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Installing and Configuring a NetBackup Failover Server on a MSCS Cluster . . . . . . . . 8 Installation Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Installation and Configuration Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
iii
Upgrading a NetBackup Failover Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Installing Non-Failover Media Servers on a MSCS Cluster . . . . . . . . . . . . . . . . . . . . . . 10 Installing the NetBackup Client on a MSCS Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Performing Backups and Restores in a MSCS Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Restoring the Cluster Quorum for Windows 2000 Systems . . . . . . . . . . . . . . . . . . . 12 Restoring Cluster Quorum to Windows 2000 Node With Directory Services . . . . 13 Preparing for Disaster Recovery of A Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Disaster Recovery of A MSCS Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Recovering the Entire Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Recovering All Shared Disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Recovering NetBackup Data on a Shared Disk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Chapter 3. NetBackup in HP Tru64 UNIX TruCluster . . . . . . . . . . . . . . . . . . . . . . . . 19 The CAA Facility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Installing a NetBackup Server on a TruCluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Installation Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Installation Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Configuring TruCluster to Use a Failover NetBackup Server . . . . . . . . . . . . . . . . . . . . 22 Tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Applying a Patch to a NetBackup Failover Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Upgrading a Failover Server to NetBackup 5.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Configuring TruCluster to Use a Non-Failover NetBackup Server . . . . . . . . . . . . . . . . 27 Apply a Patch to a NetBackup Non-Failover Server on a TruCluster . . . . . . . . . . . . . 29 Upgrading an Existing Non-Failover Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Installing Only the NetBackup Client on a TruCluster . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Performing Backups and Restores in a TruCluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
iv
NetBackup High Availability System Administrator’s Guide
Chapter 4. NetBackup in VERITAS Cluster Server on Windows 2000 and Windows 2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Installation Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Chapter 5. NetBackup in VERITAS Cluster Server on UNIX . . . . . . . . . . . . . . . . . . . 33 About the VCS Enterprise Agent for NetBackup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Detecting an Application Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Additional Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Installation Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Pre-Installation Checklist for a NetBackup Failover Server Installation . . . . . . . . . . . . 36 For VCS Solaris and Linux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Scenario 1 - Using VxVM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Scenario 2- Disk Only, No Volume Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 For VCS AIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Scenario 1 - Using VxVM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Scenario 2 - Using AIX LVMVG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Scenario 3 - Disk Only, No Volume Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 For VCS HP-UX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Scenario 1 - Using VxVM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Scenario 2 - Using Logical Volume Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Scenario 3 - Disk Only, No Volume Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Installing and Configuring a Failover NetBackup Server . . . . . . . . . . . . . . . . . . . . . . . . 44 Installation Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Installation and Configuration Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Installing NetBackup Agents and Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Upgrading a NetBackup Failover Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Upgrading a NetBackup Failover Server from 4.5FPx or 4.5 MPx to 5.1 . . . . . . . . . . . . 47 Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Upgrade Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Installing a Non-Failover NetBackup Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Contents v
Upgrading a NetBackup Non-Failover Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Installing the NetBackup Client Only . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Performing Backups and Restores in a VCS Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Chapter 6. NetBackup in Sun Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Installation Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Installation and Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Installation Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Installation and Configuration Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Installing NetBackup Agents and Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Upgrading a NetBackup Failover Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Upgrading from NetBackup 4.5FP3 or later (With Sun Microsystem’s NetBackup Agent) to 5.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Bringing the NetBackup Resource Group Online/Offline . . . . . . . . . . . . . . . . . . . . . . . 61 Performing Backups and Restores in a Sun Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Chapter 7. Configuring NetBackup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 NetBackup Configuration Guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Device Configuration Guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Configuring Devices in a Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Configuring Devices for MSCS, VCS, or Sun Cluster . . . . . . . . . . . . . . . . . . . . . 66 Configuring Devices for a TruCluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Verifying NetBackup Catalog Backup Entries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Windows Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 UNIX Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Verifying NetBackup Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Configuring Robotic Daemons for Monitoring (VCS for UNIX, TruCluster, or Sun Cluster) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
vi
NetBackup High Availability System Administrator’s Guide
Configuring Add-Ons for Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Configuring Add-ons for Monitoring in a VCS for UNIX, TruCluster, or Sun Cluster Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Configuring Add-ons for Monitoring in a MSCS Environment . . . . . . . . . . . . . . . . 71 Chapter 8. Using NetBackup to Perform Backups and Restores in a Cluster . . . . . 73 What to Back Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Backing Up Local Disks in a Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Backing Up the Shared Disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Backing Up Database Files in a Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Restoring Data in a Cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Chapter 9. Operational Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Delay in Detecting of Loss of Connection (MSCS and VCS on Windows) . . . . . . . . . . 80 MSCS Clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 Appendix A. Additional Configurations for NetBackup and Clusters . . . . . . . . . . . 83 Cluster with Tape Devices on a Shared SCSI Bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 Configuring a Shared SCSI Bus for Tape Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 Cluster with Locally-attached Storage Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 Clusters Sharing Devices on a Fibre Channel SAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Appendix B. bpclusterutil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .93
Contents
vii
viii
NetBackup High Availability System Administrator’s Guide
Preface
This guide provides information on how to install and configure NetBackup to work with different clustering solutions. NetBackup provides high-performance backups and restores for a variety of computer types. Administrators can set up schedules for automatic, unattended backups for clients anywhere in the network. These backups can be full or incremental and are managed entirely by the NetBackup server (also referred to as the NetBackup master server). This guide is intended for the NetBackup system administrator responsible for configuring NetBackup in a cluster environment. A system administrator is a person with system administrator privileges and responsibilities. This guide assumes:
◆ ◆ ◆ ◆
A basic understanding of Windows and UNIX system administration. A working understanding of the NetBackup for Windows and UNIX client software. A working understanding of the NetBackup for Windows and UNIX server software. A familiarity with the information covered in the following NetBackup manuals:
◆ ◆
NetBackup System Administrator’s Guide for UNIX, Volume I, or NetBackup System Administrator’s Guide for Windows, Volume I NetBackup Troubleshooting Guide for UNIX and Windows
◆
A working understanding of your cluster software.
ix
What Is In This Manual?
What Is In This Manual?
This guide is organized as follows:
◆ ◆ ◆ ◆
“Introduction to NetBackup High Availability” contains information on how NetBackup works in a cluster. “NetBackup in Microsoft Cluster Server” describes how to install and configure NetBackup in a Microsoft Cluster Server environment. “NetBackup in HP Tru64 UNIX TruCluster” describes how to install and configure NetBackup in a HP Tru64 UNIX TruCluster environment. “NetBackup in VERITAS Cluster Server on Windows 2000 and Windows 2003” describes installation prerequisites and post-installation instructions for configuring NetBackup in a VERITAS Cluster Server 4.1 for a Windows 2000 or Windows 2003 environment. “NetBackup in VERITAS Cluster Server on UNIX” describes how to install and configure NetBackup in a VERITAS Cluster Server in a UNIX environment. “NetBackup in Sun Cluster” describes how to install and configure NetBackup in a Sun Cluster environment. “Configuring NetBackup” includes guidelines for configuring the NetBackup Server, including device configuration. “Using NetBackup to Perform Backups and Restores in a Cluster” describes what data should be backed up in a cluster and guidelines for how to back up and restore data in a cluster. “Operational Notes” covers general information about using NetBackup in a cluster environment. “Additional Configurations for NetBackup and Clusters” provides further examples of how NetBackup can be configured in a cluster. “bpclusterutil” describes this command and how to use it to modify and configure NetBackup in a Windows cluster.
◆ ◆ ◆ ◆
◆ ◆ ◆
x
NetBackup High Availability System Administrator’s Guide
Getting Help
Getting Help
VERITAS offers you a variety of support options. Accessing the VERITAS Technical Support Web Site The VERITAS Support Web site allows you to:
◆ ◆ ◆ ◆ ◆ ◆ ◆ ◆
obtain updated information about NetBackup, including system requirements, supported platforms, and supported peripherals contact the VERITAS Technical Support staff and post questions to them get the latest patches, upgrades, and utilities view the NetBackup Frequently Asked Questions (FAQ) page search the knowledge base for answers to technical support questions receive automatic notice of product updates find out about NetBackup training read current white papers related to NetBackup
The address for the VERITAS Technical Support Web site follows:
◆
http://support.veritas.com
Subscribing to VERITAS Email Notification Service Subscribe to the VERITAS Email notification service to be informed of software alerts, newly published documentation, Beta programs, and other services. Go to http://support.veritas.com. Select a product and click “E-mail Notifications” on the right side of the page. Your customer profile ensures you receive the latest VERITAS technical information pertaining to your specific interests. Accessing VERITAS Telephone Support Telephone support for NetBackup is only available with a valid support contract. To contact VERITAS for technical support, dial the appropriate phone number listed on the Technical Support Guide included in the product box and have your product license information ready for quick navigation to the proper support group.
Preface
xi
NetBackup Manuals ▼
To locate the telephone support directory on the VERITAS web site 1. Open http://support.veritas.com in your web browser. 2. Click the Phone Support icon. A page that contains VERITAS support numbers from around the world appears. Accessing VERITAS E-mail Support
▼
To contact support using E-mail on the VERITAS web site 1. Open http://support.veritas.com in your web browser. 2. Click the E-mail Support icon. A brief electronic form will appear and prompt you to:
◆ ◆ ◆ ◆
Select a language of your preference Select a product and a platform Associate your message to an existing technical support case Provide additional contact and product information, and your message
3. Click Send Message. Contacting VERITAS Licensing For license information call 1-800-634-4747 option 3, fax 1-650-527-0952, or e-mail amercustomercare@veritas.com.
NetBackup Manuals
The following manuals, along with the online help, comprise the NetBackup documentation set. For a more detailed listing of NetBackup documents, refer to NetBackup Release Notes for UNIX and Windows. If you have a UNIX server, refer to these documents:
◆
VERITAS NetBackup System Administrator's Guide for UNIX, Volume I NetBackup_AdminGuideI_UNIXServer.pdf Explains how to configure and manage NetBackup on a UNIX server, including managing storage units, backup policies, catalogs and host properties.
xii
NetBackup High Availability System Administrator’s Guide
NetBackup Manuals
◆
VERITAS NetBackup System Administrator's Guide for UNIX, Volume II NetBackup_AdminGuideII_UNIXServer.pdf Explains additional NetBackup features such as notify scripts, enhanced authorization and authentication, and role-based security. The guide also discusses using NetBackup with AFS, Intelligent Disaster Recovery (IDR), and the BE Tape Reader.
◆
NetBackup Media Manager System Administrator’s Guide for UNIX MediaMgr_AdminGuide_Unix.pdf Explains how to configure and manage the storage devices and media on UNIX NetBackup servers. Media Manager is part of NetBackup.
◆
VERITAS NetBackup Commands for UNIX NetBackup_Commands_UNIX.pdf Describes NetBackup and Media Manager commands and processes that can be run from a UNIX command line.
If you have a Windows server, refer to these documents:
◆
VERITAS NetBackup System Administrator’s Guide for Windows, Volume I NetBackup_AdminGuideI_WinServer.pdf Explains how to configure and manage NetBackup on a Windows server, including managing storage units, backup policies, catalogs and host properties.
◆
VERITAS NetBackup System Administrator’s Guide for Windows, Volume II NetBackup_AdminGuideII_WinServer.pdf Explains additional NetBackup features such as notify scripts, enhanced authorization and authentication, and role-based security. The guide also discusses using NetBackup with AFS, Intelligent Disaster Recovery (IDR), and the BE Tape Reader.
◆
NetBackup Media Manager System Administrator’s Guide for Windows MediaMgr_AdminGuide_Windows.pdf Explains how to configure and manage the storage devices and media on Windows NetBackup servers. Media Manager is part of NetBackup.
◆
VERITAS NetBackup Commands for Windows NetBackup_Commands_Windows.pdf Describes NetBackup and Media Manager commands and processes that can be run from a Windows command prompt.
Preface
xiii
Glossary
Also refer to the following documents:
◆
VERITAS NetBackup Troubleshooting Guide for UNIX and Windows NetBackup_Troubleshoot_Guide.pdf Provides troubleshooting information for UNIX- and Windows-based NetBackup products, including Media Manager.
◆
VERITAS NetBackup Media Manager Device Configuration Guide for UNIX and Windows MediaMgr_DeviceConfig_Guide.pdf Explains how to add device drivers and perform other system-level configurations for storage devices and media servers (or SAN media servers) that are supported by NetBackup Media Manager.
Glossary
If you encounter unfamiliar terminology, consult the NetBackup online glossary. The glossary contains terms and definitions for NetBackup and all additional NetBackup options and agents. The NetBackup online glossary is included in the NetBackup help file.
▼
To access the NetBackup online glossary 1. In the NetBackup Administration Console, click Help > Help Topics. 2. Click the Contents tab. 3. Click Glossary of NetBackup Terms. Use the scroll function to navigate through the glossary.
xiv
NetBackup High Availability System Administrator’s Guide
Accessibility Features
Accessibility Features
NetBackup contains features that make the user interface easier to use by people who are visually impaired and by people who have limited dexterity. Accessibility features include:
◆ ◆
Support for assistive technologies such as screen readers and voice input (Windows servers only) Support for keyboard (mouseless) navigation using accelerator keys and mnemonic keys
For more information, see the NetBackup System Administrator’s Guide for Windows, Volume I or the NetBackup System Administrator’s Guide for UNIX, Volume I.
Preface
xv
Conventions
Conventions
The following conventions apply throughout the documentation set. Product-Specific Conventions The following term is used in the NetBackup 5.1 documentation to increase readability while maintaining technical accuracy.
◆
Microsoft Windows, Windows Terms used to describe a specific product or operating system developed by Microsoft, Inc. Some examples you may encounter in NetBackup documentation are, Windows servers, Windows 2000, Windows Server 2003, Windows clients, Windows platforms, or Windows GUI. When Windows or Windows servers is used in the documentation, it refers to all of the currently supported Windows operating systems. When a specific Windows product is identified in the documentation, only that particular product is valid in that instance. For a complete list of Windows operating systems and platforms that NetBackup supports, refer to the NetBackup Release Notes for UNIX and Windows or go to the VERITAS support web site at http://www.support.veritas.com.
Typographical Conventions Here are the typographical conventions used throughout the manuals:
Conventions Convention GUI Font Description Used to depict graphical user interface (GUI) objects, such as fields, listboxes, menu commands, and so on. For example: Enter your password in the Password field. Used for placeholder text, book titles, new terms, or emphasis. Replace placeholder text with your specific text. For example: Replace filename with the name of your file. Do not use file names that contain spaces. This font is also used to highlight NetBackup server-specific or operating system-specific differences. For example: This step is only applicable for NetBackup Enterprise Server. Code Used to show what commands you need to type, to identify pathnames where files are located, and to distinguish system or application text that is displayed to you or that is part of a code example.
Italics
xvi
NetBackup High Availability System Administrator’s Guide
Conventions Conventions (continued) Convention Key+Key Description Used to show that you must hold down the first key while pressing the second key. For example: Ctrl+S means hold down the Ctrl key while you press S.
You should use the appropriate conventions for your platform. For example, when specifying a path, use backslashes on Microsoft Windows and slashes on UNIX. Significant differences between the platforms are noted in the text. Tips, notes, and cautions are used to emphasize information. The following samples describe when each is used. Tip Used for nice-to-know information, like a shortcut.
Note Used for important information that you should know, but that shouldn’t cause any damage to your data or your system if you choose to ignore it. Caution Used for information that will prevent a problem. Ignore a caution at your own risk.
Command Usage The following conventions are frequently used in the synopsis of command usage. brackets [ ] The enclosed command line component is optional. Vertical bar or pipe (|) Separates optional arguments from which the user can choose. For example, when a command has the following format:
command arg1|arg2
In this example, the user can use either the arg1 or arg2 variable.
Preface
xvii
Conventions
Navigating Multiple Menu Levels When navigating multiple menu levels, a greater-than sign (>) is used to indicate a continued action. The following example shows how the > is used to condense a series of menu selections into one step:
❖
Select Start > Programs > VERITAS NetBackup > NetBackup Administration Console.
The corresponding actions could be described in more steps as follows: 1. Click Start in the task bar. 2. Move your cursor to Programs. 3. Move your cursor to the right and highlight VERITAS NetBackup. 4. Move your cursor to the right. First highlight and then click NetBackup Administration Console.
Terms Used in this Document
◆
Microsoft Cluster Server Microsoft Cluster Server is also referred to as MSCS in this document.
