HP P9000 Replication Manager Software Application Agent CLI User Guide Part number: TB584-96073 Eighth edition: May 2013
Legal and notice information © Copyright 2011, 2013 Hewlett-Packard Development Company, L.P. Confidential computer software. Valid license from HP required for possession, use or copying. Consistent with FAR 12.211 and 12.212, Commercial Computer Software, Computer Software Documentation, and Technical Data for Commercial Items are licensed to the U.S. Government under vendor's standard commercial license. The information contained herein is subject to change without notice.
Contents Preface .............................................................................................. 25 1 Overview of Application Agent .......................................................... 27 Application Agent features ........................................................................................................ Application Agent functions ...................................................................................................... Backing up data ........................
CCR configurations ............................................................................................................. Configuration in which only the active server is used ........................................................ Configuration in which only the standby server is used ...................................................... Configuration in which different backup servers are used for the active and standby servers ......................................................................
Setting communication timeout periods for the Protection Manager service ............................... 114 Setting up Application Agent for linkage with RAID Manager ....................................................... 115 Specifying instance numbers .............................................................................................. 115 Specifying a retry count and interval for pair status retry checking ..........................................
Determination of a mount point .................................................................................... Determination of a mount target ................................................................................... Preparations required before executing extended commands ........................................................ Checking the prerequisite conditions ................................................................................... Execution permissions for extended commands .
Precautions on operations ........................................................................................................ Precautions on operations ................................................................................................. Precautions on backup and restoration ................................................................................ Precautions on backup ......................................................................................................
Backing up a file system to a remote site ............................................................................. Operations at the local site ......................................................................................... Operations at the remote site ...................................................................................... Operations at the local site .........................................................................................
Resynchronizing copy groups ...................................................................................... Backing up an SQL Server database ............................................................................ Restoring an SQL Server database from tape ....................................................................... Restoring an SQL Server database ...............................................................................
Preparations for performing backup or restoration in a cascade configuration or multi-target configuration ................................................................................................................... 351 Creating a user script to perform a backup in a cascade configuration .................................... 355 Outline of user script files ............................................................................................ 355 Coding rules for user script files .............
Backing up and restoring a storage group ................................................................................. Precautions when backing up an Exchange database ........................................................... Precautions when restoring an Exchange database ............................................................... Precautions when restoring from a tape ......................................................................... Note on performing a recovery using roll forward ......
Performing a backup on the standby server ................................................................... 454 Restoring data when using the automatic seed function ......................................................... 458 Restoring backup data acquired on the active server (when the automatic seed function is used) ........................................................................................................................
Restoring key files (.K01 or .Knn) ........................................................................................ Restoring missing data files (.DRF or .DAT) ........................................................................... When data file (.DRF or .DAT) backups exist ................................................................. When data file (.DRF or .DAT) backups do not exist ....................................................... Restoring incorrect data files (.DRF or .DAT) .....
Glossary .......................................................................................... 559 Index ...............................................................................................
Figures 1 Overview of Application Agent ................................................................................. 28 2 Logical volume consisting of multiple copy groups ...................................................... 30 3 Data management at a remote site ........................................................................... 32 4 Backing up multiple generations ..............................................................................
29 Example of a CCR configuration (active and standby servers use the same backup server) ................................................................................................................... 61 30 Example of an SCR configuration .............................................................................. 62 31 Configuration in which an active mailbox database copy Is backed up and restored ....... 64 32 Configuration in which a passive mailbox database copy Is backed up and restored .......
55 Cluster group definition (when a backup target and shared disks are in the same cluster group) ................................................................................................................. 136 56 Cluster Group Definition (When a Backup Target and Shared Disks are in Different Cluster Groups) ...............................................................................................................
86 Procedure for restoring an SQL Server database to the primary volume ........................ 297 87 Procedure for restoring an SQL Server database from its backup tape .......................... 298 88 System configuration for backup of an SQL Server database to tape ............................ 305 89 Processing flow ..................................................................................................... 306 90 System configuration for operations using the SQL Server transaction log ......
115 System configuration for backing up and restoring data between volumes by using VSS (when an Exchange database is backed up) ............................................................. 406 116 Processing flow for backing up storage groups to a secondary volume ......................... 407 117 Processing flow for restoring storage groups on the primary volume ............................. 407 118 System configuration for backing up and restoring storage groups to and from tape ......
Tables 1 Types of data subject to backup in the SQL Server database ......................................... 51 2 Comparison of DAG configurations ........................................................................... 68 3 Types of data subject to backup in the Exchange database ........................................... 69 4 Approximate amount required to verify Exchange Server 2003 database for backup by using VSS .....................................................................................
25 Parameters in the RAID Manager-linkage definition file (DEFAULT.dat) (to specify the retry count and retry interval for checking the status of a pair) ............................................ 118 26 Values in the RETRY_WAIT parameter set for each storage system ................................ 118 27 Parameters in the RAID Manager-linkage definition file (DEFAULT.dat) (to specify the retry count and interval for checking the pair status according to the purpose) .....................
52 How to determine a mount point when the command for mounting a secondary volume is executed (when a copy group is specified for an option) ......................................... 178 53 How to determine a mount target when the command for mounting a secondary volume is executed ........................................................................................................... 179 54 Extended command execution preparation tasks needed for each kind of server ...........
81 Correspondence between extended commands and basic commands (SQL server database) ............................................................................................................ 222 82 Correspondence between extended commands and basic commands (Exchange database) ............................................................................................................ 223 83 Required permissions for each command (commands used when backing up and restoring a file system) .................
111 Items output to the detailed trace log ....................................................................... 517 112 Sources of messages (controlled objects) and corresponding message IDs .................... 517 113 Commands for displaying dictionary map file information .......................................... 522 114 Commands for displaying backup information .......................................................... 522 115 Product reference conventions ..........................................
Preface This guide describes how to manage HP P9000 Replication Manager Software.
Preface
1 Overview of Application Agent Application Agent is designed to reduce the workload on system administrators and to provide efficient and reliable data protection. This chapter describes the features and functions of Application Agent. The descriptions in this chapter apply to the Application Agent CLI. Application Agent features Application Agent is a software package that simplifies data maintenance operations, especially those dealing with backup and restoration.
Figure 1 Overview of Application Agent The following describes how Application Agent controls other software components. The numbers below correspond to those in the figure above: 1. Application Agent quiesces (pauses) file systems or databases in storage systems in order to ensure reliable backup operations. 2. By using RAID Manager, Application Agent backs up or restores volumes that have already been operated on using the functions built into the storage systems. 3.