◆
VERITAS Volume Manager VERITAS Volume Manager is also referred to as Volume Manager in this document.
◆
VERITAS Cluster Server VERITAS Cluster Server is also referred to as VCS in this document.
◆
HP Tru64 UNIX TruCluster HP Tru64 UNIX TruCluster is also referred to as TruCluster.
xviii
NetBackup High Availability System Administrator’s Guide
Introduction to NetBackup High Availability
1
Clusters provide high availability of applications and data to users. In a cluster, two or more servers (called nodes) are linked in a network, and run cluster software that allows each node access to the shared disks. If a node becomes unavailable, cluster resources migrate to an available node (this is called failover). The shared disks and the virtual server are kept available. During failover, users experience only a short interruption in service. NetBackup can protect data on shared disks, as well as data on the node’s local disks. NetBackup can also protect databases in a cluster, such as Microsoft Exchange and Oracle, and make use of NetBackup options such as Global Data Manager (GDM).
How NetBackup Protects Data in a Cluster
There are several ways to use NetBackup for protecting the data in a cluster environment. One way is to back up the data from a cluster across the network to a separate NetBackup server by installing a NetBackup client. This is the easiest method to set up and maintain since the NetBackup-specific configuration tasks for tape devices, media, and so on, are kept separate from the set up and maintenance of the cluster itself. Another approach is to install a NetBackup media server. This allows you to back up the data to tape devices that attach directly to one or more of the nodes without going across the network. A NetBackup media server can be installed on the cluster either as a single media server running as a virtual server application, or it can be installed as separate media servers, each running on a node of the cluster independently of the other. Of course, installing separate media servers on each node of the cluster does not provide failover protection of the NetBackup media server itself, but there can be other advantages to this approach. Installing a single NetBackup media server on the cluster as a virtual server application allows the media server to failover from one of the nodes to the other as described previously. This is called a NetBackup failover media server. When you install a NetBackup failover media server, you assign NetBackup a network name resource (the virtual server name), an IP address resource, and a disk resource. The network name and IP address must be unique and set up in your name resolution configuration prior to setting up a NetBackup failover server.
1
Supported Cluster Solutions
The virtual server name is used in the NetBackup storage unit and added to the server list of any NetBackup clients to be backed up to this media server. When a failover occurs, backup jobs that were running are rescheduled by using the normal NetBackup retry logic for a failed backup. You then restart the NetBackup services on another node and the backup processing resumes. Yet another approach is to install a NetBackup master server on the cluster as a virtual server application. This is called a NetBackup failover master server and as the name implies, allows the duties of the master server to failover from one of the nodes to the other if a failure occurs on the active node. This provides high availability of the NetBackup master server itself. As with a NetBackup failover media server, when you install a NetBackup failover master server, you assign it a network name resource (the virtual server name), an IP address resource, and a disk resource. The network name and IP address must be unique and set up in your name resolution configuration prior to setting up a NetBackup failover server. The virtual server name is used as the name of the master server for all media servers and clients using this master server. As above, when a failover occurs, backup jobs that were running are rescheduled using the normal NetBackup retry logic for a failed backup. You the restart the NetBackup services for the failover node and the processing resumes. Both types of NetBackup failover servers, master and media, operate in an active/passive failover configuration. The active node and the passive (or failover node) must be the same type of server, master or media. NetBackup does not support mixing the two server types in the same failover server configuration. For example, a master server on one node cannot failover to a media server on another node. Specific details of how NetBackup runs in a cluster vary depending on the configuration you use in the cluster. (For information about the supported configurations and their specific requirements, see “Additional Configurations for NetBackup and Clusters” on page 83.)
Supported Cluster Solutions
NetBackup supports the following cluster environments. For more information on installing and configuring NetBackup in that type of cluster, refer to the chapter or document listed in the Table “Supported Cluster Solutions.”
Supported Cluster Solutions Cluster Solution Microsoft Cluster Server HP Tru64 UNIX TruCluster 2 Refer to... “NetBackup in Microsoft Cluster Server” on page 7 “NetBackup in HP Tru64 UNIX TruCluster” on page 19 NetBackup High Availability System Administrator’s Guide
Supported Cluster Solutions Supported Cluster Solutions Cluster Solution VERITAS Cluster Server 4.1 on Windows 2000 and Windows 2003 VERITAS Cluster Server on UNIX Refer to... “NetBackup in VERITAS Cluster Server on Windows 2000 and Windows 2003” on page 31 “NetBackup in VERITAS Cluster Server on UNIX” on page 33 “NetBackup in Sun Cluster” on page 53
Sun Cluster
Supported NetBackup Database Agents and Options in a Cluster
To see the list of database agents and NetBackup options that are supported in a clustered environment, refer to the following section “Accessing Compatibility Lists.” For information on installing and configuring agents and options in a cluster, refer to the system administrator’s guide for that agent or option.
Chapter 1, Introduction to NetBackup High Availability
3
Limitations of Using NetBackup with Clustering Solutions
Accessing Compatibility Lists
The VERITAS support web site now contains the most current platform support information for NetBackup. To locate the latest information on the VERITAS support web site, perform the following procedure.
▼
To access the compatibility lists on the VERITAS support website 1. Go to the VERITAS support web page: www.support.veritas.com. 2. In the Select Product Family list, click NetBackup Products. 3. From the Product list, choose NetBackup Enterprise Server. 4. With the default values in each of the fields, click the Compatibility tab. The page will fresh and display the various compatibility lists in the lower portion of the window.
Limitations of Using NetBackup with Clustering Solutions
The following limitations exist for NetBackup:
◆
NetBackup does not support converting an existing non-failover NetBackup server to a highly available failover NetBackup server. You must uninstall NetBackup, then install again. All NetBackup nodes within a cluster must be of the same server type. Do not mix master and media servers in a cluster. NetBackup is not supported as an active/active application in a cluster.
◆ ◆
Media Manager Cluster Support Limitations
The following limitations have been identified to affect clustering support in Media Manager. 1. Retries on All Failover Events When failover occurs, any task that is in process needs to be able to recover and continue. Communication needs to be re-established using the new node, and all operations that were in progress need to be retried on the new node.
4
NetBackup High Availability System Administrator’s Guide
Sample NetBackup Configuration in a Cluster
2. Breaking SCSI Reservations When fail over occurs in an environment where SCSI Reserve/Release is being used (explicitly in NetBackup, or implicitly in lower-level drivers or hardware), a SCSI Reservation remains on the drive. The drive remains unusable by other nodes and other hosts until the SCSI reservation is cleared. 3. Failover for Assigned Drives and In-progress Mount/dismount Activity State information (drive status, mount status, cleaning status, robotic library status, eject/inject status, etc.) is not in a database, as these states are in process memory, which is lost at failover time. On the new node, none of this information is reestablished by bringing up the daemons/services. The “SSO drive allocation resiliency” feature addresses this only for the device allocator (vmd/DA) component for SSO. 4. No Transaction Support in All Database Operations Current tasks include making multiple database operations to reflect one event. For example, this would include scenarios where changes are made by deleting and re-adding, where changes involve multiple records that are related (such as, optical partner relationships), or where robots and drives in the robot are deleted.
Sample NetBackup Configuration in a Cluster
Example of a cluster with NetBackup and tape devices that are shared. NetBackup Virtual Server Shared Disks
Controlling node
Failover node
Shared Tape
Private Network
LAN
Chapter 1, Introduction to NetBackup High Availability
5
Sample NetBackup Configuration in a Cluster
6
NetBackup High Availability System Administrator’s Guide
NetBackup in Microsoft Cluster Server
2
This chapter provides information on how to install and configure NetBackup in a Microsoft Cluster Server environment. Refer to the NetBackup System Administrator's Guide for Windows, Volume I, for a detailed understanding of how NetBackup is installed and administered. Refer to the documentation for MSCS for a detailed understanding of how MSCS works and how it is installed and administered.
7
Installation Prerequisites
Installation Prerequisites
❏ Verify all hardware (used for nodes) is supported by MSCS and NetBackup. For a list of supported storage devices, see the NetBackup Release Notes for UNIX and Windows and the MSCS documentation. Verify that storage devices have been properly installed and configured to work in a Windows environment. Refer to “Configuring Storage Devices for Windows” in the NetBackup Installation Guide for Windows. Verify that Microsoft Cluster Server is correctly installed and configured (all versions are supported with NetBackup 5.1). NetBackup can be installed on as many nodes as MSCS supports. MSCS supports up to 4 or 8 nodes in a cluster, depending on the level of Windows that is installed. If you want the NetBackup media server to be highly available, the robotic and tape devices must be physically connected to all the nodes that would be part of the NetBackup High Availability setup. This can be done via SCSI or fibre channel.Verify, using operating system commands, that all the devices are connected properly. Refer to the Media Manager Device Configuration Guide for UNIX and Windows. Using Cluster Administrator, verify that the Possible Owners list for each disk resource that NetBackup uses includes each node to which NetBackup can failover. Verify you have the NetBackup Enterprise Server 5.1 installation media and a valid license key.
❏
❏
❏
❏ ❏
Installing and Configuring a NetBackup Failover Server on a MSCS Cluster
The Microsoft Cluster Administrator must be installed on all NetBackup servers (including NetBackup administration clients) that you use to remotely administer the NetBackup failover server (master or media). You must also have administrator rights to the server cluster.
Installation Notes
◆ ◆
These instructions assume that this is a new installation of a NetBackup failover server. The same type of NetBackup server must be installed on each node to which NetBackup may failover. You cannot install a master server and a media server in a failover configuration.
8
NetBackup High Availability System Administrator’s Guide
Installing and Configuring a NetBackup Failover Server on a MSCS Cluster
◆ ◆
If you install a NetBackup failover media server, ensure that all media server nodes reference the same NetBackup master server name during the install. After configuring NetBackup as a cluster group in MSCS, do not change the group name.
Installation and Configuration Instructions
▼
To install a NetBackup failover server on a MSCS cluster 1. On each desired node in the cluster, install the NetBackup master or media server. Follow the instructions for installing NetBackup in a clustered environment as described in the NetBackup Installation Guide for Windows. 2. Add licenses for any NetBackup agents or options (note that some agents and options require a separate installation). See the appropriate NetBackup documentation for the agent or option for installation and configuration instructions. 3. After the installation is complete, continue with the configuration of NetBackup.
◆ ◆
Refer to “NetBackup Configuration Guidelines” on page 64 for guidelines on configuring NetBackup. Refer to “Device Configuration Guidelines” on page 64 and “Configuring Devices in a Cluster” on page 65 for guidelines and instructions on configuring devices in a cluster. Note that devices must be configured on each node in the cluster. See “Verifying NetBackup Catalog Backup Entries” on page 68 for instructions on verifying that the entries for the NetBackup Catalog backups are correct. By default, NetBackup options (such as NetBackup Advanced Reporter) will not cause NetBackup to failover if they fail. You can configure NetBackup options to failover NetBackup if a service or the option fails. For more information, see “Configuring Add-ons for Monitoring in a MSCS Environment” on page 71.
◆ ◆
Chapter 2, NetBackup in Microsoft Cluster Server
9
Upgrading a NetBackup Failover Server
4. When you have completed the configuration of NetBackup, verify that NetBackup can fail over properly in the cluster. Refer to “Verifying NetBackup Configuration” on page 69. Note Under normal circumstances, cluster configuration is one of the final steps in installing NetBackup in a cluster. If this step is not done or does not complete successfully, you can use the bpclusterutil command to perform this step. The “bpclusterutil -c” command runs the cluster configuration step that the installation performs. Using the -a option with the appropriate option name configures the NetBackup option in a cluster. You can run this command multiple times. Only run this command from the active node. See the Appendix “bpclusterutil” on page 89 for more information.
Upgrading a NetBackup Failover Server
▼
To upgrade a NetBackup failover server 1. Ensure a good backup of your cluster environment exists that includes the NetBackup databases and catalogs. 2. Follow the instructions for upgrading NetBackup in a clustered environment as described in the NetBackup Installation Guide for Windows. Note By default, NetBackup options (such as NetBackup Advanced Reporter) will not cause NetBackup to failover if they fail. You can configure NetBackup options to failover NetBackup if a service or the option fails. For more information, see “Configuring Add-ons for Monitoring in a MSCS Environment” on page 71. Note NetBackup does not support converting an existing non-failover NetBackup server to a highly available failover NetBackup server.
Installing Non-Failover Media Servers on a MSCS Cluster
Instead of installing a NetBackup failover media server on a MSCS cluster, you can install a non-failover NetBackup media server on each desired node of the cluster. There is no cluster failover protection of the NetBackup media server with this type of installation. A media server is installed on a cluster just as it is in a non-clustered environment.
10
NetBackup High Availability System Administrator’s Guide
Installing the NetBackup Client on a MSCS Cluster ▼
To install a NetBackup non-failover media server on a MSCS cluster 1. Review “Installation Prerequisites” on page 8. 2. Install the NetBackup media server software on each desired node in the cluster. Refer to the NetBackup Installation Guide for Windows for instructions. 3. Configure each NetBackup media server. Refer to the following manuals for information on configuring media servers:
◆ ◆ ◆
NetBackup Installation Guide for Windows NetBackup Media Manager System Administrator's Guide for Windows NetBackup System Administrator's Guide for Windows, Volume I
Installing the NetBackup Client on a MSCS Cluster
If you are backing up the data across the network to a separate NetBackup server, you can install the NetBackup client on the cluster. In this case, you must install the NetBackup client on each desired node of the cluster. You set up this configuration in nearly the same way as standalone clients. However, if you have problems with name resolution when trying to back up data on the cluster (either local data or shared data), consider setting the Required Network Interface parameter, for each client, to the fully qualified name of the node where the NetBackup client is installed. The NetBackup client is installed on a cluster just as it is in a non-clustered environment. Refer to the NetBackup Installation Guide for Windows for information on installing the NetBackup client.
Performing Backups and Restores in a MSCS Cluster
The process of configuring and performing backups and performing restores in a cluster is similar no matter which clustering solution you are using. Guidelines and instructions for backups and restores are covered in the chapter “Using NetBackup to Perform Backups and Restores in a Cluster” on page 73. The following section covers information on restoring the cluster quorum that is specific to a MSCS environment.
Chapter 2, NetBackup in Microsoft Cluster Server
11
Performing Backups and Restores in a MSCS Cluster
Restoring the Cluster Quorum for Windows 2000 Systems
The cluster quorum is backed up as part of System State for each node. To restore the cluster quorum, other nodes in the cluster must be offline. If you cannot take the other nodes in the cluster offline before restoring the cluster quorum, you can create the following registry entry to force the restore of the cluster quorum:
HKEY_LOCAL_MACHINE\Software\VERITAS\NetBackup\CurrentVersion\Config\ Cluster_Database_Force_Restore
Note The actual registry entry above must be on a single line. If required, create this registry value as data type REG_SZ with a value of YES. When this value is set, the cluster service is stopped for any nodes that are online. Restoring the quorum to a node that is running Active Directory or to a node that is currently a domain controller, requires additional steps not listed here. If Active Directory is running on the target server or the target server is currently a domain controller, see “Restoring Cluster Quorum to Windows 2000 Node With Directory Services” on page 13.
▼
To restore the cluster quorum 1. From the NetBackup Backup, Archive, and Restore interface, select the backup image (or range of images) that contains the backup of the System State for this node of the cluster. Then, select the node of the tree labeled System_State as the directory to be restored. For details, see the section on performing restores in the NetBackup Backup, Archive, and Restore Getting Started Guide. 2. If the disk where cluster quorum previously resided has been changed (The disk could have been replaced with a new one, or the disk configuration could have been changed so that the cluster quorum now resides on a different disk.), you may have to set the registry entry described above to force the restore to continue. This enables the drive letter of the disk that the cluster quorum was on to remain the same, even if the configuration has changed and the disk signatures contained in the restore media do not match the disk signatures contained in the cluster quorum. 3. Continue the restore operation as explained in the section on performing restores in the NetBackup Backup, Archive, and Restore Getting Started Guide. 4. When the restore operation is complete, either reboot the cluster node or use the Microsoft Cluster Administrator to restart the cluster service on any nodes where it was stopped, whichever is appropriate.
12
NetBackup High Availability System Administrator’s Guide
Performing Backups and Restores in a MSCS Cluster
Restoring Cluster Quorum to Windows 2000 Node With Directory Services
To restore the quorum to a node that is running Active Directory or to a node that is currently a domain controller, the node must be in Directory Services Restore Mode. The cluster services cannot run in this mode, so the cluster quorum must be restored separately, after System State is restored and the node has been rebooted.
▼
To restore the cluster quorum to a node running Active Directory or that is currently a domain controller 1. From the NetBackup Backup, Archive, and Restore interface, select the backup image (or range of images) that contains the backup of the System State for this node of the cluster. Then, select the node of the tree labeled System_State as the directory to be restored. For details, see the section on performing restores in the NetBackup Backup, Archive, and Restore Getting Started Guide. 2. Exclude the cluster quorum from being restored with System State at this time by expanding the System_State node and deselecting the Cluster_Database node of the tree. 3. Continue the restore operation by following the steps on performing restores in the NetBackup Backup, Archive, and Restore Getting Started Guide. 4. When the restore operation is complete, reboot the cluster node in safe (repair) mode (restart the computer and press F8 when prompted to select an operating system) and then select Directory Services Restore Mode. 5. Start the NetBackup service(s). 6. In the NetBackup Backup, Archive, and Restore interface, expand the System_State node and select only the Cluster_Database node. 7. If the disk where the cluster quorum previously resided has been changed, you may have to set the registry entry described above to force the restore to continue. The disk could have been replaced with a new one, or the disk configuration could have been changed so that the cluster quorum now resides on a different disk. This enables the drive letter of the disk that the cluster quorum was on to remain the same, even if the configuration has changed and the disk signatures contained in the restore media do not match the disk signatures contained in the cluster quorum. 8. Continue the restore operation by following the steps in the section on performing restores in the NetBackup Backup, Archive, and Restore Getting Started Guide.
Chapter 2, NetBackup in Microsoft Cluster Server
13
Performing Backups and Restores in a MSCS Cluster
9. When the restore has completed, reboot the target node. 10. When the restore operation is complete, use the Microsoft Cluster Administrator to restart the cluster service on any nodes on which it was stopped.
14
NetBackup High Availability System Administrator’s Guide
Preparing for Disaster Recovery of A Cluster
Preparing for Disaster Recovery of A Cluster
If a disaster occurs, the following information is required to successfully recover the cluster.
General Cluster Information Cluster name Cluster IP address and subnet mask Cluster node names Node IP addresses Local and shared drive letters and partition scheme Cluster Groups Group name Preferred nodes Failover/failback policies Cluster Resources Resource name Resource type Group membership Possible owners Resource dependencies Restart and Looks Alive/Is Alive properties Resource-related parameters _____________________________________________ _____________________________________________ _____________________________________________ _____________________________________________ _____________________________________________
_____________________________________________ _____________________________________________ _____________________________________________ _____________________________________________ _____________________________________________ _____________________________________________ _____________________________________________ _____________________________________________ _____________________________________________ _____________________________________________ _____________________________________________ _____________________________________________
Chapter 2, NetBackup in Microsoft Cluster Server
15
Disaster Recovery of A MSCS Cluster
Disaster Recovery of A MSCS Cluster
Several scenarios are discussed in this section:
◆ ◆ ◆
To recover the entire cluster, see the next topic, “Recovering the Entire Cluster”. To recover all shared disks, see “Recovering All Shared Disks” on page 17. To recover the shared disk that the NetBackup application resides on, see “Recovering NetBackup Data on a Shared Disk” on page 17.
Recovering the Entire Cluster
▼
To recover the nodes to their pre-disaster state 1. On the first node you want to recover, reinstall the Windows operating system, including the last service pack applied before the failure. 2. Install any other drivers or applications to make the node operational. Note After booting the nodes in a cluster, ensure that the drive letters match the original cluster configuration. If the original configuration does not match, you can use the Disk Administrator, to some extent, to control the hard drive numbering scheme that Windows devises. 3. On all remaining nodes you need to recover, reinstall Windows, including the last service pack applied before the failure. Also, see the previous note. 4. Reinstall the cluster services and bring the cluster online. 5. Reinstall the same NetBackup software on the cluster that was installed before the failure. For details, see the earlier sections that describe installing a NetBackup client or media server on a cluster, whichever applies. 6. If a NetBackup failover server was installed on the cluster or separate NetBackup media servers were installed on the nodes of the cluster as described earlier, use the NetBackup bprecover command to restore the NetBackup catalog information before continuing. See the NetBackup Troubleshooting Guide for UNIX and Windows for more information on recovering the NetBackup catalog database. 7. Restore the data files through the virtual server. 8. If you must restore a database to the shared drives, see the system administrator’s guide for your database agent to continue the restore.
16
NetBackup High Availability System Administrator’s Guide
Disaster Recovery of A MSCS Cluster
Recovering All Shared Disks
▼
To recover all shared disks 1. Uninstall the cluster software from both nodes. 2. Replace and repartition the shared disks. 3. Reinstall the cluster software. 4. Ensure all NetBackup shared disks are assigned the same drive letters used when NetBackup was originally installed and configured. 5. To reconfigure NetBackup for the cluster, run the following command.
bpclusterutil -c
6. If a NetBackup failover server was installed on the cluster as described earlier, use the NetBackup bprecover command to restore the NetBackup catalog information on the shared disk before continuing. See the NetBackup Troubleshooting Guide for UNIX and Windows for more information on recovering the NetBackup catalog database. 7. Use NetBackup to restore any data on the shared disks. For details on performing restores, see the NetBackup Backup, Archive, and Restore Getting Started Guide.
Recovering NetBackup Data on a Shared Disk
If a NetBackup failover server was installed on the cluster as described earlier and you must recover the shared disk that was used for the NetBackup failover server, use the NetBackup bprecover command to restore the NetBackup catalog information on the shared disk. See the NetBackup Troubleshooting Guide for UNIX and Windows for more information on recovering the NetBackup catalog database.
Chapter 2, NetBackup in Microsoft Cluster Server
17
Disaster Recovery of A MSCS Cluster
18
NetBackup High Availability System Administrator’s Guide
NetBackup in HP Tru64 UNIX TruCluster
3
This chapter provides information on how to install, configure and use NetBackup Server in a TruCluster environment. For information about NetBackup, refer to the NetBackup System Administrator's Guide for UNIX, Volume I. For information about TruCluster, refer to the TruCluster User's Guide.
19
The CAA Facility
The CAA Facility
Since all nodes in a TruCluster have access to all installed software within the cluster, there must to be a way to differentiate the single node that is to run the server software. In a typical NetBackup installation, a script starts and stops the server software. In a TruCluster, this script would be executed on all machines in the cluster. As a result, a number of processes are started on each machine. Where these machines are configured as servers this is problematic. HP’s CAA (Cluster Application Availability) facility can control the node on which NetBackup will run. The CAA facility provides monitoring and resource control and the ability to switch applications between nodes in a cluster. Once NetBackup is registered with CAA, this facility can be used to control upon which node an application can be active. When configuring a failover server, you will run the cluster_config script, which will register NetBackup with CAA and create the scripts used to start, stop, and monitor NetBackup. When configuring a non-failover server, these steps need to be performed manually. CAA provides two controls: caa_start
caa_stop brings the NetBackup daemons online takes the NetBackup daemons offline
Installing a NetBackup Server on a TruCluster
This section describes how to install NetBackup as a non-failover or failover server on a TruCluster.
Installation Notes
◆ ◆
These instructions assume that this is a new installation of a NetBackup server. Converting an existing non-failover NetBackup server to a highly available failover NetBackup server is not supported.
20
NetBackup High Availability System Administrator’s Guide
Installing a NetBackup Server on a TruCluster
Prerequisites
❏ Verify all hardware (used for nodes) is supported by TruCluster and NetBackup. For a list of supported storage devices, see the NetBackup Release Notes for UNIX and Windows and the TruCluster documentation. Verify that TruCluster 5.1a or 5.1b is correctly installed and configured. (Failover server only) Set up a virtual name for NetBackup in NIS, DNS or /etc/hosts. This hostname is used as a logical hostname during the NetBackup installation and configuration of the cluster. It is not assigned to any physical node, but is attached to the NetBackup application resource, enabling NetBackup to failover to the appropriate node during the failover process. (Failover server only) Make sure that each node in the cluster, on which you wish to install NetBackup, is rsh equivalent. As the root user you need to be able to perform a remote login to each node without entering a password. This is only necessary for installation and configuration of the NetBackup server and any NetBackup agents and options. Once installation and configuration is complete this is no longer required. If you want the NetBackup media server to be highly available, the robotic and tape devices must be physically connected to all the nodes that would be part of the NetBackup High Availability setup. This can be done via SCSI or fibre channel.Verify, using operating system commands, that all the devices are connected properly. Refer to the Media Manager Device Configuration Guide for UNIX and Windows. Verify you have the NetBackup Enterprise Server 5.1 installation media and a valid license key.
❏ ❏
❏
❏
❏
Installation Instructions
▼
To install NetBackup as a NetBackup failover server in a TruCluster 1. Follow the instructions for installing NetBackup Server software as described in the NetBackup Installation Guide for UNIX. The NetBackup server software (master or media) can be installed on any node in the cluster. During installation, specify the cluster alias as the server name. 2. Continue with the section “Configuring TruCluster to Use a Failover NetBackup Server” on page 22.
Chapter 3, NetBackup in HP Tru64 UNIX TruCluster
21
Configuring TruCluster to Use a Failover NetBackup Server ▼
To install NetBackup as a NetBackup non-failover server in a TruCluster Note When the NetBackup is installed in a cluster as a failover server, the NetBackup client will only run on the node that is also running the NetBackup server. 1. Follow the instructions for installing NetBackup Server software as described in the NetBackup Installation Guide for UNIX. The NetBackup server software (master or media) can be installed on any node in the cluster. During installation, specify the specific node name as the server name. 2. Continue with the section “Configuring TruCluster to Use a Non-Failover NetBackup Server” on page 27.
Configuring TruCluster to Use a Failover NetBackup Server
▼
To configure for a failover server 1. From the /usr/openv/netbackup/bin/cluster/ directory, run the cluster_config configuration utility. 2. For the TruCluster NetBackup service/profile name, indicate a new name for this configuration. 3. Specify the cluster alias to be used by NetBackup. If this is a new cluster alias, indicate the cluster alias IP address to be used by NetBackup. This should already be configured in the /etc/hosts file, or in DNS, or NIS. Also provide the cluster alias IP subnet to be used by CAA. This is the network subnet for the cluster alias IP address. For instance, an IP address of 10.80.136.119 on a network with a netmask of 255.255.248.0 would have a subnet of 10.80.136.0. 4. Provide a list of the cluster members on which NetBackup may run. Separate member names with a space and list them in the preferred order. 5. On the summary screen, review the information you provided. Note If you chose to use a pre-existing cluster alias, the IP and subnet fields will be blank.
◆ ◆
If the information is incorrect, choose No, and the configuration script will start back at the beginning. If the information is correct, choose Yes.
22
NetBackup High Availability System Administrator’s Guide
Configuring TruCluster to Use a Failover NetBackup Server
The script will:
◆ ◆ ◆ ◆ ◆
Create and register a CAA profile for the NetBackup application. Create and register a cluster alias (if a new alias was specified). Remove the S77NetBackup and K77NetBackup scripts from /sbin/rc0.d and /sbin/rc2.d directories. Update the bp.conf and vm.conf files. Start the NetBackup server.
6. You are now able to start and stop NetBackup with the following commands:
caa_start profile_name caa_stop profile_name
For example, with a caa registered profile named “netbackup” for NetBackup server, you can start and stop the application with the following commands:
caa_start netbackup caa_stop netbackup
You must use the caa commands in place of the standard NetBackup start and stop script. The standard script does not track which node the software is running on and allows multiple instances of NetBackup within the cluster. 7. Continue with the configuration of NetBackup.
◆ ◆
Refer to “NetBackup Configuration Guidelines” on page 64 for guidelines on configuring NetBackup. Refer to “Device Configuration Guidelines” on page 64 and “Configuring Devices in a Cluster” on page 65 for guidelines and instructions on configuring devices in a cluster. Note that devices must be configured on each node in the cluster. See “Verifying NetBackup Catalog Backup Entries” on page 68 for instructions on verifying that the entries for the NetBackup Catalog backups are correct.
◆
8. Install any NetBackup agents and options. a. Freeze NetBackup by running the following command:
caa_stop profile_name
b. Install the agents and options as you would in a non-clustered environment. Refer to the system administrator’s guides for those agents or options.
Chapter 3, NetBackup in HP Tru64 UNIX TruCluster
23
Configuring TruCluster to Use a Failover NetBackup Server
c. Unfreeze NetBackup by running the following command:
caa_start profile_name
By default, robotic daemons and NetBackup options (such as NetBackup Advanced Reporter) will not cause NetBackup to failover if they fail. You can configure robot daemons and NetBackup options to failover NetBackup if a daemon or the option fails. For more information, see “Configuring Robotic Daemons for Monitoring (VCS for UNIX, TruCluster, or Sun Cluster)” on page 70 and “Configuring Add-ons for Monitoring in a VCS for UNIX, TruCluster, or Sun Cluster Environment” on page 71. 9. When you have completed configuration of NetBackup, verify that NetBackup can failover properly in the cluster. Refer to “Verifying NetBackup Configuration” on page 69.
Tips
◆
By default, NetBackup attempts to complete a backup job two times (within 12 hours) before allowing the job to fail. It is possible that scheduler attempts may be exhausted before the failover of the NetBackup media server complete. If this happens, increase the Scheduled Backup Attempts setting to 5 or 6. You can also avoid this problem by setting the Wakeup Interval setting to a lower value. The Wakeup Interval setting should be set lower to increase the likelihood that NetBackup will be able to initiate all scheduled backups. If you experience caa_stop failures for NetBackup, you may want to consider increasing the script timeout value. The default is 180 seconds but sometimes this is not enough for a heavily loaded machine to shut down all of the robotic daemons. The same is true of machines with several tape drives attached in a SAN configuration. To see if this solution is correct for you, use “vmps” to check which daemons are running immediately after the error. If it is a robotic daemon, increasing the timeout will most likely fix your problem. The script timeout value is part of TruCluster profile property; use the caa_profile TruCluster command to change this value.
◆ ◆
24
NetBackup High Availability System Administrator’s Guide
Applying a Patch to a NetBackup Failover Server
Applying a Patch to a NetBackup Failover Server
Follow these instructions when applying patches to a NetBackup failover server. See the NetBackup Release Notes for UNIX and Windows for the latest information on applying a specific patch.
▼
To apply a patch to a NetBackup failover server 1. Ensure a good backup of your cluster environment exists that includes the NetBackup databases and catalogs. 2. Login to the node NetBackup is running on. 3. Stop the NetBackup server using the command caa_stop profile name. 4. Complete the patch install. 5. Start NetBackup using the command caa_start profile name. 6. Verify that NetBackup can failover properly in the cluster. Refer to “Verifying NetBackup Configuration” on page 69.
Upgrading a Failover Server to NetBackup 5.1
This section describes how to upgrade failover server.
▼
To upgrade a failover server 1. Ensure a good backup of your cluster environment exists that includes the NetBackup databases and catalogs. 2. Stop the NetBackup server using the following command.
caa_stop profile_name
3. In the vm.conf:
◆ ◆ ◆
Remove any MM_SERVER_NAME or MM_HOST_NAME entries from this file. Verify that any CLUSTER_NAME entries are defined as the virtual server name. If it exists, change the REQUIRED_INTERFACE entry to CLUSTER_NAME and define it as the virtual name.
Chapter 3, NetBackup in HP Tru64 UNIX TruCluster
25
Upgrading a Failover Server to NetBackup 5.1
4. Install NetBackup. Follow the instructions for “Upgrading NetBackup on UNIX Servers and Clients” as described in the NetBackup Installation Guide for UNIX. 5. When installation is complete run the following command.
caa_start profile_name
6. Upgrade any NetBackup agents and options you have purchased. a. Freeze NetBackup by running the following command:
caa_stop profile_name
b. Upgrade agents and options as you would in a non-clustered environment. Refer to the system administrator’s guides for those agents or options. By default, robotic daemons and NetBackup options (such as NetBackup Advanced Reporter) will not cause NetBackup to failover if they fail. Robot daemons and NetBackup options can be configured to failover NetBackup if a daemon or the option fails. For more information, see “Configuring Robotic Daemons for Monitoring (VCS for UNIX, TruCluster, or Sun Cluster)” on page 70 and “Configuring Add-ons for Monitoring in a VCS for UNIX, TruCluster, or Sun Cluster Environment” on page 71. c. Unfreeze NetBackup by running the following command:
caa_start profile_name
7. Verify that NetBackup can failover properly in the cluster. Refer to “Verifying NetBackup Configuration” on page 69.
26
NetBackup High Availability System Administrator’s Guide
Configuring TruCluster to Use a Non-Failover NetBackup Server
Configuring TruCluster to Use a Non-Failover NetBackup Server
The following procedure describes how to use the CAA facility to start NetBackup and keep track of the node on which it is active. For more information on this, see “The CAA Facility” on page 20.