Restoring data Application Agent provides various commands that use volume replication functions to restore a database or file system from a backup. You can use whichever restore command most suits your specific needs. Having two options for restoring data is useful because it decreases the amount of expertise required to restore file systems or databases managed by DBMSs.
Pair management using copy groups In Application Agent, a primary and secondary volume pair whose synchronization is started and stopped by using volume replication or RAID Manager functions is referred to as a copy group (or pair volume). The following figure shows a logical volume that consists of multiple copy groups.
In a cluster environment using Microsoft Cluster Service or Windows Server Failover Clustering, you can select whether to perform data restoration while cluster resources that monitor the DBMS are in the online state or in the offline state. For a restore operation in which the cluster resources are in the offline state, you need to take all the databases that are using the DBMS services offline.
Figure 3 Data management at a remote site Managing data at a remote site allows you have a secondary volume in a storage system that is physically different from the storage system where the primary volume resides. Application Agent uses the Continuous Access Software or Continuous Access Journal copy type to back up or restore volumes to or from a remote site.
When selecting a secondary volume for a backup, you can either manually select the volume you want to use or have Application Agent choose it for you. In order to have Application Agent choose a secondary volume for you, you need to use the volume copy function. When manually selecting a volume, specify the generation name unique to the desired volume.
Figure 5 Backing up multiple generations when a copy group is locked The order in which secondary volumes are used for backups does not change regardless of which generation you might decide to restore at some point in time. The number of generations that Application Agent can manage, and the backup and restore operations available are all dependent on the volume replication functions of the particular storage system.
2 Application Agent system configurations This chapter describes the system configurations and volume configurations required for using Application Agent as well as provides precautions applicable to those configurations. The descriptions in this chapter apply to the Application Agent CLI. Basic configurations for backing up and restoring data This section describes the basic system configurations and provides precautions applicable to using Application Agent to back up and restore data.
Figure 6 Example of a system configured to back up and restore data within a storage system Configurations in which data can be backed up and restored between two storage systems Application Agent supports the use of Continuous Access Software and Continuous Access Journal for backing up and restoring data on the storage systems of remote sites, as a safeguard in the event of storage system failure at a local site.
Figure 7 Continuous Access Software or Continuous Access Journal system configuration Business Copy and Continuous Access Software multi-target configuration Application Agent supports a multi-target configuration for Business Copy and Continuous Access Software. The following figure shows the Business Copy and Continuous Access Software multi-target configuration diagram.
Figure 8 Business Copy and Continuous Access Software multi-target configuration Business Copy and Continuous Access Journal multi-target configuration Application Agent supports a multi-target configuration for Business Copy and Continuous Access Journal. The following figure shows the Business Copy and Continuous Access Journal multi-target configuration diagram.
Figure 9 Business Copy and Continuous Access Journal multi-target configuration Continuous Access Software and Continuous Access Journal multi-target configuration Application Agent supports a multi-target configuration for Continuous Access Software and Continuous Access Journal. The following figure shows the Continuous Access Software and Continuous Access Journal multi-target configuration diagram.
Figure 10 Continuous Access Software and Continuous Access Journal multi-target configuration Business Copy (multi-generation), Continuous Access Software, or Continuous Access Journal configuration Application Agent supports a system configuration for Business Copy (multi-generation), and Continuous Access Software, or Continuous Access Journal. The following figure shows the Business Copy (multi-generation) and Continuous Access Software, or Continuous Access Journal configuration diagram.
Figure 11 Business Copy (multi-generation), Continuous Access Software, or Continuous Access Journal configuration Precautions on backing up and restoring data between storage systems • Regarding pair creation of paired volumes between storage system Application Agent does not automatically create paired volumes when it backs up data between a local site and a remote site. Users must create a pair of paired volumes in advance.
Figure 12 Example of configuration for concurrent backup from multiple primary volumes to multiple secondary volumes (remote copy) • Regarding start of RAID Manager instances between storage systems RAID Manager instances between a local site and a remote site are not automatically started. Start the RAID Manager instances on the remote site in advance.
Figure 13 Example of a system configured with a tape device Configurations for operation in cluster environments This section describes configurations for operation when a database server and a file server are used in a cluster environment.
Figure 14 Example of a cluster configuration of the operation standby type (Active-Passive) To execute Application Agent commands in a cluster configuration of the operation standby type (Active-Passive), you need to assign a storage destination for the dictionary map files used by each virtual server. For about the dictionary map files in a cluster configuration, see “Creation in an Active-Passive cluster configuration” on page 130.
Figure 15 Example of a cluster configuration of the mutual standby type (Active-Active) To execute Application Agent commands in a cluster configuration of the mutual standby type (Active-Active), create a dictionary map files for each virtual server. Additionally, you need to assign a storage destination for the dictionary map files used by each virtual server. For details about the dictionary map files in a cluster configuration, see “Creation in an Active-Active cluster configuration” on page 132.
Figure 16 Example of a system configured to use VSS to back up data Requirements and precautions for backing up file systems This subsection explains requirements and precautions when using VSS to back up file systems. • P9000 RMXP Shadow Copy Provider must be installed both in the file server and backup server in advance.
• Do not mount another volume in a directory on the volume to be backed up. If another volume is mounted, mounting and restoration for the secondary volume might fail. • If you change the settings of the following files, use the Services pane in the Windows to restart the Protection Manager service: • Application Agent configuration definition file (init.conf) • RAID Manager-linkage definition file (DEFAULT.
Figure 17 Example of a system configured to restore data on a host that differs from the host used for backup Configurations for using the log-shipping functionality You can use the log-shipping function to back up an SQL Server database. When you use this function, the system needs two servers: a source server that distributes log files, and a destination server to which the log files are distributed.
Figure 18 Example of a system configuration for the log-shipping functionality Cascade configurations When the backup is of an SQL Server database, Application Agent can handle the backup to a secondary volume at a remote destination. The backup configuration in this case is called a cascade configuration because backup configuration has multiple hierarchies. For data backup and restoration in a cascade configuration, use user scripts.
Figure 19 Example of a cascade configuration (when the backup is of an SQL Server database) Multi-target configurations When the backup is of an SQL Server database, Application Agent can handle a multi-target configuration that backs up data to a remote destination and to a secondary volume with a local environment at the same time. For data backup and restoration in multi-target configuration, use user scripts.
Figure 20 Example of a multi-target configuration (when the backup is of an SQL Server database) Requirements and precautions for backing up SQL Server databases The table below describes the types of data subject to backup in the SQL Server database. Backed-up databases differ depending on option specifications in the backup command (the drmsqlbackup or EX_DRM_SQL_BACKUP command).