▼
To configure for a non-failover server 1. Move the following: /usr/openv/netbackup/bin/cluster/trucluster/caa_netbackup to /usr/openv/netbackup/bin/cluster/trucluster/caa_netbackup.orig 2. Copy the following: /var/cluster/caa/template/template.scr to /usr/openv/netbackup/bin/cluster/trucluster/caa_netbackup 3. Remove the S77NetBackup and K77NetBackup scripts from rc0.d and rc2.d directories, because the CAA facility is used to start and stop NetBackup. 4. In the script, add pointers to the start and stop scripts and to the daemon(s) to be monitored (to determine if the application is active). The daemon to be monitored for the media server is vmd. For the master server the daemons to monitor are vmd, bprd and bpdbm. “netbackup” is the name used for the NetBackup application. To start NetBackup services the script should call:
/usr/openv/netbackup/bin/goodies/netbackup start
To stop the NetBackup services the script should call:
/usr/openv/netbackup/bin/goodies/bp.kill_all FORCEKILL
5. Once the script is complete, a CAA profile must be made. Run the following commands:
caa_profile -create netbackup -t application -a /usr/openv/netbackup/bin/cluster/trucluster/caa_netbackup -p restricted -h
This creates a profile that uses the completed script, names the application “netbackup,” and restricts the operation of that application to the named node.
Chapter 3, NetBackup in HP Tru64 UNIX TruCluster 27
Configuring TruCluster to Use a Non-Failover NetBackup Server
6. Assuming NetBackup has been installed, the application needs to be registered. Use the following command.
caa_register profile_name
7. In the /etc/inetd.conf file, comment out the lines that define the NetBackup listeners (bpcd, vnetd, vopied, bpjava). 8. Login to the node NetBackup will be running on (the restricted hostname specified in step 5) and add the following lines for the NetBackup listeners to the /etc/inetd.conf.local file.
bpcd stream tcp nowait root /usr/openv/netbackup/bin/bpcd bpcd vnetd stream tcp nowait root /usr/openv/bin/vnetd vnetd vopied stream tcp nowait root /usr/openv/bin/vopied vopied bpjava-msvc stream tcp nowait root /usr/openv/netbackup/bin/bpjava-msvc bpjava-msvc -transient
9. You can now start and stop NetBackup with the following commands:
caa_start profile_name caa_stop profile_name
For example, with a caa registered profile named “netbackup” for NetBackup server, the application you can start and stop NetBackup with the following commands:
caa_start netbackup caa_stop netbackup
You should use the caa commands in place of the standard NetBackup start and stop script. The standard script does not track which node the software is running on and allows multiple instances of NetBackup within the cluster. 10. Continue with the configuration of NetBackup. Refer to the NetBackup Installation Guide for UNIX for information on configuring the NetBackup server software. 11. Install any NetBackup agents and options you have purchased. Install agents and options as you would in a non-clustered environment. Refer to the NetBackup guides for those agents or options.
28
NetBackup High Availability System Administrator’s Guide
Apply a Patch to a NetBackup Non-Failover Server on a TruCluster
Apply a Patch to a NetBackup Non-Failover Server on a TruCluster
Note Converting an existing non-failover NetBackup server to a highly available failover NetBackup server is not supported. Follow these instructions when applying patches to a non-failover server.
▼
To apply a patch to a NetBackup non-failover server 1. Ensure a good backup of your cluster environment exists that includes the NetBackup databases and catalogs. 2. Login to the node NetBackup is running on (restricted hostname specified when creating the profile). 3. Stop NetBackup server using the command caa_stop profile name. 4. Install NetBackup 5.1. 5. In the /etc/inetd.conf file, make sure the NetBackup listeners (bpcd, vnetd, vopied, bpjava) are commented out. 6. To turn off the NetBackup listeners, refresh the inetd daemon on all cluster nodes. 7. For each node on which NetBackup is installed, verify that the NetBackup listeners (bpcd, vnetd, vopied, bpjava) are included in the /etc/inetd.conf.local file. 8. Start NetBackup using the command caa_start profile name.
Chapter 3, NetBackup in HP Tru64 UNIX TruCluster
29
Upgrading an Existing Non-Failover Server
Upgrading an Existing Non-Failover Server
Note Converting an existing non-failover NetBackup server to a highly available failover NetBackup server is not supported.
▼
To upgrade a non-failover server 1. Ensure a good backup of your cluster environment exists that includes the NetBackup databases and catalogs. 2. Refer the instructions for upgrading NetBackup on UNIX servers and clients as described in the NetBackup Installation Guide for UNIX.
Installing Only the NetBackup Client on a TruCluster
The client software can be installed on any node in the cluster (because the software is run from the shared /usr file system). Specify the name of one of the nodes in the cluster as the client name. This will limit the execution of the client to that node.
Performing Backups and Restores in a TruCluster
The process of configuring and performing backups and performing restores in a cluster is similar no matter which clustering solution you are using. Guidelines and instructions for backups and restores are covered in the chapter “Using NetBackup to Perform Backups and Restores in a Cluster” on page 73.
30
NetBackup High Availability System Administrator’s Guide
NetBackup in VERITAS Cluster Server on Windows 2000 and Windows 2003
4
This chapter provides information on the prerequisites for using NetBackup as a clustered application in a VERITAS Cluster Server (VCS), version 4.1, for Windows 2000 or Windows 2003 environment. For complete details on installing and configuring NetBackup in this environment, please refer to the VERITAS Cluster Server Administrator's Guide. VCS is a high-availability solution for cluster configurations. With VCS you can monitor systems and application services, and restart services on a different system when hardware or software fails. For information about VCS, refer to the VERITAS Cluster Server User's Guide.
31
Installation Prerequisites
Installation Prerequisites
This section contains information about the requirements that must be met prior to installing and configuring the VCS Enterprise Agent for NetBackup and the NetBackup server in a VCS environment. ❏ Verify that VCS 4.1 and NetBackup 5.1 support your hardware. For a list of supported storage devices, see the VERITAS Cluster Server System Administrator’s Guide and the NetBackup Release Notes for UNIX and Windows. Verify that storage devices have been properly installed and configured to work in a Windows environment. Refer to “Configuring Storage Devices for Windows” in the NetBackup Installation Guide for Windows. Verify that you have correctly installed and configured VCS 4.1. NetBackup can be installed on as many nodes as VCS 4.1 supports. Follow the steps in the VERITAS Cluster Server Installation Guide. The VCS NetBackup agent is installed with VCS 4.1; no separate installation is required.
❏
❏
Note All VCS servers must be at the same Windows Service Pack level, and at the same VCS version and VCS Service Pack level. ❏ ❏ ❏ ❏ Verify that VCS Cluster Manager is installed. Refer to the VERITAS Cluster Server Installation Guide for more information. Verify the network and nodes support the TCP/IP protocol. Verify that you have an IP address and host name to be assigned to the NetBackup resource in VCS. If you want the NetBackup media server to be highly available, the robotic and tape devices must be physically connected to all the nodes that would be part of the NetBackup High Availability setup. This can be done using SCSI or fibre channel.Verify that all the devices are connected properly. Refer to the Media Manager Device Configuration Guide for UNIX and Windows. Verify you have the NetBackup Enterprise Server 5.1 installation media and a valid license key.
❏
Configuration
Refer to the VERITAS Cluster Server Administrator's Guide for details on configuring NetBackup to work with the VCS NetBackup agent.
32
NetBackup High Availability System Administrator’s Guide
NetBackup in VERITAS Cluster Server on UNIX
5
This chapter describes the steps to install and configure a NetBackup server in a VCS for UNIX environment. VCS is a high-availability solution for cluster configurations. VCS enables you to monitor systems and application services, and to restart services on a different system when hardware or software fails. For information about VCS, refer to the VERITAS Cluster Server User's Guide.
33
About the VCS Enterprise Agent for NetBackup
About the VCS Enterprise Agent for NetBackup
The VCS Enterprise Agent for NetBackup monitors specific NetBackup resources, determines the status of these resources, and starts or stops NetBackup according to external events. The agent includes VCS type declarations and agent executables, and is represented by the NetBackup resource type. When the agent detects an application failure, the NetBackup service group is failed over to another system in the cluster. Specific agent operations include: Online Executes the following command: /usr/openv/netbackup/bin/goodies/netbackup start Offline Executes the following command: /usr/openv/netbackup/bin/goodies/bp.kill_all FORCEKILL Monitor Scans the process table for vmd, bprd, and bpdbm, etc. The specific processes to be monitored can be configured using the cluster configuration utility. The agent provides two levels of application monitoring: basic and detailed (in-depth). See “Additional Monitoring” on page 34 for more information. Stops all NetBackup daemons after detecting an unexpected offline operation or an ineffective online operation.
Clean
Detecting an Application Failure
The agent detects an application failure if the monitor process cannot locate the specific monitored processes (such as vmd, bprd, bpdbm.) in the process table.
Additional Monitoring
Additional monitoring capabilities are available for robots and add-on products. Refer to the “Configuring Robotic Daemons for Monitoring (VCS for UNIX, TruCluster, or Sun Cluster)” on page 70 and “Configuring Add-Ons for Monitoring” on page 70.
34
NetBackup High Availability System Administrator’s Guide
Installation Prerequisites
Installation Prerequisites
❏ Verify that both VCS and NetBackup support your hardware. For a list of supported storage devices, see the NetBackup Release Notes for UNIX and Windows and the VCS documentation. Verify that VCS 2.1 or later is correctly installed and configured on Solaris, HP-UX, AIX, or Linux. Follow the steps in the VERITAS Cluster Server Installation Guide. The shared disk must be configured and accessible to all cluster nodes on which NetBackup will be installed. Configure a virtual name for NetBackup in NIS, DNS or /etc/hosts. This hostname is used as a logical hostname during the NetBackup installation and configuration of the cluster. It is not assigned to any physical node, but is attached to the NetBackup application resource, allowing NetBackup to failover to the appropriate node during the failover process. Make sure that each node in the cluster, on which you wish to install NetBackup, is rsh equivalent (On HP the command is remsh). As the root user you need to be able to perform a remote login to each node in the cluster without entering a password. This is only necessary for installation and configuration of the NetBackup server and any NetBackup agents and options. Once installation and configuration is complete this is no longer required. If you want the NetBackup media server to be highly available, the robotic and tape devices must be physically connected to all the nodes that would be part of the NetBackup High Availability setup. This can be done using SCSI or fibre channel.Verify that all the devices are connected properly. Refer to the Media Manager Device Configuration Guide for UNIX and Windows. Verify you have the NetBackup Enterprise Server 5.1 installation media and a valid license key.
❏ ❏ ❏
❏
❏
❏
Chapter 5, NetBackup in VERITAS Cluster Server on UNIX
35
Pre-Installation Checklist for a NetBackup Failover Server Installation
Pre-Installation Checklist for a NetBackup Failover Server Installation
Note The configuration utility will treat all attribute values globally.
Virtual Name for NetBackup: IP Address: Subnet mask Network Device Name (NIC) Nodes on which NetBackup server is installed Node Name __________________________________ __________________________________ __________________________________ __________________________________
__________________________________ __________________________________ __________________________________ __________________________________
IP addresses __________________________________ __________________________________ __________________________________ __________________________________
36
NetBackup High Availability System Administrator’s Guide
Pre-Installation Checklist for a NetBackup Failover Server Installation
For VCS Solaris and Linux
Scenario 1 - Using VxVM
Resource Example
Disk Group Resource: DiskGroup: __________________________ StartVolumes: ________________________ StopVolumes: ________________________ Volume Resource: (optional) Volume: _____________________________ Mount Resource MountPoint: _________________________ BlockDevice: _________________________ FS Type: _____________________________ Mount Option: _______________________ Fsck Option: _________________________ /opt/VRTSnbu /dev/vx/dsk/nbuvol vxfs the required value (this is optional) (if you add other options, -y is also required) nbuvol nbudg 0 or 1 0 or 1
Chapter 5, NetBackup in VERITAS Cluster Server on UNIX
37
Pre-Installation Checklist for a NetBackup Failover Server Installation
Scenario 2- Disk Only, No Volume Manager
Resource Example
Mount Resource MountPoint: _____________________________ BlockDevice:_____________________________ FS Type:_________________________________ Mount Option: ___________________________ Fsck Option: _____________________________ /opt/VRTSnbu /dev/dsk/c1t1dos3 vxfs (this is optional) (if you add other options, -y is also required)
38
NetBackup High Availability System Administrator’s Guide
Pre-Installation Checklist for a NetBackup Failover Server Installation
For VCS AIX
Scenario 1 - Using VxVM
Resource Example
Disk Group Resource: DiskGroup: __________________________ StartVolumes: ________________________ StopVolumes: ________________________ Volume Resource: Volume: _____________________________ Mount Resource MountPoint: _________________________ BlockDevice: _________________________ FS Type: _____________________________ Mount Option: _______________________ Fsck Option: _________________________ /opt/VRTSnbu /dev/vx/dsk/nbuvol vxfs is required (optional) (if you add other options, -y is also required) nbuvol nbudg 0 or 1 0 or 1
Chapter 5, NetBackup in VERITAS Cluster Server on UNIX
39
Pre-Installation Checklist for a NetBackup Failover Server Installation
Scenario 2 - Using AIX LVMVG
Resource Example
LVMVG Resource: VolumeGroup:___________________________ MajorNumber: __________________________ Disks: __________________________________ ImportvgOpt: ___________________________ VaryonvgOpt: ___________________________ SyncODM: ______________________________ Mount Resource: MountPoint:_____________________________ BlockDevice: ____________________________ FS Type: ________________________________ Mount Option: __________________________ Fsck Option:_____________________________ /opt/VRTSnbu /dev/nbudev jfs2 (optional) (if you add other options, -y is also required) nbuvg 58 hdisk1 or hdisk2 (optional) (optional) (optional)
40
NetBackup High Availability System Administrator’s Guide
Pre-Installation Checklist for a NetBackup Failover Server Installation
Scenario 3 - Disk Only, No Volume Manager
Resource Example
Mount Resource: MountPoint: _____________________________ BlockDevice:_____________________________ FS Type:_________________________________ Mount Option: ___________________________ Fsck Option: _____________________________ /opt/VRTSnbu /dev/dsk/c1t1dos3 jfs2 (optional) (if you add other options, -y is also required)
Chapter 5, NetBackup in VERITAS Cluster Server on UNIX
41
Pre-Installation Checklist for a NetBackup Failover Server Installation
For VCS HP-UX
Scenario 1 - Using VxVM
Resource Example
Disk Group Resource: DiskGroup: __________________________ StartVolumes: ________________________ StopVolumes: ________________________ Volume Resource: Volume: _____________________________ Mount Resource: MountPoint:__________________________ BlockDevice: _________________________ FS Type: _____________________________ Mount Option: _______________________ Fsck Option:__________________________ /opt/VRTSnbu /dev/vx/dsk/nbuvol vxfs is required (optional) (if you add other options, -y is also required) nbuvol nbudg 0 or 1 0 or 1
Scenario 2 - Using Logical Volume Manager
LVMVolume Group Resource: Volume Group: _______________________ LVMLogical Volume Resource: Volume Group: _______________________ Logical Group:________________________ vg0a lvol1 vg0a
42
NetBackup High Availability System Administrator’s Guide
Pre-Installation Checklist for a NetBackup Failover Server Installation
Scenario 3 - Disk Only, No Volume Manager
Mount Resource: MountPoint: _____________________________ BlockDevice: _____________________________ FS Type: _________________________________ Mount Option: ___________________________ Fsck Option: _____________________________ /opt/VRTSnbu /dev/dsk/c1t1dos3 hfs (optional) (if you add other options, -y is also required)
Chapter 5, NetBackup in VERITAS Cluster Server on UNIX
43
Installing and Configuring a Failover NetBackup Server
Installing and Configuring a Failover NetBackup Server
This section describes how to install and configure a new NetBackup failover server. If you are upgrading from NetBackup 5.0 to 5.1, refer to “Upgrading a NetBackup Failover Server” on page 46.
Installation Notes
◆ ◆
These instructions assume that this is a new installation of a NetBackup failover server. You must install the same type of NetBackup server on each node to which NetBackup may failover. You cannot install a master server and a media server in a failover configuration. If you are installing a NetBackup failover media server, ensure that all media server nodes reference the same NetBackup master server name during the install. Do not install older versions of the VCS for UNIX agent. Use the version of the agent provided with NetBackup 5.1. This agent is installed along with the NetBackup 5.1 Server software.
◆ ◆
Installation and Configuration Instructions
▼
To install and configure a NetBackup failover server 1. Fill out the appropriate sections, based on your environment, of the checklist in the section “Pre-Installation Checklist for a NetBackup Failover Server Installation” on page 36. 2. Make sure the shared disk is not mounted on any node in the cluster. 3. Follow the instructions for installation and initial configuration in the NetBackup Installation Guide for UNIX. You must install on each node that you want NetBackup to run on. The virtual name should be used for the server name during the install procedure. 4. From the primary node, run the cluster_config utility located in /usr/openv/netbackup/bin/cluster directory. 5. Answer questions based on the information gathered in step 1.
44
NetBackup High Availability System Administrator’s Guide
Installing and Configuring a Failover NetBackup Server
6. Continue with the configuration of NetBackup.
◆ ◆
Refer to “NetBackup Configuration Guidelines” on page 64 for guidelines on configuring NetBackup. Refer to “Device Configuration Guidelines” on page 64 and “Configuring Devices in a Cluster” on page 65 for guidelines and instructions on configuring devices in a cluster. Note that devices must be configured on each node in the cluster. See “Verifying NetBackup Catalog Backup Entries” on page 68 for instructions on verifying that the entries for the NetBackup Catalog backups are correct.
◆
7. Install any NetBackup agents or options you have purchased. Refer to the next section, “Installing NetBackup Agents and Options.” 8. When you have completed configuration of NetBackup, verify that NetBackup can failover properly in the cluster. Refer to “Verifying NetBackup Configuration” on page 69.
Installing NetBackup Agents and Options
You should install NetBackup agents and options on all cluster nodes where the NetBackup server is installed. In addition, the NetBackup group must be frozen before you install any agents or options. By default, robotic daemons and NetBackup options (such as NetBackup Advanced Reporter) will not cause NetBackup to failover if they fail. Robot daemons and NetBackup options can be configured to failover NetBackup if a daemon or the option fails. For more information, see “Configuring Robotic Daemons for Monitoring (VCS for UNIX, TruCluster, or Sun Cluster)” on page 70 and “Configuring Add-ons for Monitoring in a VCS for UNIX, TruCluster, or Sun Cluster Environment” on page 71.
▼
To install a NetBackup agent or option 1. Freeze the NetBackup group: hagrp -freeze nbu_group 2. On each inactive node in the cluster on which NetBackup is installed, install the agent or option. See the appropriate NetBackup documentation for the agent or option for installation and configuration instructions. 3. On the active node, install the agent or option. 4. Unfreeze the NetBackup group.
Chapter 5, NetBackup in VERITAS Cluster Server on UNIX
45
Upgrading a NetBackup Failover Server
Upgrading a NetBackup Failover Server
Follow these instructions when upgrading a NetBackup failover server from version 5.0 to 5.1. If upgrading from 4.5 FPx or 4.5 MPx, refer to “Upgrading a NetBackup Failover Server from 4.5FPx or 4.5 MPx to 5.1” on page 47. Note VERITAS highly recommends that you use the latest VCS for UNIX agent that is installed with NetBackup 5.1. Older versions of the agent are not supported. Also, the cluster_config script is only supported with versions of the VCS for UNIX agent included with NetBackup 5.0 or later.
▼
To upgrade a NetBackup failover server 1. Ensure a good backup of your cluster environment exists that includes the NetBackup databases and catalogs. 2. Freeze the NetBackup group. 3. On each inactive node to which NetBackup may failover, install the NetBackup server software. Follow the instructions for “Upgrading NetBackup on UNIX Servers and Clients” as described in the NetBackup Installation Guide for UNIX. If required to specify the server name, provide the virtual name of the server. 4. On the active node, install the NetBackup server software. Follow the instructions for “Upgrading NetBackup on UNIX Servers and Clients” as described in the NetBackup Installation Guide for UNIX. 5. On each node on which NetBackup server is installed, upgrade any NetBackup agents or options. Refer to the section, “Installing NetBackup Agents and Options” on page 45. For detailed instructions refer to the system administrator’s guides for those agents and options. 6. Unfreeze the NetBackup group. 7. Start the daemons with the hagrp -online command. 8. Verify that NetBackup can failover properly in the cluster. Refer to “Verifying NetBackup Configuration” on page 69.
46
NetBackup High Availability System Administrator’s Guide
Upgrading a NetBackup Failover Server from 4.5FPx or 4.5 MPx to 5.1
Upgrading a NetBackup Failover Server from 4.5FPx or 4.5 MPx to 5.1
The process of upgrading a NetBackup failover server from version 4.5FPx or 4.5MPx to 5.1 includes upgrading the databases on the shared disk, removing NetBackup from the cluster, removing NetBackup from the nodes, and, finally, reinstalling NetBackup on the cluster. Note VERITAS highly recommends that you use the latest VCS for UNIX agent that is installed with NetBackup 5.1. Older versions of the agent are not supported. The cluster_config script is also not supported with earlier versions of the VCS for UNIX agent. Caution A NetBackup failover server must be upgraded according to the instructions below. If these steps are not followed, NetBackup may not function properly in the cluster.
Requirements
◆ ◆
A good backup must exist of your cluster environment, including the NetBackup databases and catalogs. The VCS executable directory is included in your path environment variable. The default path is /opt/VRTSnbu/bin.
Upgrade Instructions
▼
To upgrade a NetBackup failover server 1. Freeze the NetBackup group. 2. Edit the bp.conf and vm.conf files. a. If a REQUIRED_INTERFACE entry exists, replace it with a CLUSTER_NAME entry. Otherwise add a new CLUSTER_NAME entry. This entry should be defined as the virtual server name. b. For a master server, ensure that the first SERVER entry in the bp.conf file matches the CLUSTER_NAME entry.
Chapter 5, NetBackup in VERITAS Cluster Server on UNIX
47
Upgrading a NetBackup Failover Server from 4.5FPx or 4.5 MPx to 5.1
3. On the active node, install the NetBackup server software. Follow the instructions for “Upgrading NetBackup on UNIX Servers and Clients” as described in the NetBackup Installation Guide for UNIX. 4. Unfreeze the NetBackup group. 5. Bring the NetBackup group offline. This will unmount the shared disk. 6. Put the cluster configuration into writable mode:
haconf -makerw
7. Remove the old NetBackup group, agent, and type from the cluster. a. Remove each resource in the NetBackup group. Use the following command:
hares -delete
b. Remove the NetBackup group. Use the following command:
hagrp -delete
c. On each node in the cluster, stop the NetBackup agent:
haagent -stop NetBackup -sys
d. Remove the NetBackup resource type.
hatype -delete NetBackup
e. Remove the agent from each node in the cluster:
◆ ◆ ◆ ◆
On Solaris, run pkgrm VRTSvcsnb On AIX, run installp -u VRTSvcsnb.rte On HP, run swremove VRTSvcsnb On Linux, run rpm -ev VRTSvcsnb
Then, run the following command: rm -rf /opt/VRTSvcs/bin/NetBackup 8. Dump the cluster configuration and return to read-only mode:
haconf -dump -makero
48
NetBackup High Availability System Administrator’s Guide
Upgrading a NetBackup Failover Server from 4.5FPx or 4.5 MPx to 5.1
9. On each node to which NetBackup may failover: a. Remove the NetBackup software.
◆
On Solaris, run pkgrm to remove the server software and any options or agents. Answer “no” when asked “Are you running pkgrm as part of an upgrade process?”. On all other platforms, remove everything in the /usr/openv directory.
◆
b. Install the NetBackup server software. Follow the instructions for “Installing NetBackup Server Software” as described in the NetBackup Installation Guide for UNIX. The virtual name should be used for the server name during the install procedure. Note The installation may appear to have stopped while NetBackup attempts to set the global device database hostname to the virtual hostname. This process has to timeout since the virtual hostname is not yet active on any of the nodes. Note Do not perform an upgrade installation on the nodes. NetBackup 5.1 is installed differently in a cluster than NetBackup 4.5FPx and 4.5MPx. An upgrade installation is only performed on the shared disk (as in step 4). 10. To configure the cluster, run the following command:
/usr/openv/netbackup/bin/cluster/cluster_config -vcs_agent_upgrade
When prompted for the block device, use the same block device for the shared disk that you used when installing NetBackup 4.5_FPx or 4.5_MPx. It is recommended that you use the default mount point, /opt/VRTSnbu. 11. If you have devices attached to your failover server, run the Device Discovery wizard from each node in the cluster. Refer to “Device Configuration Guidelines” on page 64 and “Configuring Devices in a Cluster” on page 65 for guidelines and instructions on configuring devices in a cluster. Use the following command to switch to each node:
hagrp -switch nbu_group -to
12. On each node on which NetBackup server is installed, upgrade any NetBackup agents or options. Refer to the section, “Installing NetBackup Agents and Options” on page 45. For detailed instructions refer to the system administrator’s guides for those agents and options.
Chapter 5, NetBackup in VERITAS Cluster Server on UNIX
49
Upgrading a NetBackup Failover Server from 4.5FPx or 4.5 MPx to 5.1
13. Verify that NetBackup can failover properly in the cluster. Refer to “Verifying NetBackup Configuration” on page 69. 14. Start the daemons on the desired node.
hagrp -online nbu_group -sys
50
NetBackup High Availability System Administrator’s Guide
Installing a Non-Failover NetBackup Server
Installing a Non-Failover NetBackup Server
▼
To install a NetBackup non-failover server on a VCS for Solaris cluster 1. Review the “Installation Prerequisites” on page 35. 2. Install the NetBackup server software on the local disk of each desired node in the cluster. Follow the instructions for installing NetBackup server software as described in the NetBackup Installation Guide for UNIX. 3. Configure each NetBackup server. Refer to the NetBackup Installation Guide for UNIX, and the NetBackup System Administrator's Guide for UNIX, Volume I, for information on configuring NetBackup server software.
Upgrading a NetBackup Non-Failover Server
Note Converting an existing non-failover NetBackup server to a highly available failover NetBackup server is not supported. 1. Ensure a good backup of your cluster environment exists that includes the NetBackup databases and catalogs. 2. On each node on which NetBackup is installed, install the NetBackup server software on the local disk. Follow the instructions for upgrading NetBackup on UNIX servers and clients as described in the NetBackup Installation Guide for UNIX.
Installing the NetBackup Client Only
Install NetBackup client software on a local disk on all desired nodes in the cluster. Follow the instructions for a standard client install as described in the NetBackup Installation Guide for UNIX.
Performing Backups and Restores in a VCS Cluster
The process of configuring and performing backups and performing restores in a cluster is similar no matter which clustering solution you are using. Guidelines and instructions for backups and restores are covered in the chapter “Using NetBackup to Perform Backups and Restores in a Cluster” on page 73.
Chapter 5, NetBackup in VERITAS Cluster Server on UNIX 51
Performing Backups and Restores in a VCS Cluster
52
NetBackup High Availability System Administrator’s Guide
NetBackup in Sun Cluster
6
This chapter provides information on how to install and configure NetBackup in a Sun Cluster. Refer to the NetBackup System Administrator's Guide for UNIX, Volume I, for a detailed understanding of how NetBackup is installed and administered. Refer to the documentation for Sun Cluster for a detailed understanding of how Sun Cluster works, how it is installed and administered, and for information on the SunPlex cluster system.
53
Limitations
Limitations
NetBackup does not support the HAStorage Plus option for Sun Cluster.
Installation Prerequisites
Complete the following tasks before installing and configuring NetBackup in a Sun Cluster. ❏ Verify that your hardware is supported by both Sun Cluster and NetBackup. For a list of supported storage devices, see the NetBackup Release Notes for UNIX and Windows and the Sun Cluster documentation. Verify that Sun Cluster 3.0, Update 3, or Sun Cluster 3.1, Update 1, is correctly installed and configured. Setup a virtual name for NetBackup in NIS, DNS or /etc/hosts. This hostname will be used as a logical hostname during the NetBackup installation and configuration of the cluster. It will not be assigned to any physical node, but will be attached to the NetBackup application resource, allowing NetBackup to failover to the appropriate node during the failover process. The shared disk must be configured and accessible to all cluster nodes on which NetBackup will be installed. See the Sun Cluster documentation for more information on how to create and configure a shared disk. Make sure the shared disk has enough space for the NetBackup database files and directories. If you want the NetBackup media server to be highly available, the robotic and tape devices must be physically connected to all the nodes that would be part of the NetBackup High Availability setup. This can be done via SCSI or fibre channel.Verify, using OS commands, that all the devices are connected properly. Refer to the Media Manager Device Configuration Guide for UNIX and Windows. Make sure that each node in the cluster, on which you wish to install NetBackup, is rsh equivalent. As the root user you need to be able to perform a remote login to each node in the cluster without entering a password. This is only necessary for installation and configuration of the NetBackup server and any NetBackup agents and options. Once installation and configuration is complete this is no longer required. You must be able to mount the disk on all nodes at the same time (i.e., a global file system). Verify you have the NetBackup Enterprise Server 5.1 installation media and a valid license key.
❏ ❏
❏
❏
❏
❏ ❏
54
NetBackup High Availability System Administrator’s Guide
Installation and Configuration
Installation and Configuration
This section describes procedures for setting up and administering NetBackup server as a failover application in a Sun Cluster.
Installation Notes
◆ ◆
These instructions assume that this is a new installation of a NetBackup failover server. You must install the same type of NetBackup server on each node to which NetBackup may failover. You cannot install a master server and a media server in a failover configuration. If you are installing a NetBackup failover media server, ensure that all media server nodes reference the same NetBackup master server name during the install. The NetBackup resource group name in a Sun Cluster is always scnb-harg. This cannot be configured by the user. Two Sun Cluster resource files are installed with NetBackup and should not be edited: scnb.VRTS and scnb.conf.
◆ ◆ ◆
Installation and Configuration Instructions
▼
To install and configure NetBackup in a Sun Cluster 1. Install the NetBackup on each node to which NetBackup may failover. Follow the instructions for installing NetBackup server software as described in the NetBackup Installation Guide for UNIX. 2. On the primary node, execute the following command: /usr/openv/netbackup/bin/cluster/cluster_config 3. You will be prompted for a hostname. This is used as a logical hostname and the directory path for the shared NetBackup space. Enter the same hostname and shared path for all the nodes. The NetBackup resource group is called sncb-harg. The NetBackup resource is called scnb-hars. 4. Follow the onscreen directions to configure the type of NetBackup server you will use.
Chapter 6, NetBackup in Sun Cluster
55
Installation and Configuration
5. Freeze the NetBackup group by executing the following:
scrgadm -c -j scnb-hars -y Retry_count=0 /usr/cluster/bin/scswitch -M -n -j scnb-hars
Note Additional commands are required to properly freeze and unfreeze NetBackup in a SunCluster. This is due to Sun bug #5017826. When NetBackup is started under pmf with automatic restarts enabled, pmf will intervene if the tag dies, irrespective of whether the monitor is enabled or not. To freeze NetBackup, execute: scrgadm -c -j scnb-hars -y Retry_count=0 scswitch -n -M -j scnb-hars To unfreeze NetBackup, execute: scswitch -e -M -j scnb-hars scrgadm -c -j scnb-hars -y Retry_count=X Where X is the default Retry_count value of 2, or whatever value you specified before installing NetBackup. 6. On each node on which NetBackup server is installed, install any NetBackup agents or options. Refer to the next section, “Installing NetBackup Agents and Options.” 7. Unfreeze the NetBackup group:
/usr/cluster/bin/scswitch -M -e -j scnb-hars scrgadm -c -j scnb-hars -y Retry_count=X
Note The default value for the Retry_count is 2. If you have not used this default, then specify the value set before installing NetBackup. 8. Continue with the configuration of NetBackup.
◆ ◆
Refer to “NetBackup Configuration Guidelines” on page 64 for guidelines on configuring NetBackup. Refer to “Device Configuration Guidelines” on page 64 and “Configuring Devices in a Cluster” on page 65 for guidelines and instructions on configuring devices in a cluster. Note that devices must be configured on each node in the cluster. See “Verifying NetBackup Catalog Backup Entries” on page 68 for instructions on verifying that the entries for the NetBackup Catalog backups are correct.