Database subject to backup#1 User database Distribution database Type of file subject to backup Backup file name Storage destination for backup files Data file Same as the backup source file name Secondary volume Transaction log file Same as the backup source file name Secondary volume Meta file#2 Varies depending on the VDI metafile storage directory specified with the -drmsqlinit command#3 Data file Same as the backup source file name Secondary volume Transaction log file Same as the bac
• Do not use the name META_database-ID (ten digits).dmp for a data file of a database or transaction log file. • Backed-up data cannot be restored or recovered by specifying the -undo option during execution of the drmsqlinit command when there is no registered directory for storing the UNDO file. • Backup for the transaction log cannot be executed during execution of the drmsqlinit command when there is no registered directory for storing the transaction log backup file.
database is backed up and restored along with the database configuration file of the master database when the system database is backed up and restored. • For SQL Server 2008, the data compression function is not supported.
Figure 21 Allocation of database configuration files Allocating multiple database configuration files to one volume When multiple database configuration files are allocated to a volume, all databases need to be backed up or restored in a group. The figure below shows how backup and restoration are performed in this case. Figure 22 When multiple database configuration files are allocated to a volume Allocating metafiles A metafile is a file SQL Server outputs during backup processing.
Figure 23 Metafile allocation Assume that you have a cluster system in which database servers are clustered using Microsoft Cluster Service or Windows Server Failover Clustering, and one of the cluster servers has failed over to the other server. In this case, if a metafile has been stored on a volume in the storage system, the primary volume can be restored from the secondary volume.
Figure 24 Metafile allocation (in the cluster configuration) Using a dynamic disk configuration • Do not allocate a data file made up of multiple instances to a file system (drive) of a disk group subject to backup and restoration. • Allocate the following directories and files to a file system (drive) of a disk group different from the one to which data files (*.mdf and *.ndf) are allocated.
Using a cluster configuration • Do not register the service resources of multiple SQL Server instances into a cluster group. • In a cluster environment, specify a user on each node for the owner of a database subject to backup. Local users with the same user name and password on different nodes are not considered to be the same user. Therefore, use a domain user account that is common to all nodes.
• When VSS backup is used with SCC, backups can be performed only at the active server. However, when VSS backup is used with CCR, backups can also be performed at the standby server. • Ability to reduce processing loads on the active server by performing backups on the standby server. In this manual, the function that performs seed processing automatically when restoring data is called the automatic seed function. For details about seed processing, see the Exchange Server 2007 manual.
Figure 27 Example of a CCR configuration (only the standby server is used) Configuration in which different backup servers are used for the active and standby servers The following figure shows an example of a CCR configuration in which different backup servers are used for the active and standby servers.
Figure 29 Example of a CCR configuration (active and standby servers use the same backup server) SCR configurations Application Agent provides VSS backup and restoration functionality for SCR configuration sources. Limitations on the target configuration are the same as for Exchange Server 2007. For details about the limitations for Exchange Server 2007, see the Exchange Server 2007 documentation. The following shows an example of an SCR configuration.
Figure 30 Example of an SCR configuration Application Agent's restore processing provides the SCR seed function that executes seed processing (data resync) in an SCR configuration. The user can set whether the SCR seed function is to be used. The SCR seed function is used in the following cases: • When a point-in-time restore operation is performed. • When data integrity between the source and target is known to have been lost when a roll-forward restore operation was performed.
an active mailbox database copy, you can use the data in a passive mailbox database copy to continue operation. Application Agent provides backup and restore functionalities for mailbox database copies in the event of a logical error occurring due to issues such as incorrect email transmission and viruses. The restore functionality performs seed processing, which resynchronizes data by copying the passive mailbox database copy to the restored active mailbox database copy.
Figure 31 Configuration in which an active mailbox database copy Is backed up and restored Primary volume 1 is the active mailbox database copy, primary volume 2 is the passive mailbox database copy, and the secondary volume is managed by the backup server. The active side manages all the services and backup and restore operations, and the passive side temporarily takes over execution of the active side's services if needed.
Figure 32 Configuration in which a passive mailbox database copy Is backed up and restored Primary volume 1 is the active mailbox database copy, primary volume 2 is the passive mailbox database copy, and the secondary volume is managed by the backup server. The active side manages services, and the passive side takes over execution of the active side's services if needed and also performs backup and restore operations.
Figure 33 Configuration in which an active mailbox database copy and a passive mailbox database copy are backed up and restored by different backup servers Primary volume 1 is the active mailbox database copy, and primary volume 2 is the passive mailbox database copy. Secondary volume 1 is managed by backup server 1, and secondary volume 2 is managed by backup server 2. The data synchronization between the active side and the passive side is accomplished via log shipping.
Backing up and restoring the active mailbox database copy and the passive mailbox database copy by using the same backup server The following figure shows an example of a DAG configuration in which an active mailbox database copy and a passive mailbox database copy are backed up and restored by the same backup server.
If logical errors, such as those caused by viruses, occur on both the active side and the passive side, the data is restored from the tape device to secondary volume 1, and then from secondary volume 1 to primary volume 1. The restored data is then applied to primary volume 2 via seed processing. Comparison of DAG configurations The following table compares the four types of DAG configurations that use Application Agent. Refer to this table when setting up your system.
Requirements and precautions for backing up Exchange databases The following table describes the types of data subject to backup in the Exchange database. Table 3 Types of data subject to backup in the Exchange database Database subject to backup Files subject to backup Data files (Exchange Server Information Store) Exchange Server storage group Exchange Server Information Store Extension *.edb *.stm# Transaction log files *.log Checkpoint files *.chk Data files *.edb Transaction log files *.
• Maximum length: 64 characters • Characters that cannot be used = ; \ / , • The operations shown below change information pertaining to the Exchange Server on the Active Directory, so if you perform any of these operations you will no longer be able to restore backup data obtained before these operations (that is, you will no longer be able to use the drmexgrestore or EX_DRM_EXG_RESTORE command): • Modification of the paths of the Information Store, log, and system files.
• P9000 RMXP Shadow Copy Provider must be installed in advance, on both the database server and the backup server. If P9000 RMXP Shadow Copy Provider has not been installed on those servers when Application Agent is installed, P9000 RMXP Shadow Copy Provider will be installed together with Application Agent. • The backup server is required even if data is not backed up to tape. This is because the backup server imports VS snapshots and verifies the Exchange database.
the verification, we recommend that you use the drmexgverify or EX_DRM_EXG_VERIFY command to verify the database before performing restoration or backing up data to tape. Note that, although an option is provided for the drmexgbackup or EX_DRM_EXG_BACKUP command to delete transaction logs after backup without verification, if this option is used when the database is corrupted, roll-forward recovery might not be able to be performed.