◆
56
NetBackup High Availability System Administrator’s Guide
Installation and Configuration
9. When you have completed configuration of NetBackup, verify that NetBackup can fail over properly in the cluster. Refer to “Verifying NetBackup Configuration” on page 69.
Installing NetBackup Agents and Options
To install agents and options, NetBackup daemons must be brought offline without causing a failover. To accomplish this, Sun Cluster must be configured to stop monitoring NetBackup, but keep NetBackup resources online. By default, robotic daemons and NetBackup options (such as NetBackup Advanced Reporter) will not cause NetBackup to failover if they fail. You can configure robot daemons and NetBackup options to failover NetBackup if a daemon or the option fails. For more information, see “Configuring Robotic Daemons for Monitoring (VCS for UNIX, TruCluster, or Sun Cluster)” on page 70 and “Configuring Add-ons for Monitoring in a VCS for UNIX, TruCluster, or Sun Cluster Environment” on page 71.
▼
To install NetBackup agents and options 1. Place NetBackup in the unmanaged state by executing the following:
scrgadm -c -j scnb-hars -y Retry_count=0 /usr/cluster/bin/scswitch -M -n -j scnb-hars
2. On each inactive node on which NetBackup server is installed, install or upgrade any agents or options. See the appropriate NetBackup documentation for the agent or option for installation and configuration instructions. 3. Perform the same installation on the active node. 4. After completing the installs for the agents or options, execute the following:
/usr/cluster/bin/scswitch -M -e -j scnb-hars scrgadm -c -j scnb-hars -y Retry_count=X
Note The default value for the Retry_count is 2. If you have not used this default, then specify the value set before installing NetBackup. The NetBackup resources will be brought online and Sun Cluster will resume monitoring the NetBackup resource.
Chapter 6, NetBackup in Sun Cluster
57
Upgrading a NetBackup Failover Server
Upgrading a NetBackup Failover Server
Follow these instructions if you are upgrading from NetBackup 5.0 to 5.1. If you are upgrading from NetBackup 4.5 FP3, see “Upgrading from NetBackup 4.5FP3 or later (With Sun Microsystem’s NetBackup Agent) to 5.1” on page 59. Note These upgrade instructions only support upgrading if you are using a globally shared disk. NetBackup 5.1 does not support upgrading an HASP configuration. If you are using an HASP configuration, convert your disk to be a global disk (each machine can access the disk simultaneously).
▼
To upgrade a NetBackup failover server 1. Ensure a good backup of your cluster environment exists that includes the NetBackup databases and catalogs. 2. Put NetBackup in the unmanaged state by executing the following command:
scrgadm -c -j scnb-hars -y Retry_count=0 /usr/cluster/bin/scswitch -M -n -j scnb-hars
3. On each inactive node on which NetBackup server is installed, upgrade NetBackup. Follow the instructions for “Upgrading NetBackup on UNIX Servers and Clients” as described in the NetBackup Installation Guide for UNIX. 4. On the active node, perform the same upgrade installation. 5. On each node on which NetBackup server is installed, upgrade any NetBackup agents or options. Refer to the section “Installing NetBackup Agents and Options” on page 57. For detailed instructions refer to the system administrator’s guides for those agents and options. 6. Execute the following command:
/usr/cluster/bin/scswitch -M -e -j scnb-hars scrgadm -c -j scnb-hars -y Retry_count=X
Note The default value for the Retry_count is 2. If you have not used this default, then specify the value set before installing NetBackup. The NetBackup resources will be brought online and Sun Cluster will resume monitoring the NetBackup resource. 7. Verify that NetBackup can failover properly in the cluster. Refer to “Verifying NetBackup Configuration” on page 69.
58 NetBackup High Availability System Administrator’s Guide
Upgrading from NetBackup 4.5FP3 or later (With Sun Microsystem’s NetBackup Agent) to 5.1
Upgrading from NetBackup 4.5FP3 or later (With Sun Microsystem’s NetBackup Agent) to 5.1
Follow these instructions if you have installed NetBackup failover server on a Sun Cluster and are using Sun Microsystem’s agent for NetBackup 4.5FP3 or later. Caution A NetBackup failover server must be upgraded according to the instructions below. If these steps are not followed, NetBackup may not function properly in the cluster. Note These upgrade instructions only support upgrading if you are using a globally shared disk. NetBackup 5.1 does not support upgrading an HASP configuration. If you are using an HASP configuration convert your disk to be a global disk (each machine can access the disk simultaneously).
▼
To upgrade NetBackup with Sun Microsystem’s agent 1. Ensure a good backup of your cluster environment exists that includes the NetBackup databases and catalogs. 2. Bring NetBackup offline with the following command:
/usr/cluster/bin/scswitch -F -g netbackup resource group
3. Create a temporary resource group while for upgrading NetBackup:
scrgadm -a -g scnb-harg
Note You must use scnb-harg as the name of the resource group. This is a temporary group that will be removed after the upgrade. 4. Register the virtual name in the above group:
scrgadm -a -L -g scnb-harg -l virtual-name
5. Bring the NetBackup group online:
scswitch -Z -g scnb-harg
6. In the bp.conf and vm.conf files, change all REQUIRED_INTERFACE entries to CLUSTER_NAME.
Chapter 6, NetBackup in Sun Cluster
59
Upgrading from NetBackup 4.5FP3 or later (With Sun Microsystem’s NetBackup Agent) to 5.1
7. Copy all NetBackup databases and files on the shared disk back to their default locations. This includes but is not limited to the following: /usr/openv/db/var /usr/openv/netbackup/db /usr/openv/netbackup/bp.conf /usr/openv/volmgr/database /usr/openv/volmgr/vm.conf 8. On all the other nodes, remove all the links to files and databases on the shared disk. 9. On each of the nodes, edit the bp.conf so the CLIENT_NAME entry is set to the hostname of the node. 10. On the shared disk, copy the netbackup directory to netbackup.orig. 11. On the active node, perform an upgrade installation. Follow the instructions for “Upgrading NetBackup on UNIX Servers and Clients” as described in the NetBackup Installation Guide for UNIX. 12. During the install, choose to start up all the daemons. 13. Perform the same upgrade on the inactive nodes. 14. On the inactive nodes, bring the daemons down:
/usr/openv/netbackup/bin/goodies/bp.kill_all FORCEKILL
15. Remove the temporary scnb-harg group that was created for the upgrade by running the following commands:
scswitch -F -g scnb-harg scswitch -n -j scnb-hars scswitch -n -j virtual-name scrgadm -r -j scnb-hars scrgadm -r -j virtual-name scrgadm -r -g scnb-harg
60
NetBackup High Availability System Administrator’s Guide
Bringing the NetBackup Resource Group Online/Offline
16. On the active node, run the cluster_config utility. 17. Bring the cluster online with using the following command:
/usr/cluster/bin/scswitch -Z scnb-harg
From now on start and stop NetBackup with scswitch. The resource group is named scnb-harg. The name of the resource group is not configurable.
Bringing the NetBackup Resource Group Online/Offline
To bring the NetBackup resource group online or offline, special Sun Cluster commands must be used. See the Sun Cluster documentation for more information on these and other commands.
Sun Cluster commands To: Enter this command:
bring the NetBackup resource group online bring the NetBackup resource group offline verify that the NetBackup resource is online
/usr/cluster/bin/scswitch -z -g scnb-harg -h
/usr/cluster/bin/scswitch -F -g scnb-harg
/usr/cluster/bin/scstat -g
Performing Backups and Restores in a Sun Cluster
The process of configuring and performing backups and performing restores in a cluster is similar no matter which clustering solution you are using. Guidelines and instructions for backups and restores are covered in the chapter “Using NetBackup to Perform Backups and Restores in a Cluster” on page 73.
Chapter 6, NetBackup in Sun Cluster
61
Performing Backups and Restores in a Sun Cluster
62
NetBackup High Availability System Administrator’s Guide
Configuring NetBackup
7
Configuration of NetBackup includes configuring storage devices, volumes, the catalog backup, and creating backup policies. This configuration can be completed most easily by using the NetBackup Configuration Wizards in the NetBackup Administration Console. Refer to the NetBackup System Administrator's Guide for Windows, Volume I, or the NetBackup System Administrator's Guide for UNIX, Volume I, for more details for instructions on performing these steps manually (See the section, “Configuring NetBackup Without Wizards.”). Before configuring NetBackup, refer to the following sections in this chapter.
Section “NetBackup Configuration Guidelines” on page 64 “Device Configuration Guidelines” on page 64 “Configuring Devices in a Cluster” on page 65 “Verifying NetBackup Catalog Backup Entries” on page 68
Description Guidelines to follow when configuring the NetBackup server in a cluster. Guidelines to follow when configuring devices in a cluster. Instructions for configuring devices in a cluster.
Applicable only to NetBackup installed in a MSCS or VCS for NT cluster. Instructions for verifying the NetBackup Catalog entries.
63
NetBackup Configuration Guidelines
NetBackup Configuration Guidelines
◆
The NetBackup server list must be set up correctly on each node where NetBackup is configured as a failover server. If NetBackup is installed as a master server in the cluster, the virtual server name must appear first, followed by the node names within the NetBackup group. If NetBackup is installed as a media server in the cluster, then the name of the master must appear first in the server list and the virtual server name must be included elsewhere in the server list. NetBackup depends on names to route information between participants in the backup and restore process. Therefore, it is very important that all name usage be consistent within a NetBackup configuration. Specific node names and cluster aliases should not be used interchangeably. (VCS, MSCS, and Sun Cluster only) CLIENT_NAME should be set to the name of the node. CLUSTER_NAME should be set to the virtual server name. With the exception of NetBackup in a TruCluster, when making changes to the NetBackup configuration on the active node, be sure to apply the same changes to each node in the cluster. Any configuration changes should be verified by moving the NetBackup server to each node to check for the correct setting. Configuration information for backup policies, storage units, and the NetBackup catalog backup is stored on the shared disk and cannot be configured separately for each node. Use the NetBackup Administration Console as much as possible to configure and administer NetBackup.
◆
◆ ◆ ◆
◆ ◆
◆
Device Configuration Guidelines
The following are guidelines for avoiding problems as you configure the devices for a clustered NetBackup server:
◆
If there are devices attached to one node but not to the other node, they are available only when the node where they are attached is online. For a NetBackup failover server, it is usually best to attach all of the devices to each node on which NetBackup is installed using shared SCSI or fibre channel connections as necessary to share the devices. For NetBackup media server failover, configure devices from the NetBackup master server using the cluster alias or virtual server name as the NetBackup media server name.
◆
64
NetBackup High Availability System Administrator’s Guide
NetBackup Configuration Guidelines
In a TruCluster, device paths should be the same on each node in the cluster.
◆
To keep the configuration as straightforward as possible, specify the NetBackup master server as the volume database host for robots in the cluster and as the standalone volume database host for standalone tape devices in the cluster. For tape devices that are in a robotic library, ensure the robot drive number field is set correctly according to the drive numbering scheme implemented by the manufacturer of the robotic library. Note that the first drive in the robot is always considered robot drive number 1 for NetBackup; if the manufacturer’s drive numbering scheme starts with a different number such as 0, adjust it accordingly. Ensure that the robot numbers you define for robots used by the failover server are consistent on all servers that use that robot. If the robot number defined for the robot on one node does not match the number defined on another node for the same robot, problems will occur when attempting backups. When adding devices or including a new device host, specify the virtual server name of the NetBackup failover server in the Device host field in the Media and Device Management utility. For automatic discovery of devices, specify the virtual server name as the device host to scan. Always use the virtual server name of the NetBackup failover server in the NetBackup media server field (Windows interface) or Media server field (Java interface) when creating storage units that reside on the cluster.
◆
◆
◆
◆ ◆
Configuring Devices in a Cluster
This section describes briefly how to configure devices in a clustered environment. For complete details on this topic, refer to the NetBackup Media Manager System Administrator's Guide for Windows or NetBackup Media Manager System Administrator's Guide for UNIX. Note To accommodate disparate SCSI connections on each node of the cluster, you must configure devices on each node in a MSCS, VCS, Sun Cluster environment. First, configure the devices on the active node. Then move the NetBackup group to another node and configure the devices on that node, move NetBackup to another node and configure the devices on that node, etc. In a TruCluster, devices need only be configured once.
Chapter 7, Configuring NetBackup
65
NetBackup Configuration Guidelines
Configuring Devices for MSCS, VCS, or Sun Cluster
▼
To configure devices for MSCS, VCS, or Sun Cluster 1. Review the guidelines in the section, “Device Configuration Guidelines.” 2. On the active node, run the Device Configuration wizard. This wizard will automatically discover and configure the devices and create storage units for those devices. Using the Device Configuration wizard is the recommended method of configuring devices. To add robots and drives manually, refer to the “Configuring and Managing Devices section in the NetBackup System Administrator's Guide for Windows, Volume I, or NetBackup System Administrator's Guide for UNIX, Volume I. Note Be sure to use the virtual server name as the device host to scan. Note If you are using one or more standalone tape drives, the Standalone Volume Database Host setting must be changed to the appropriate NetBackup server setting (usually the NetBackup master server). NetBackup does not change this setting automatically, so you must verify the setting is correct. 3. Move the NetBackup group to another node. 4. Refresh the view in the Media and Device Management utility. 5. Run the Device Configuration wizard. Again, use the virtual server name when specifying the device host. Note You may observe some transient errors while the cluster is partially configured. For example, the NetBackup Administration Console may indicate that it cannot synchronize the global database. This error will be resolved once the devices are configured on all nodes in the cluster.
66
NetBackup High Availability System Administrator’s Guide
NetBackup Configuration Guidelines
6. Continue configuring the devices on each node in the cluster by repeating step 3 through step 5. Note If device paths are not displayed after refreshing the Devices node or running tpconfig, the active node of the cluster is not configured properly. Note If you wish to configure NetBackup to failover if a robotic device fails, refer to the section, “Configuring Robotic Daemons for Monitoring (VCS for UNIX, TruCluster, or Sun Cluster).”
▼
To add new devices Follow the same procedure as when initially configuring devices in a cluster. First, configure the devices on the active node. Then move the NetBackup group to another node and configure the devices on that node, move NetBackup to another node and configure the devices on that node, and so on.
Configuring Devices for a TruCluster
▼
To configure devices in a TruCluster 1. Review the guidelines in the section, “Device Configuration Guidelines.” 2. On the active node, run the Device Configuration wizard. This wizard will automatically discover and configure the devices and create storage units for those devices. Using the Device Configuration wizard is the recommended method of configuring devices. To add robots and drives manually, refer to the “Configuring and Managing Devices section in the NetBackup System Administrator's Guide for Windows, Volume I, or NetBackup System Administrator's Guide for UNIX, Volume I. Note Be sure to use the virtual server name as the device host to scan. Note If you are using one or more standalone tape drives, the Standalone Volume Database Host setting must be changed to the appropriate NetBackup server setting (usually the NetBackup master server). NetBackup does not change this setting automatically, so you must verify the setting is correct. Note If you wish to configure NetBackup to failover if a robotic device fails, refer to the section, “Configuring Robotic Daemons for Monitoring (VCS for UNIX, TruCluster, or Sun Cluster).”
Chapter 7, Configuring NetBackup
67
NetBackup Configuration Guidelines
Verifying NetBackup Catalog Backup Entries
After you have configured the NetBackup server, the Files list entries for the NetBackup Catalog need to be verified and changed, if necessary. The Files list for the Catalog Backup can be accessed as follows:
Windows Servers
▼
To configure the Catalog Backup file list 1. Open the NetBackup Administration Console, 2. Right-click on Catalog and choose Configure NetBackup Catalog Backup. 3. Click on the Files tab. Verify that there are entries in the files list for the Catalog Backup for each of the nodes to which NetBackup can failover. The list must contain three entries for each node: nodename:install_path\NetBackup\db nodename:install_path\Volmgr\database nodename:install_path\NetBackup\var Where nodename is the physical node name of that node and install_path is the path where NetBackup is installed on the local drive. The list must also contain the following two entries for the virtual server: virtual_server_name:shared_disk_path\NetBackup\db virtual_server_name:shared_disk_path\Volmgr\database virtual_server_name:shared_disk_path\Volmgr\misc Where virtual_server_name is the virtual server name of the NetBackup failover server and shared_disk_path is the path to the NetBackup files on the shared disk.