• The following table lists the approximate amount of time required to verify a database for Exchange Server 2010. Table 6 Approximate amount required to verify Exchange Server 2010 database for backup by using VSS Backup server system configuration Data file capacity (*.edb) Number of transaction log files (*.log) Amount of time required for database verification 100 GB 5 30 minutes • CPU: 3.
The table below lists the conditions for allocating files when restoring backup data in units of Information Stores. Table 7 Conditions for allocating files when restoring backup data in units of Information Stores Files on the same disk Restoration target No other files *.edb Information Store that is a restoration target Information Stores in the same storage group Information Stores in different storage groups *.edb *.stm *.edb *.edb -- I A A N N I -- A A N N *.stm *.stm I *.
• Restore only IS12 individually • Restore IS13 and IS14 at the same time (you cannot restore just one) Notes: • When restoring backup data in units of Information Stores, you must apply (roll forward) the transaction log at the time of restoration. For details on precautions relating to executing commands at restoration, see the description for the drmexgrestore command in the manual HP P9000 Replication Manager Software Application Agent CLI Reference Guide.
• When using VSS, the following conditions must be satisfied regarding the relationship between the cluster resource and the Exchange Information Store resource to be operated: For Exchange Server 2003: Except for the relationship between the cluster resource and the Exchange System Attendant resource to be operated, do not set the Exchange Information Store resource to go offline when a cluster resource goes offline.
CCR configuration • With a CCR configuration, you must manage dictionary map files and backup catalogs separately for the active and standby servers. The dictionary map files and backup catalogs acquired by each of the servers are used by the respective database servers and by the backup servers used by these database servers. • If the active and standby servers use the same backup server, the active and standby servers must use different copy group names for the disk to which the backups are saved.
• The server used for seed processing has sufficient performance (such as CPU and disk performance) to adequately run the installed applications (such as Application Agent, Exchange Server, and tape backup management softwares). SCR configuration • If you use the SCR seed function, the same version of Application Agent as the source must be installed on all targets.
• To restore the backup data, the mailbox database copy to be restored must be in the active status. If it is not in the active status, a message indicating an error is displayed. • Backup data can be restored only by the Exchange Server that backed up the data. An Exchange Server cannot restore backup data that was obtained by a different Exchange Server. • Correct execution of the replication function after restoration requires seed processing.
Figure 36 Configuration for managing the primary volume and secondary volume from the same server (without using Application Agent) • A copy group must consist of pair volumes for the same application. When a copy group consists of more than one pair volume and there are pair volumes for different applications in the copy group, an unexpected backup or restoration might be performed. • When using Business Copy, configure the pair volume in the storage system.
drmXXdisplay indicates the drmfsdisplay, drmsqldisplay, or drmexgdisplay command. • When using Continuous Access Journal, define h0 (h and a generation number) for the MU# setting in the RAID Manager configuration definition file (horcmn.conf). If 0 is defined for MU#, the Continuous Access Journal information will not be stored in the dictionary map when you execute a drmXXdisplay command with the -refresh option specified.
Figure 37 System configuration (when the primary volume for Business Copy and the primary volume for Continuous Access Software are the same LDEV) When using the configuration shown in Figure 37 to restore the paired volume for Business Copy by using Application Agent, set the status of the paired volume for Continuous Access Software to SMPL or PSUS (SSUS).
Figure 38 System configuration (when the secondary volume for Business Copy and the primary volume for Continuous Access Software are the Same LDEV) When using the configuration shown in Figure 38 to back up or restore the paired volume for Business Copy using Application Agent, set the status of the paired volume for Continuous Access Software to SMPL or PSUS (SSUS).
Figure 39 System configuration (In a Business Copy cascade configuration, when the volume subject to backup operates both as the primary volume and the secondary volume) In a Business Copy cascade configuration, Application Agent only supports backup and restore operations for the pair volumes of the primary volumes used in the database or the file system.
If not using VSS: It is possible to define consistency groups to make files of the same file types of multiple storage groups that are always restored together are in a same consistency group. However, if rolling forward is not performed when restore and recovery operations are performed, it is possible to define consistency groups to make all files of the storage groups that are always restored together are in a same consistency group.
Figure 40 In XP Continuous Access Asynchronous/Continuous Access Journal configuration: restoring UserDB1 and UserDB2 together or separately Figure 41 In XP Continuous Access Asynchronous/Continuous Access Journal configuration: restoring UserDB1 and UserDB2 together • For information about precautions to take when changing a volume configuration, see “Precautions when changing a volume configuration” on page 92.
CMD-IP-address-PORT-number), take measures based on the confirmed information, and then start Application Agent CLI operation once again. Volume configurations usable by Application Agent The table below describes volume configurations that Application Agent can back up and restore.
Figure 42 Examples of a disk set configuration and of a relationship between a disk group and a disk set In this manual, unless it is necessary to distinguish between Veritas Storage Foundation and LVM disk groups, both groups are generically referred to as a disk group. Application Agent can also simultaneously back up a basic disk and a disk group. When a dynamic disk is being used, Application Agent supports both of the following configurations: • One disk group is used as one logical volume.
Examples of disk group configurations usable by Application Agent If you configure a disk group with logical volume manager, Application Agent can be used only when one or more physical volumes constitute one disk group. You cannot divide one physical group into multiple disk groups. The following figure shows examples of disk group configurations usable by Application Agent.
Figure 44 Data backup and restoration in a system configuration where multiple disk groups exist in a file system In this figure, disk groups DG01 and DG02 each of which consists of two physical volumes are configured. Three logical volumes are created on DG01 and one logical volume is created on DG02. Mount points D:, E:, and F:, are assigned to DG01 and mount point G: assigned to DG02.
Example: Restore only the file system created in disk group DG02 after backing up disk groups DG01 and DG02. PROMPT> drmfsrestore 0000000001 -resync -target G: KAVX0001-I The drmfsrestore command will now start. KAVX0019-I Copying file(s) now... KAVX0002-I The drmfsrestore command will now end. PROMPT> Requirements and precautions for volume configurations This section describes the requirements and precautions for volume configurations when Application Agent is used.
• If one disk group contains multiple physical volumes, Application Agent performs volume duplication operation for backup only by the number of physical volumes. • Do not define the following as copy groups, and do not specify them as backup targets: directories at a directory mount point, or volumes mounted to a drive. Also, configure directory mount points that are backup targets so that they are on the same directory level.
• • • • A disk group was deleted or disabled. A logical volume allocated to a disk group was added or deleted. A physical volume belonging to a disk group is added, deleted, or replaced. A change was made in the RAID Manager configuration definition file (horcmn.conf), such as the adding or deleting of a pair volume.