UNIX Servers
▼
To configure the Catalog Backup file list 1. Open the NetBackup Administration Console, 2. Right-click on Catalog and choose Configure NetBackup Catalog Backup. 3. Click on the Files tab.
68
NetBackup High Availability System Administrator’s Guide
NetBackup Configuration Guidelines
4. Verify that there are entries in the files list for the Catalog Backup for each of the nodes to which NetBackup can failover. virtual_name:/usr/openv/var virtual_name:/path/to/shared/netbackup/db virtual_name:/path/to/shared/volmgr/database/*DB virtual_name:/path/to/shared/volmgr/database/robotic_def virtual_name:/path/to/shared/volmgr/database/ltidevs If there is anything present in the volmgr/database directory besides what is listed above (for example, NDMP devices) add entries for these to the Files list.
Verifying NetBackup Configuration
After completing configuration of storage devices, volumes, the catalog backup, and backup policies, verify that NetBackup fails over properly in the cluster. 1. Confirm that the NetBackup is online on the primary node, then determine if the following can be completed successfully:
◆ ◆ ◆ ◆
scheduled backups user-directed backups NetBackup database backups server- and user-directed restores
2. Switch the NetBackup resource group to another node and perform the same tests as above. Repeat for each node in the cluster on which NetBackup is installed.
Chapter 7, Configuring NetBackup
69
Configuring Robotic Daemons for Monitoring (VCS for UNIX, TruCluster, or Sun Cluster)
Configuring Robotic Daemons for Monitoring (VCS for UNIX,
TruCluster, or Sun Cluster)
By default, robotic daemons will not cause NetBackup to failover if they fail. You can configure robot daemons to failover NetBackup if a daemon fails. Note You must configure robots before performing the following instructions. Refer to the NetBackup Installation Guide for Windows or NetBackup Installation Guide for UNIX for instructions on configuring storage devices.
▼
To configure NetBackup to failover if a robotic daemon fails 1. Configure your robots on each node in the cluster. 2. Run the following script:
/usr/openv/netbackup/bin/cluster_config -r
3. Choose a (Add). 4. Select your robot from the list. Multiple robots can be selected.
▼
To remove a robotic daemon from monitoring 1. Run the following script:
/usr/openv/netbackup/bin/cluster_config -r
2. Choose d (Delete). 3. Select the robot type you would like to remove.
Configuring Add-Ons for Monitoring
By default, add-on products (such as NetBackup Advanced Reporter) will not cause NetBackup to failover if they fail. These products can be configured to failover NetBackup if an add-on product fails.
70
NetBackup High Availability System Administrator’s Guide
Configuring Add-Ons for Monitoring
Configuring Add-ons for Monitoring in a VCS for UNIX, TruCluster, or Sun Cluster Environment
▼
To monitor an add-on and failover NBU when the add-on fails 1. Run the following script:
/usr/openv/netbackup/bin/cluster_config -a
2. Choose a (Add). 3. Select your add-on from the list. Multiple add-ons can be selected.
▼
To remove an add-on from monitoring 1. Run the following script:
/usr/openv/netbackup/bin/cluster_config -a
2. Choose d (Delete). 3. Select the add-on you wish to delete from the list.
Configuring Add-ons for Monitoring in a MSCS Environment
▼
To monitor an add-on service and failover NetBackup when the add-on service fails 1. Open the Cluster Administrator. 2. Select the NetBackup group. 3. In the right-hand pane, double-click on the resource you want to monitor. The properties for the resource are displayed. 4. Click on the Advanced tab. 5. In the Restart group, select Affects the group. 6. Click Apply. 7. Click OK.
Chapter 7, Configuring NetBackup
71
Configuring Add-Ons for Monitoring ▼
To remove an add-on service from monitoring 1. Open the Cluster Administrator. 2. Select the NetBackup group. 3. In the right-hand pane, double-click on the resource you want to remove from monitoring. The properties for the resource are displayed. 4. Click on the Advanced tab. 5. In the Restart group, deselect Affects the group. 6. Click Apply. 7. Click OK.
72
NetBackup High Availability System Administrator’s Guide
Using NetBackup to Perform Backups and Restores in a Cluster
This chapter provides information on what data should be backed up in a cluster and guidelines on how to back up and restore data in a cluster. Specific instructions for performing backups and restores can be found in the NetBackup Backup, Archive, and Restore Getting Started Guide and in the system administrator guides for NetBackup’s options and agents.
8
73
What to Back Up
What to Back Up
Back up the following to protect all data in the cluster, including file systems and databases:
◆ ◆ ◆
Local disks on each node. (See “Backing Up Local Disks in a Cluster” on page 74.) All disks that attach to the shared SCSI bus. (See “Backing Up the Shared Disks” on page 75.) Virtual servers, which may contain data or contain database applications. Use NetBackup database agents to back up databases. (See “Backing Up Database Files in a Cluster” on page 76.) User backups that are run on individual nodes of the cluster generally run as a backup of the node, not the NetBackup virtual server. Because of this, it is usually easier and less confusing to use scheduled backups rather than user backups to protect the data in the cluster.
MSCS and VCS for Windows Clusters In addition to the items above, be sure to back up the System State on each node. For MSCS clusters, the cluster quorum, which contains recovery information for the cluster and information about changes to the cluster configuration, is included in the System State backup (Windows 2000 or later only). (See “Backing Up Local Disks in a Cluster” on page 74.)
Backing Up Local Disks in a Cluster
To protect the data on the node’s local disks, configure a policy that includes the cluster node names in the client list. In addition to this policy, you must also configure a separate policy to back up the shared disks (see “Backing Up the Shared Disks” on page 75). The following are guidelines for configuring the policy to back up local disks:
◆ ◆
If NetBackup is installed on the cluster as a NetBackup failover media server, for the policy storage unit, select the storage unit created for the failover media server. If NetBackup is installed on the cluster as separate NetBackup media servers, configure a policy for each node. Each policy should include the cluster node name as a client. For each policy, select the storage unit created for the NetBackup media server that is installed on that cluster node.
74
NetBackup High Availability System Administrator’s Guide
Backing Up the Shared Disks
◆
In the Backup Selections list, add the ALL_LOCAL_DRIVES directive. Then add the shared disks to the Exclude List for each client (or, for UNIX clients, create an exclude list for the policy that performs backups of those clients). Adding the shared disks to the Exclude List prevents the shared disks from being backed up with each client’s local disks. For Windows 2000 or later, always include System_State:\ in the file list to back up the System State for each cluster node. In a policy for an MSCS cluster, System_State:\ will also back up the cluster quorum information.
◆
Backing Up the Shared Disks
To protect the data on the disks that attach to the shared SCSI bus, configure a policy that includes the virtual server name associated with the shared disk in the Clients list. In addition to this policy, you must also configure a separate policy to back up the local disks (see “Backing Up Local Disks in a Cluster” on page 74). The following are guidelines for configuring the policy to back up the shared disks:
◆ ◆
If NetBackup is installed on the cluster as a NetBackup failover media server, for the policy storage unit, select the storage unit created for the failover media server. If NetBackup is installed on the cluster as separate NetBackup media servers, you can select either a storage unit that was created for one of the media servers or select Any Available. If a storage unit that does not have local, direct access to the shared disk drives is used for the backup, the data is backed up across the network. This occurs even if the storage unit refers to a NetBackup media server running on a node of the cluster that does not currently have control of the shared disk drives. In the Backup Selections list, specify the drive letters or mounted directories of the shared disks.
◆
◆
Note At the time the backup is initiated, if the cluster resource that defines the virtual server name of the cluster is controlled by one node of the cluster, while the cluster resource that defines the shared disks is controlled by another node of the cluster, the backup will fail. To prevent this condition, ensure that the virtual server name and the shared disks are configured to failover together so that the same node always controls them.
◆
For Windows 2000 clusters, the System State information is included in the backup of the node’s local disks, so it is not necessary to include System_State:\ in the file list for this policy.
Chapter 8, Using NetBackup to Perform Backups and Restores in a Cluster
75
Backing Up Database Files in a Cluster
Backing Up Database Files in a Cluster
Database applications are installed on a cluster as virtual servers. To protect the data for these virtual servers, install the appropriate NetBackup database agent on each node of the cluster (With NetBackup for Windows, database agents are installed along with NetBackup server and the NetBackup client.) and a policy needs to be created for that database agent. When configuring a policy for the application/database in the cluster, always use the virtual server name of the application/database as the client name in the policy. For complete installation and configuration instructions for a particular database agent, see the NetBackup documentation for that agent.
Restoring Data in a Cluster
For all file restore operations, use the procedures on performing restores in the NetBackup Backup, Archive, and Restore Getting Started Guide. When restoring files to the shared disk drives, restore those files to the virtual server name. When restoring individual database files, restore those files to the virtual server name that corresponds to the client where the database application is installed. Note Because of the multiple virtual names that refer to the same computer in a cluster environment, it is possible for files to be backed up in the context of more than one client name. Careful planning of policies should avoid this problem. However, still remember that when restoring to a cluster, it may be necessary to browse more than one client name and perform more than one restore to restore all of the files that you need. If you are using the Backup, Archive, and Restore interface on a NetBackup client that is installed on a cluster, additional configuration steps may be necessary to restore files that were backed up using the virtual server name. Since the Backup, Archive, and Restore interface operates in the context of that client’s name, you must perform an redirected restore to restore the files on the shared disk that were backed up using the virtual server name. NetBackup allows this redirected restore operation only if the necessary configuration is performed on the NetBackup master server. This configuration consists of creating the install_path\NetBackup\db\altnames or /usr/openv/netbackup/db/altnames directory on the master server and adding files. See the “Managing Client Restores” section of the NetBackup System Administrator's Guide for Windows, Volume I, or NetBackup System Administrator's Guide for UNIX, Volume I. There may be other situations that require the appropriate altnames directory entries to be created on the master server. If while attempting to restore files from the client, the operation fails with the error 131 client is not validated to use this server message, you must set up the altnames directory to allow the operation to succeed. One example is when the required interface parameter is set to a valid network name for the
76 NetBackup High Availability System Administrator’s Guide
Restoring Data in a Cluster
client but it does not match the NetBackup client name parameter for that client. Note that this may often be true for NetBackup clients in a cluster. Of course, using a server-directed restore is always an option and avoids the need to set up the altnames directory. Example Assume the cluster virtual server name is TOE and the cluster node names are TIC and TAC. Files on the shared disk must be backed up by a NetBackup policy that includes TOE in the client list.
▼
To perform a server-directed restore of files on the shared disk Set both the source client and the destination client to TOE. The server-directed restore does not have to know which node is in control of the shared disk at the time of the restore.
▼
To perform a user-directed restore of files on the shared disk 1. Create the following files on the master server. For a UNIX server: /usr/openv/netbackup/db/altnames/tic /usr/openv/netbackup/db/altnames/tac For a Windows server: shared_drive_install_path\NetBackup\db\altnames\tic shared_drive_install_path\NetBackup\db\altnames\tac 2. In both files, add the virtual server name TOE on one line in the file. 3. Determine which node (TIC or TAC) has control of the shared disk. 4. Start the Backup, Archive, and Restore interface on that node and select the virtual server name (TOE) as the source client and the server (On Windows machines, choose File > Specify NetBackup Machines. On UNIX machines, choose Actions > Specify NetBackup Machines). 5. Browse the backed-up files by using the virtual server name (TOE) from the shared disk and restore them as needed.
Chapter 8, Using NetBackup to Perform Backups and Restores in a Cluster
77
Restoring Data in a Cluster
78
NetBackup High Availability System Administrator’s Guide
Operational Notes
Please note the following when using NetBackup in a clustered environment:
◆
9
If a NetBackup server fails over and there are backup jobs running, one of two things will happen:
◆
If Checkpoint Restart is enabled for backup jobs for the policy, any backups that were running will be resumed at the last checkpoint, assuming that the schedule window permits resuming the job. If Checkpoint Restart is not enabled for backup jobs, any backup jobs that were running will fail and show up as failed jobs in the Activity Monitor or NetBackup reports. Regularly scheduled backup jobs are retried according to the normal retry logic, assuming that the schedule window and so on permit the retry of the job. After a failover occurs, the user must restart user-directed backups and manual backups.
◆
Refer to the NetBackup System Administrator's Guide for Windows, Volume I, or the NetBackup System Administrator's Guide for UNIX, Volume I, for more information on Checkpoint Restart.
◆ ◆
If a NetBackup server fails over and there are restore jobs running, the user must manually resume the restores. After a failover occurs on a NetBackup failover server (master or media), it can take a while for the tape devices and media servers to be reset to a point where they can be used again by NetBackup. For example, a tape may have to be rewound before being available. Until the device is again ready for use by NetBackup, you may see errors such as 213 (no storage units available for use) or 219 (the required storage unit is unavailable). If you have one of the NetBackup administration interfaces currently connected to a NetBackup failover server, it may be necessary to refresh the interface after a failover occurs.
◆
79
Delay in Detecting of Loss of Connection (MSCS and VCS on Windows)
Delay in Detecting of Loss of Connection (MSCS and VCS on Windows)
There may be a delay in detecting the loss of a connection from a NetBackup Windows master server to a media server. In certain situations, there can be a delay on a NetBackup Windows master server before it detects that the connection to a media server has been aborted. For example, if a media server goes down while running a backup, there may be a delay on the master server before it detects that the media server is no longer available. While at first it may appear that there is a problem with the NetBackup Windows master server, this delay is actually a result of a certain TCP/IP configuration parameter on Windows called KeepAliveTime that is set to 7,200,000 (two hours, in milliseconds) by default. More information about the KeepAliveTime and other associated TCP/IP configuration parameters on Windows may be found in the following Microsoft Knowledge Base articles: Q140325 and Q120642. The effect of this delay is that NetBackup jobs running on that media server appear to be active for a period of time after the connection to the media server has gone down. In some cases this can result in an undesirable delay before the current backup job fails and is subjected to the normal NetBackup retry logic for execution on a different media server, if one is available. This delay is especially noticeable when the media server in question is a NetBackup failover media server running in a Microsoft Cluster Server (MSCS) environment. NetBackup relies upon the retry logic of the NetBackup master server to restart NetBackup jobs that were running on the NetBackup failover media server when a failover occurs. It may be advantageous to modify the KeepAliveTime configuration parameter on the NetBackup Windows master server for the reasons described above. However, this must be done using extreme caution because the parameter is a system wide parameter that affects all TCP/IP communications for that system. Also, it may be advantageous to modify this parameter on Windows media servers that use the failover master server.
80
NetBackup High Availability System Administrator’s Guide
MSCS Clusters
MSCS Clusters
This section describes items that should be noted before using NetBackup as a failover server in a MSCS environment. In addition to what is covered here, refer to the information above for further notes that are relevant to all cluster environments.
◆
If there are no tape devices configured on the controlling node of the NetBackup failover server, the NetBackup Device Manager service resource takes itself offline since it is not needed. If you do not intend to configure tape devices for the NetBackup failover server, you may want to delete the NetBackup Device Manager service resource from the NetBackup group. This way you can avoid seeing the group in a warning or failed state. For three- or four-node MSCS clusters, the MSCS Move Group command lists all of the nodes in the cluster, not just the ones used for the NetBackup failover server. You must pick a node that is valid for the NetBackup failover server to use. (MSCS will display all nodes in the cluster, even those nodes on which NetBackup has not been installed.)
◆
Chapter 9, Operational Notes
81
MSCS Clusters
82
NetBackup High Availability System Administrator’s Guide
Additional Configurations for NetBackup and Clusters
Several cluster configurations are supported by NetBackup; including the following:
◆
A
Cluster with Tape Devices on a Shared SCSI Bus Cluster with Locally-attached Storage Devices Clusters Sharing Devices on a Fibre Channel SAN
◆ ◆
Combinations of these configurations can also be used.
83
Cluster with Tape Devices on a Shared SCSI Bus
Cluster with Tape Devices on a Shared SCSI Bus
In this configuration, a NetBackup failover server (master or media) is installed on the cluster. Only one of the NetBackup servers installed in the cluster is active at any point in time. Each node in the cluster has access to the tape devices. Access to the tape devices is controlled by the fact that only one node of the NetBackup failover server can be active at any point in time.