Backing up and restoring dynamic disks by dynamically or statically recognizing secondary volumes Dynamically recognizing secondary volumes: • When Application Agent dynamically recognizes the secondary volume, the Demount Device Warning message is output to the Windows event log. Statically recognizing secondary volumes: • Do not rescan disk information with disk management, Veritas Enterprise Administrator, or the diskpart or vxassist command.
formed. To view the backup signatures that were created when the backup was performed, execute the drmdevctl command by specifying the backup ID and the -sigview option.
Application Agent system configurations
3 Preparation for using Application Agent This chapter explains the flow of environment setup required for using Application Agent, and gives details on each step in the environment setup. The descriptions in this chapter apply to the Application Agent CLI. Setting up an environment for Application Agent Table 4-1 shows the procedures for setting up an environment for Application Agent and the tasks required for each server.
Environment setup task File server or database server Backup server File name Sections and documents for reference Specify settings for linkage with RAID Manager. Needed Needed RAID Manager-linkage definition file (DEFAULT.dat) Setting up Application Agent for linkage with RAID Manager, page 115 Create dictionary map files. Needed Needed -- Creating dictionary map files, page 129 Cluster information configuration definition file (cluster-name.
Environment setup task File server or database server Backup server File name Sections and documents for reference Specify settings for batch operations. Needed Needed Any file Creating a list file, page 159 Update the dictionary map files. Needed Needed#4 -- Updating a dictionary map file, page 163 -- Preparations required before executing extended commands, page 179 Perform preparation for executing extended commands. Needed Needed Create a user script.
When a primary volume exists in the backup server, the dictionary map files must be updated. When the primary volume does not exist in the backup server, updating of the dictionary map files is not necessary. Setting up RAID Manager To back up data by using Application Agent, you need to create a RAID Manager configuration definition file (horcmn.conf, when n is the instance number) for each instance, to match the intended usage. The location of the RAID Manager configuration definition file (horcmn.
• • • • Back up multiple primary volumes to a single generation Back up multiple primary volumes to multiple generations Back up or restore between storage systems Set up a multi-target or cascade configuration Backing up one primary volume to multiple generations This subsection describes the configuration in which one primary volume is to be backed up to multiple generations.
Figure 48 Volume configuration and configuration definition to back up multiple primary volumes to a single generation Backing up multiple primary volumes to multiple generations This subsection describes the configuration in which multiple primary volumes are to be backed up to multiple generations.
Figure 49 Volume configuration and configuration definition to back up multiple primary volumes to multiple generations Backup or restore between storage systems Application Agent allows you to backup or restore with mixed volume replication functionalities between storage systems such as Business Copy and Continuous Access Software or Continuous Access Journal. In this case, setup to enable a remote copy is required in the RAID Manager configuration definition file (horcmn.conf).
Figure 50 Example of a configuration definition for backup or restore between storage systems The following shows an example of the RAID Manager configuration definition file for a database server Host0 in the local site in the above configuration example. In this example, multiple secondary volumes are allocated to one primary volume, and a separate configuration definition file is created for each backup target server. • RAID Manager configuration definition file (horcm0.
• RAID Manager configuration definition file (horcm1.conf) HORCM_DEV ##dev_group TC01 TC01 HORCM_INST ##dev_group TC01 dev_name DEV31 DEV32 port# CL2-C CL2-C ip_address Host2 TargetID 8 8 LU# 6 7 MU# LU# 6 7 MU# h1 h1 LU# 6 7 MU# h2 h2 service 12502/udp • RAID Manager configuration definition file (horcm2.
than one volume. You need to take this into consideration when specifying the RAID Manager configuration definition file. The following shows an example of coding in a configuration definition file for a multi-target configuration and cascade configuration. Figure 51 Configuration definition for a multi-target configuration and a cascade configuration Setting up operation of Application Agent Application Agent operation settings are specified in the Application Agent configuration definition file (init.
checking the status of the cluster resource, change the following parameters. In a non-cluster environment, these parameter must not be changed. Leave the default settings as is. Table 10 Parameters in init.conf (to set the retry count and retry interval for checking the status of cluster resources) Parameter Specifiable values (default) Description • In this parameter, specify the maximum retry count for checking the status of cluster resources. Generally, specify a value other than 0.
Specifying the retry count and interval for process status retry check When Application Agent performs a cold backup or restoration, it activates or deactivates the service of the DBMS product to be backed up or restored. When Application Agent activates or deactivates the service, it checks the active/inactive status of the process. You can change the retry count and retry interval for checking the status of a process by using the following parameters.
Specifying the retry count and interval for command execution status retry check If you want to change the retry count and retry interval for retries of checking the status of the command being executed by Application Agent, change the following parameters. Table 13 Parameters in init.conf (to specifying a retry count and a retry interval for command execution status retries checking) Parameter Specifiable values (default) Description • Specify the maximum retry count.
• Make sure that the previous backup information has been deleted. Use the drmsqlcat command to confirm that the backup information has been deleted. To delete the backup information, use the drmresync command. • Make sure that the status of the copy group for which operations are to be performed is PVOL_PAIR. Use the RAID Manager pairdisplay command to confirm the status of the copy group.
Even after execution of the drmsqlrecover command, you can check integrity from SQL Server Enterprise Manager. If YES is specified for this option, integrity after recovery will be checked using SQL Server's DBCC CHECKDB. Note also that the DBCC CHECKDB execution time heavily depends on the database size, including the number of indexes in the database and the number of lines per table. For further information, see the SQL Server online help.
Cluster type CLU_MSCS_RESTORE Veritas Cluster Server -- Restore target Yes Cluster resource status OFFLINE Legend: --: Not applicable Yes: Can be restored • SQL Server Database You can restore a user database while cluster resources are in the online state in a Microsoft Cluster Service or Windows Server Failover Clustering environment. If the restoration target contains a system database (master, model, msdb, or distribution), the SQL Server service is stopped momentarily.
If you specify -no_resync as the restoration option: Table 20 Parameter in CLU_MSCS_RESTORE and cluster resource Status (SQL Server database 2) Cluster type Microsoft Cluster Service or Windows Server Failover Clustering Veritas Cluster Server CLU_MSCS_RESTORE Restoration target S U Cluster resource status (cluster control) S+U OFFLINE Yes Yes Yes None# ONLINE Yes Yes Yes None# -- Yes Yes Yes None# Legend: --: Not applicable Yes: Can be restored S: System database U: User database # If
Setting up parallel execution of verification processes for Exchange database backup When an Exchange database is being backed up, you can specify the number of verification processes that are executed in parallel during VSS backup, using the following parameter. Table 22 Parameter in init.conf (setup of parallel execution of verification processes) Parameter Default (specifiable value) Description Specify the number of verification processes to be executed in parallel during VSS backup.
Setting up Application Agent for linkage with RAID Manager Do not set HORCMINST or HORCC_MRCF in the system environment variable. Also, do not set HORCM_CONF or HORCMPERM in the system environment variable or the user environment variable. For the RAID Manager settings, add or change the parameters in the RAID Manager-linkage definition file (DEFAULT.dat).
HORCMINST_AUX=1 HORCMINST=2 HORCMINST_AUX=3 • In a cluster configuration, all servers making up the cluster must share the HORCMINST and HORCMINST_AUX instance numbers in DEFAULT.dat. You can create two identical DEFAULT.dat files, and then place one DEFAULT.dat file on one server and the other DEFAULT.dat file on the other server. Table 24 Parameters in the RAID Manager-linkage definition file (DEFAULT.
Figure 52 Example of specifying a RAID Manager instance number in the RAID Manager-linkage definition file (DEFAULT.dat) The following conditions are assumed for this example: • A RAID Manager instance (instance number: 1) has already been activated on the file/database server. • A RAID Manager instance (instance number: 2) has already been activated on the backup server. • RAID Manager configuration definition files (horcm1.conf and horcm2.
The pair status checking retry interval should be 1.5 times the estimated pair operation time in a backup system configuration. By default, Application Agent checks the pair status every 10 seconds, up to 3600 times. This default setting assumes that a backup or restoration will finish in 10 hours (10 seconds * 3600 = 600 minutes).
specify any purpose-specific parameters, the values of RETRY_TIME and RETRY_WAIT are used. The following table shows the parameters you can add. Table 27 Parameters in the RAID Manager-linkage definition file (DEFAULT.dat) (to specify the retry count and interval for checking the pair status according to the purpose) Parameter Specifiable values (default) Description Specifies the maximum number of retries for checking the completion of pair resynchronization.
Parameter Description Specifiable values (default) Specifies the maximum number of times a retry can be performed for checking whether pair splitting has finished. Specify this parameter to check the pair status after pairsplit is executed for backup or restoration. SPLIT_RETRY_TIME If you specify 0, Application Agent continues to retry checking until it succeeds in checking the status.
Parameter Specifiable values (default) Description Specifies the retry interval for checking whether resynchronization on restoration has finished. You can specify a value in units of 10 milliseconds. RESTORE_RETRY_WAIT When adding this parameter, you must consider the following: the number of disks to be restored, the disk sizes, and the difference in data amount between the primary and secondary volumes when at the time the restoration is performed#2.
The smaller the retry interval value (RESYNC_RETRY_WAIT, SPLIT_RETRY_WAIT, RESTORE_RETRY_WAIT, BUSY_RETRY_WAIT, or RESTORE_DELAY_RETRY_WAIT), the larger the CPU load. Refer to Table 26 and specify a value appropriate to the storage system to be used. Specifying a retry count and interval for RAID Manager command busy retry checking If a RAID Manager command returns a busy status, Application Agent tries to execute the command again.
• Write parameters in the form parameter-name=parameter-value. • Use an equals character (=) between a parameter name and a parameter value. • Specify a parameter value within the valid range. The following values can be specified in copy parameter definition files. Table 29 Parameter in copy parameter definition files Parameter Description Specifiable values RETRY_TIME In this parameter, specify the maximum retry count for checking the status of a pair.
Parameter Description Specifiable values RESTORE_DELAY_RETRY_TIME Specify the maximum number of retries that will be performed to confirm whether the volume is ready for resynchronization. You need to specify this parameter if you want to check whether the pair status allows execution of pairresync -restore for restoration. If you specify 0, Application Agent continues to retry checking until it can check the pair status.
For details about the values that can be specified as the track size, see the RAID Manager documentation. Table 30 Parameter in the RAID Manager-linkage definition file (DEFAULT.
• drmfsbackup • drmsqlbackup • drmexgbackup • drmresync Table 31 Parameter in the RAID Manager-linkage definition file (DEFAULT.dat) (to specify whether to create a pair automatically) Parameter Description Specifiable values (default) Specifies whether Application Agent automatically creates paired volumes. PAIR_CREATE# If DISABLE is specified, Application Agent will not automatically create a pair.
Setting up operation mode when automatically selecting copy groups Usually, Application Agent automatically selects the secondary volume used as a backup destination. However, when Application Agent is copying data to multiple secondary volumes by locking the copy group, one generation might be given different generation numbers.
Specifying the installation path for RAID Manager Specifying the installation destination for the RAID Manager commands to be started by Application Agent. If RAID Manager is to be installed on a different drive than Application Agent, specify the installation destination for RAID Manager. If RAID Manager is to be installed on the same drive as Application Agent, the installation path for RAID Manager is set automatically when Application Agent is installed.
Creating dictionary map files Application Agent automatically creates dictionary map files during installation. If you want the dictionary map files to be created in a particular directory, or if you are installing Application Agent in a cluster configuration, you must use one of the procedures explained in “Creation in a non-cluster configuration or on a backup server” on page 129 to “Creation in a cascade configuration for Exchange database backup” on page 134 to create the dictionary map files.
Creation in an Active-Passive cluster configuration In an Active-Passive cluster configuration, you need to create dictionary map files in a shared directory that can be referenced from both the active and standby servers. The figure below provides an example of creating dictionary map files in an Active-Passive cluster configuration.
1. Change the locations of the dictionary map files and backup catalog. In the init.conf configuration definition files of Application Agent on the active and standby servers respectively, add the DRM_DB_PATH=directory-on-the-shared-disk;virtual-server-name entry. When editing the files, note the following points: • Add the entry to the last line of the init.conf file. • After directory-on-the-shared-disk;virtual-server-name, always enter a linefeed code.
Creation in an Active-Active cluster configuration In an Active-Active cluster configuration, you need to create dictionary map files on each virtual server. Note also that the created dictionary map files must be placed in a shared directory that can be referenced from both servers, in case a failure should occur in either server. The figure below provides an example of creating dictionary map files in an Active-Active cluster configuration.
1. Set the locations of the dictionary map files and backup catalog. In the init.conf configuration definition files of Application Agent on the active and standby servers respectively, add a DRM_DB_PATH=directory-in-shared-disk;virtual-server-name entry for each virtual server. When editing the files, note the following points: • Add the entry to the last line of the init.conf file. • In directory-in-shared-disk, specify a cluster resource of virtual-server-name.
3. When Veritas Cluster Server is used, execute the drmclusinit command to register cluster resource information. For details, see “Creating a cluster information configuration definition file when using Veritas cluster server” on page 137. Creation in a cascade configuration for Exchange database backup To back up an Exchange database in a cascade configuration, you must create dictionary map files in an arbitrary directory at both the local and remote sites.