Example of a cluster with NetBackup and tape devices on a shared SCSI bus
NetBackup Virtual Server
Controlling Node Shared SCSI Bus for Disks
Failover Node
Shared SCSI Bus for Tape Private Network LAN
The devices are configured on each of the failover server nodes independently from the other node. This allows different SCSI bus numbers, and so on, on each of the nodes to refer to the same device. Remember, however, that any NetBackup storage units you create for the failover server must be configured with the actual number of devices. For example, do not increase the number of drives since there are multiple connections to the same tape device. In certain situations, it is possible for a SCSI bus reset to be issued for the SCSI bus where the shared tape devices attach. Therefore, do not connect tape devices and shared disk drives to the same SCSI bus; connect each to separate SCSI buses.
Configuring a Shared SCSI Bus for Tape Devices
Before configuring a shared SCSI bus for tape devices, read the following carefully. To configure tape devices on a shared SCSI bus, you must have the following.
◆ ◆
SCSI cables SCSI terminators
84
NetBackup High Availability System Administrator’s Guide
Cluster with Tape Devices on a Shared SCSI Bus
◆ ◆
A SCSI adapter in each cluster server to provide a shared external bus between the nodes At least one tape device on the shared bus
The tape devices must connect to a bus that uses the same method of transmission as the device (single-ended, differential, or low-voltage differential). Only one transmission method can be used on a single SCSI bus. Terminate the SCSI bus at both ends so commands and data can be transmitted to and from all devices on the bus. Each SCSI bus must have two terminators and they must be at each end of the segment. If a tape device is in the middle of the bus as shown in the preceding figure (“Example of a cluster with NetBackup and tape devices on a shared SCSI bus”) remove any internal termination in that device. If the tape device is at the end of the bus as shown in the following figure (“Example of a shared bus with tape devices at the end of the bus”) and the tape device has internal termination, you can use the device’s internal termination to terminate the bus.
Example of a shared bus with tape devices at the end of the bus
NetBackup Virtual Server
Controlling Node Shared SCSI Bus for Disks Shared SCSI Bus for Tape Private Network
Failover Node
LAN
◆
SCSI adapters. This method is not recommended because if the server is disconnected from the shared bus, or if there is a power supply failure, the bus may not be properly terminated and be inoperable. Pass-through (or feed-through) SCSI terminators. You can use these with SCSI adapters and with some tape devices. If the tape device is at the end of the bus, you can attach a pass-through SCSI terminator to terminate the bus. The internal terminators in the tape device must be disabled.
◆
Appendix A, Additional Configurations for NetBackup and Clusters
85
Cluster with Tape Devices on a Shared SCSI Bus
Tip
To ensure termination if a power supply failure occurs, turn off the on-board terminators on the SCSI controller (using the method that the host adapter manufacturer recommends) and physically terminate the controller with a terminator. Y cables. You can use these with some tape devices. If the tape device is at the end of the bus, you can attach a terminator to one branch of the Y cable to terminate the bus. The internal terminators in the tape device must be disabled. Trilink connectors. You can use these with some tape devices. If the tape device is at the end of the bus, you can attach a terminator to one of the trilink connectors to terminate the bus. The internal terminators in the tape device must be disabled.
◆
◆
Besides terminating the bus, Y-cables and trilink connectors also let you isolate the devices from the shared bus without affecting the bus termination. You can maintain or remove that device without affecting the other devices on the shared SCSI bus.
▼
To configure a shared SCSI bus for tape devices 1. Install the SCSI controllers for the shared SCSI bus. Ensure that the SCSI controllers for the shared SCSI bus are using different SCSI IDs. For example, on the controlling node, set the SCSI controller ID to 6 and on the failover node set the SCSI controller ID to 7. 2. Prepare the SCSI controllers for the shared SCSI bus. For details, see the documentation provided by your SCSI host adapter manufacturer. Note Do not have power on both nodes while configuring the systems. Or, if both nodes have power on, do not connect the shared SCSI buses to both nodes. 3. Connect the shared SCSI tape devices to the cable, connect the cable to both nodes, then terminate the bus segment using one of the methods discussed in the previous section.
86
NetBackup High Availability System Administrator’s Guide
Cluster with Locally-attached Storage Devices
Cluster with Locally-attached Storage Devices
In this configuration, a separate NetBackup media server is installed on each desired node, storage devices are locally attached to those nodes, and no failover protection is available for the NetBackup media servers.
Cluster with NetBackup and locally-attached storage devices
Controlling Node Locally-Attached Storage Devices Shared SCSI Bus for Disks Heartbeat
Failover Node
Locally-Attached Storage Devices
LAN
Appendix A, Additional Configurations for NetBackup and Clusters
87
Clusters Sharing Devices on a Fibre Channel SAN
Clusters Sharing Devices on a Fibre Channel SAN
In this configuration, one or more clusters attach to a fibre channel storage area network (SAN), with a NetBackup failover server (master or media) installed on the cluster. Tape devices are shared through the SAN. This configuration offers the normal advantages of a SAN backup environment such as performing backups locally to tape devices through the fibre connections, instead of across the network to a separate NetBackup server. This configuration insures that backups and restores will never use the network for data communications. The prioritized storage unit group makes this possible by using the local storage unit if it is available, but in the case of a machine failover, the other storage unit will be used. In the case of the machine failover, the virtual print server will have moved to the other node as well, so this keeps the backups/restores local. For complete details on configuring devices in a cluster with a fibre channel SAN, refer to the NetBackup Media Manager System Administrator's Guide for Windows or the NetBackup Media Manager System Administrator's Guide for UNIX.
88
NetBackup High Availability System Administrator’s Guide
bpclusterutil
B
In an MSCS cluster, under normal circumstances, cluster configuration is one of the final steps in installing NetBackup in a cluster. If this step is not done or does not complete successfully, the bpclusterutil command may be used perform this step. The “bpclusterutil -c” command runs the cluster configuration step performed by installation. Using the -a option with the appropriate option name configures the NetBackup option in a cluster. This command can be run multiple times. Only run this command from the active node. NAME bpclusterutil - Modify and configure NetBackup in a Windows cluster. SYNOPSIS install_path\NetBackup\bin\bpclusterutil -c [-np] install_path\NetBackup\bin\bpclusterutil -a [GDM|NBAR|NDMP] install_path\NetBackup\bin\bpclusterutil -addnode NodeName install_path\NetBackup\bin\bpclusterutil -delete install_path\NetBackup\bin\bpclusterutil -display install_path\NetBackup\bin\bpclusterutil -depend ServiceName DependServiceName install_path\NetBackup\bin\bpclusterutil [-r NETBACKUPTYPE MASTER|MEDIA -r CLUSTERTYPE MSCS|VCS -r CLUSTERGROUPNAME -r CLUSTERNAME -r NETWORKSELECTION -r VIRTUALSERVERNAME -r NETBACKUPINSTALLPATH
89
-r NETBACKUPSHAREDDRIVE -r VIRTUALIPADDRESS -r VIRTUALSUBNETMASK -r NODENAMES -r GLOBDBHOST -r CLUSTEREDSERVICES -r MONITOREDSERVICES ] DESCRIPTION This command is available on NetBackup master and media servers and is only applicable to clusters in a Windows environment. This command will allow the user to set up the registry entries used for cluster configuration and then to configure the cluster as well as configuring NetBackup add on options. OPTIONS -a [GDM|NBAR|NDMP] Configures the specified NetBackup add-on in a cluster. -addnode Adds a node name to the possible owners list of all the resources in the NetBackup Group. -c Configures NetBackup in a cluster based on the values specified in the registry -delete Deletes the registry entries used to configure NetBackup in a cluster. -depend ServiceName DependServiceName Sets the cluster resource for ServiceName to have a dependency on DependServiceName. -display Displays the registry entries used for clustering configuration. -r NETBACKUPTYPE MASTER|MEDIA CLUSTERTYPE MSCS|VCS CLUSTERGROUPNAME CLUSTERNAME NETWORKSELECTION
90
NetBackup High Availability System Administrator’s Guide
VIRTUALSERVERNAME NETBACKUPINSTALLPATH NETBACKUPSHAREDDRIVE VIRTUALIPADDRESS VIRTUALSUBNETMASK NODENAMES GLOBDBHOST CLUSTEREDSERVICES MONITOREDSERVICES Sets the corresponding NetBackup cluster configuration registry entry to the specified value. If NODENAMES is specified, only one node name can be entered. Execute the command multiple times to add node names. If CLUSTEREDSERVICES is specified, only one service name can be entered. Execute the command multiple times to add service names. If MONITOREDSERVICES is specified, only one service name can be entered. Execute the command multiple timesto add service names.
Appendix B, bpclusterutil
91
92
NetBackup High Availability System Administrator’s Guide
Index
A accessibility features xv active/active application, NetBackup as a, in a cluster 4 Advanced Reporter, supported in a cluster 3 altnames directory 76, 77 assigned drives 5 B backups and failovers 79 in a Microsoft Cluster Server (MSCS) 11, 74 in a TruCluster 30 of local disks in a cluster 74 of shared disks 75 what to back up in a cluster 74 bpclusterutil 10, 89–91 C caa_profile 24 caa_start 20 caa_stop 20 failures 24 Checkpoint Restart 79 CLIENT_NAME 60, 64 Cluster Application Availability (CAA), using with NetBackup 20 cluster configurations example 5 mixed server types not supported 2 NetBackup client 1 NetBackup master server 2 NetBackup media server 1 cluster quorum, restoring 12–14 cluster solutions HP Tru64 UNIX TruCluster 2 limitations 4 Microsoft Cluster Server 2 D database agents, supported in a cluster 3 databases protected in a cluster 1 transaction support in 5 device configuration adding new devices 67 attaching devices to all nodes 64 automatic discovery 65 Device host 65 guidelines 64–65 instructions 65 instructions for MSCS 66 instructions for Sun Cluster 66 instructions for TruCluster 67 instructions for VCS 66 specifying cluster alias or virtual server name 64 standalone volume database host 65 virtual server name 65 93 Sun Cluster 3 supported environments 2 VERITAS Cluster Server on UNIX 3 VERITAS Cluster Server on Windows 2000/2003 3 CLUSTER_NAME 47, 59, 64 clustering with NetBackup, overview 1 compatibility lists 4 configuration of NetBackup verifying 69 configuration, of NetBackup 64 examples 83 NetBackup catalog backups 64 policies 64 server list 64 storage units 64 usage of cluster aliases 64 usage of node names 64
volume database host 65 device host 65, 66 disaster recovery, Microsoft Cluster Server (MSCS) 15–17 drives, assigned 5 F failover and manually refreshing Administration Console 79 and storage unit errors 79 definition 1 when backup jobs running 79 when restore jobs running 79 failover master servers 2 failover media servers 1, 2 G GDM, supported in a cluster 3 Glossary. See NetBackup Help. K K77NetBackup 23, 27 KeepAliveTime 80 L limitations media manager in a cluster 4 NetBackup in a cluster 4 local disks backups of 74 loss of connection, to NetBackup Windows server 80 M Microsoft Cluster Administrator, installing on all NetBackup servers 8 Microsoft Cluster Server (MSCS) backup in a 11, 74 bpclusterutil 89–91 device configuration 66 disaster recovery 15–17 hardware requirements 8 installation notes 8 installation prerequisites 8 installing a NetBackup failover server 9 installing non-failover media servers 10 installing the NetBackup client 11 loss of connection to NetBackup server 80 monitoring NetBackup options 71 N
Move Group command 81 NetBackup agents, installing on a failover server 9 NetBackup cluster group name 9 NetBackup Device Manager 81 NetBackup failover media server 9 NetBackup options, installing on a failover server 9 Possible Owners list 8 restores in a 11, 76 upgrading a NetBackup failover server 10 MM_HOST_NAME 25 MM_SERVER_NAME 25 mount/dismount activity 5 NDMP, supported in a cluster 3 NetBackup Catalog Backup entries, verifying UNIX servers 68–69 Windows servers 68 NetBackup Device Manager 81 NetBackup failover server installing in a MSCS 8, 9 uninstalling from a UNIX cluster See NetBackup Installation Guide for UNIX uninstalling from a Windows cluster See NetBackup Installation Guide for Windows upgrading in a MSCS 10 NetBackup options monitoring on a MSCS cluster 71 monitoring on UNIX clusters 71 supported in a cluster 3 non-failover servers, converting to failover servers 4 P patches apply to a TruCluster failover server 25 applying to a TruCluster non-failover server 29–?? patches, applying to a TruCluster non-failover server ??–29 platforms supported for NetBackup in a cluster 4 R Required Network Interface, for a client in a Microsoft Cluster Server (MSCS) 11
94
NetBackup High Availability System Administrator’s Guide
restores and failovers 79 data in a cluster 76 in a Microsoft Cluster Server (MSCS) 11, 76 in a TruCluster 76 performing in a TruCluster 30 robotic daemons, monitoring on UNIX clusters 70 robotic libraries 65 S S77NetBackup 23, 27 Scheduled Backup Attempts, in a TruCluster 24 SCSI Reservations 5 scstat 61 scswitch 56, 57, 58, 59, 61 server types, mixing 4 shared disks and storage units 75 SSO drive allocation resiliency 5 supported in a cluster 3 Standalone Volume Database Host 65, 66, 67 storage units and NetBackup failover media servers 75 and NetBackup non-failover media servers 75 virtual server as NetBackup media server 65 Sun Cluster backups in a 61, 74 cluster_config 55 device configuration 66 hardware requirements 54 installation and configuration 55–57 installation prerequisites 54 monitoring NetBackup options 57, 71 monitoring robotic daemons 57, 70 NetBackup agents, installing or upgrading 57 NetBackup options, installing or upgrading 57 T
NetBackup resource group bringing offline 61 bringing online 61 verifying it is online 61 restores 61, 76 rsh equivalency 54 shared disk configuration 54 upgrading a NetBackup 4.5FP3 failover server 59–61 upgrading a NetBackup failover server 58 tpconfig 67 transaction support in database operations 5 TruCluster 30 backups in a 30, 74 caa_profile 24 caa_stop failures 24 cluster_config 22–23 device configuration 67 device paths 65 failover server configuration 22–24 hardware requirements 21 installation notes 20 installation prerequisites 21 installing a NetBackup failover server 20–22 installing a NetBackup non-failover server 20–21 installing the NetBackup client 30 monitoring daemons for a non-failover server 27 monitoring NetBackup options 26, 71 monitoring robotic daemons 26, 70 NetBackup agents installing on a failover server 23, 26 installing on a non-failover server 28 NetBackup options installing on a failover server 23, 26 installing on a non-failover server 28 non-failover server configuration 27–28 registering NetBackup 28 restores in a 30, 76 Scheduled Backup Attempts 24 setting up a virtual name 21 upgrading a failover media server 25–26 upgrading a failover server 25 Wakeup Interval setting 24
Index
95
U uninstalling NetBackup server from a UNIX cluster See NetBackup Installation Guide for UNIX from a Windows cluster See NetBackup Installation Guide for Windows V Vault, supported in a cluster 3 vcs_agent_upgrade 49 VERITAS Cluster Server (VCS) for UNIX backups in a 51, 74 Clean 34 cluster_config 46, 47 detecting an application failure 34 hardware requirements 35 installation notes 44 installing a non-failover server 51 installing and configuring a NetBackup failover server 44–45 installing the NetBackup client only 51 Monitor 34 monitoring NetBackup options 34, 71 monitoring robotic daemons 70 monitoring robots 34 NetBackup agents, installing 45 NetBackup options, installing 45 NetBackup options, monitoring 45 non-failover server, converting to a failover server 51 Offline 34 Online 34 pre-installation checklist 36 AIX 39–41 HP-UX 42–43 Solaris and Linux 37–38 restores 76 restores in a 51 robotic daemons, monitoring 45
rsh equivalency 35 shared disk configuration 35 supported NetBackup agent for 46, 47 upgrading a failover server 46, 47–50 upgrading a non-failover server 51 virtual name for NetBackup 35 See also the VERITAS Cluster Server User’s Guide VERITAS Cluster Server (VCS) for Windows 2000/2003 agent for NetBackup 32 backups in a 74 configuration of NetBackup See the VERITAS Cluster Server Administrator’s Guide device configuration 66 installation of NetBackup See the VERITAS Cluster Server Administrator’s Guide loss of connection, to NetBackup Windows server 80 NetBackup failover media server 44, 55 requirements 32 restores 76 See also the VERITAS Cluster Server Installation Guide See also the VERITAS Cluster Server System Administrator’s Guide virtual server name 1, 2, 65 and backups 75 as the device host 66, 67 in storage units 65 in the Clients list 75 volume database host 65 W Wakeup Interval setting in a TruCluster 24 Wakeup Interval setting, in a TruCluster 24
96
NetBackup High Availability System Administrator’s Guide