• How to create a cluster information configuration definition file (when Veritas Cluster Server is used) A cluster information configuration definition file needs to be created only when Veritas Cluster Server is used as the cluster software. If the cluster software is Microsoft Cluster Service or Windows Server Failover Clustering, you do not need to create a cluster information configuration definition file.
Figure 55 Cluster group definition (when a backup target and shared disks are in the same cluster group) How to Define Cluster Groups (When a Backup Target and Shared Disks Are to Be in Separate Cluster Groups) In a cluster configuration, if you execute multiple extended commands concurrently in a single virtual server, define the following cluster groups separately: a cluster group that contains backup target, and a cluster group that contains shared disks used by Application Agent.
Figure 56 Cluster Group Definition (When a Backup Target and Shared Disks are in Different Cluster Groups) Creating a cluster information configuration definition file when using Veritas cluster server When you are using Veritas Cluster Server, execute the drmclusinit command on the database or file server to register cluster resource information.
2. Register the following items according to the output messages: Table 37 Parameters for the cluster software Entry name Explanation Virtual server name (required) Enter the virtual server name to be managed by the cluster software. Resource group name (required) Enter the service group name. • When the backup is of a file system: Enter a hyphen (-). Name of the instance or application to be backed up (required) • When the backup is of an SQL Server database: Enter the SQL instance name.
• Execute the drmclusinit command on both the active server and standby server. • The resource information to be registered by using the drmclusinit command must exactly match the information registered in the cluster software. • The virtual server name to be registered by using the drmclusinit command must satisfy the following conditions: - The specified virtual server name is the same as VirtualName in the Lanman resource of Veritas Cluster Server.
The information to be registered includes the following: • • • • Storage directory for the VDI metafiles VDI generation timeout period Storage directory for the UNDO log files Storage directory for the backup files (for transaction logs) For details on the SQL Server database information to be registered, see the description of the drmsqlinit command in the manual HP P9000 Replication Manager Software Application Agent CLI Reference Guide.
4. Check the registration. To check your registration, execute the drmsqlinit command with the -v option specified: PROMPT> drmsqlinit -v instance-name Notes • If your system is in a cluster configuration. or if you use Continuous Access Software or Continuous Access Journal functionality, you need to execute the drmsqlinit command on each server. • Execute the drmsqlinit command after setting up the SQL Server instance.
Specify this option if the command that operates the SQL Server database terminates with an error due to an SQL Server login request timeout (error message: KAVX1008-E, detailed message DRM11013, Code: 0). You can use the SQL_LOGIN_TIMEOUT parameter to specify the SQL Server login timeout option in sqlinit_instance-name.conf. A login request timeout is likely to occur in environments with high network loads, which may result from a large-scale database.
Restored object S U System configuration Non-cluster Restoration type SQL Server automatic recovery processing Parameter specification Restoration while the cluster resources are online Not performed Not needed -- Not performed Not needed Restoration while the cluster resources are offline Performed Needed# Restoration while the cluster resources are online Not performed Not needed -- Not performed Not needed Cluster -- Yes (partial) Non-cluster Legend Yes: Can be restored --: Not a
2. Specify the parameter in the sqlinit_instance-name.conf file. Write the parameter starting from the far left, without spaces. Delimit the parameter name and timeout value by an equal sign (=), and do not enter space or tab characters before or after the equal sign. Lines beginning with # are assumed to be comment lines. If a parameter is specified two or more times, the earliest value written is effective. • To specify the SQL Server login timeout option Add the SQL_LOGIN_TIMEOUT=login-timeout-value.
Specifying settings for using VSS If you want to use VSS to perform an online backup of a file system or an Exchange database, the following settings are required: • The Application Agent and RAID Manager environments must be set up so that Application Agent can perform VSS backups. Moreover, it is necessary to create a pair before backup. When creating a pair, specify the -m noread option in the paircreate command.
Application-Agent-installation-directory\DRM\conf Table 42 vsscom.conf parameters Parameter Description Values that can be set BACKUP_SERVER Specifies the host name or IP address of the backup server. 1 to 256 one-byte characters Specifies the number of times backup is retried if a Writer timeout occurs during backup. If 0 is specified, backup is not retried.
Parameter Description Values that can be set EXG_VERIFY_RETRY_INTERVAL Specifies, in seconds, the intervals at which backup is retried when an Exchange database verification error occurs. Usually, this parameter does not need to be changed This value is not required when the backup is of a file system. 0 to 99999 • If you plan to back up Exchange databases, install Exchange Management Console on the backup server. If Exchange Server 2003 is used, also install the ESEUTIL command.
Set the system environment variables for the file server or database server, as shown in the table below. Table 43 System environment variables required to use VSS (file server or database server) Environment variable Description Specify the following value: VSXPRMENVF Application-Agent-installation-destination\DRM\conf\vssprv.conf# # This file is for internal use of Application Agent. You do not need to specify this file. After setting the above system environment variables, restart the OS.
• A backup with a name identifying the generation specified This section describes the system setup for each server. Set the system environment variables for the file server or database server, as shown in the table below. Table 45 System environment variables required to use VSS (file server or database server) System Environment Variable Description VSXPHORCMINST_LOCAL Specifies the instance number of the RAID Manager that manages the primary volume used in VSS backup.
• In a configuration where multiple file servers or database servers are sharing one backup server, manage all the secondary volumes in a single RAID Manager instance on the backup server. Creating a cascade configuration definition file for Exchange database backup To back up an Exchange database in a cascade configuration, you must edit the definition file for the cascade configuration on the database server.
Parameter #1 Description Maximum character length #2 Omission Specify the number of the RAID Manager instance that is managing Remote Copy in the cascade configuration. INSTANCE_NUMBER COPY_GROUP_NUM All of the Remote Copy copy groups must be managed by this instance. If you are using VSS to perform the backup, the instance specified in this parameter must be the same as the instance that is being managed by Application Agent. Enter the number of Remote Copy copy groups in the cascade configuration.
#1 Parameter Description #2 Maximum character length Omission Specify the timeout value for the pair resynchronization processing at the remote site as a value between 0 and 86400 (in seconds). If you specify 0, this processing does not time out. If you omit this specification, 3600 is assumed. PRESPLIT_TIMEOUT Only parameter values in sections that are considered to be for the remote site are used. As a rule, set this value to roughly 1.
The following table provides details about the handling of the backup command's -cascade option and sections. Table 48 How the -cascade option and the sections are evaluated Local site Remote site Option Database server -cascade When the server specified in the DB_SERVER parameter is evaluated as being the local server, the values in that section are used. Backup server Database server Backup server The value specified for the BK_SERVER parameter defined in the section described to the left is used.
2. When the following message appears, enter the name of the tape backup management software. KAVX0411-I Enter the backup management product name. Enter one of the following values: If you are using NetBackup: NBU If you are using Backup Exec: BEWS 3. When the following message appears, enter the number of backup catalog storage days. KAVX0417-I Enter the number of days to store the backup catalog. For details on the value to be specified, see the notes at the end of this procedure. 4.
Application-Agent-installation-directory\DRM\conf\tape\NBU_sample.dat • Backup Exec Application-Agent-installation-directory\DRM\conf\tape\BEWS_sample.dat Configuration parameters The parameters in the configuration definition file for tape backup are in the order indicated in the table below.
Parameter Description Specify the absolute path of the mount point of the volume to be backed up (use a maximum of 255 bytes). If no entry for the mount point to be backed up can be found, values specified for CLASS or POLICY and for SCHEDULE under "MOUNT_POINT=default" will be used. Specify the mount point in the tape backup configuration definition file in the order in which Application Agent automatically mounts them during a tape backup.
Parameter Description Specify the absolute path of the mount point of the volume to be backed up (use a maximum of 255 bytes). If no entry for the mount point to be backed up can be found, a value specified for SCRIPT_FILE under "MOUNT_POINT=default" will be used. If the value of this parameter is set to default, the value specified by SCRIPT_FILE will be used for all directories. Specify the mount point to be mounted during a tape backup.
PATH1= SUBDIRS1= Parameters other than INST_PATH, NBU_MASTER_SERVER, INCLUDE_EXEC, and PARALLEL_COUNT can be specified multiple times for a mount point. However, note the following: • Always leave the line before the MOUNT_POINT line blank. • Specify the configuration definition parameters in the order listed in the table above. • When entering default ("default" must be lowercase) for MOUNT_POINT, the entry must come after all the other entries.
User-defined configuration definition file When you execute the drmmediabackup command or the drmmediarestore command with the -bup_env option (that is, when specifying a specific configuration definition file for each command), make sure that you create the corresponding configuration definition file in advance. When you execute the drmmediabackup or drmmediarestore command without the -bup_env option, do not delete the default configuration definition file.
• File name You can use alphanumeric characters to specify any name. • File storage location This file can be stored in any location. In a cluster environment, however, this file should be stored in a shared directory that can be accessed from both servers. • Contents of a file Specify copy group names, following the rules listed below: Write each copy group name from the leftmost column. Copy group names are case-sensitive. A line beginning with a number sign (#) is handled as a comment line.
• drmsqllogbackup File name of the list file Use alphanumeric characters to specify the file name. List file storage location List files are stored in the following locations: • For a database list file: This file can be stored in any location. However, in a cluster environment, we recommend that you store this file in a shared directory that can be accessed from both servers.
In a cluster environment, however, this file should be stored in a shared directory that can be accessed from both servers. Contents of the transaction log list file The contents of the file must follow these coding rules: • • • • • • • Enter the name of the database, followed by the name of the transaction log file. Enclose the name of the database in square brackets "[" and "]". Enter the transaction log files for each database in the order in which they are to be applied.
Updating a dictionary map file After environment setup is complete, the next step is to update the dictionary map files. By updating the dictionary map files, you can confirm that the environment has been set up appropriately. To update the dictionary map files: • When the backup is of a file system: Execute the drmfsdisplay command with the -refresh option specified. • When the backup is of an SQL Server database: Execute the drmsqldisplay command with the -refresh option specified.
You can create a copy group mount definition file even if the configuration is other than the above. If a copy group mount definition file does not exist, Application Agent determines a mount point automatically. Once you create a copy group mount definition file, you can perform the following operations: • You can mount the secondary volume to a specified mount point.
1. Conceal all secondary volumes from the backup server. To conceal all secondary volumes from the backup server, execute the drmdevctl command on the backup server. BKServer > drmdevctl -detach BKServer > 2. Use the Windows disk management functionality or a RAID Manager command to confirm that the secondary volumes are concealed. The following example, uses the inqraid command of RAID Manager: BKServer >echo hd0-1 | inqraid -CLI DEVICE_FILE PORT SERIAL LDEV CTG Harddisk0 Harddisk1 BKServer > 3.
4. Confirm that backup catalogs have been created for all generations. To confirm that the backup catalogs are created for all generations, on the file server or database server execute a command to display backup information (drmfscat command, drmsqlcat command, or drmexgcat command). The number of backup information items should match the number of generations. When the backup is of a file system: FSServer > drmfscat mp_on_dg01.dat INSTANCE: mp_on_dg01.
START-TIME: 2004/12/15 10:20:00 END-TIME: 2004/12/15 10:23:00 HOSTNAME: DB_SVR1 : : When the backup is of an Exchange database: DBServer > drmexgcat STORAGEGROUP: STR1 BACKUP-ID:0000000001 BACKUP-MODE: ONLINE STORAGEGROUP: STR1 ORIGINAL-ID:0000000001 START-TIME:2004/12/15 10:00:00 END-TIME: 2004/12/15 10:01:00 HOSTNAME:EXG_SRV1 : : STORAGEGROUP: STR1 BACKUP-ID:0000000002 BACKUP-MODE: ONLINE STORAGEGROUP: STR1 ORIGINAL-ID:0000000002 START-TIME:2004/12/15 10:10:00 END-TIME: 2004/12/15 10:11:00 HOSTNAME:EXG_SR
6. Transfer exported backup catalog information from the file server or database server to the backup server. Transfer the temporary file exported from the file server or the database server to the backup server. To transfer, execute the ftp (file transfer protocol) command. In the following example, the FTP root folder is assumed to be C:\FTP_ROOT: FSServer> ftp backup-server-name ftp> Username: (Enter the login name) ftp> password: (Enter the password) ftp> binary ftp> put C:\tmp\0000000001.
9. Unmount the secondary volume. To unmount the secondary volume on the backup server, execute the drmumount command. The secondary volume becomes concealed as the drmumount command is executed. BKServer > drmumount 0000000001 KAVX0001-I The drmumount command will now start. KAVX0406-I Unmounting will now start. Mount point = X: KAVX0407-I Unmounting has finished. Mount point = X: KAVX0002-I The drmumount command will now end. 10.
11. Check the contents of the created copy group mount definition file. The created copy group mount definition file is stored in the Application-Agent-installation-directory\DRM\conf\vm. Use a text editor to open the created copy group mount definition file, and check the contents of the file. Change the FS values (mount point of the mount target) as necessary. If you do not create a copy group mount definition file, this step is not necessary.
#configuration file for copy-group mounting [CG_DEF] COPY-GROUP-NUM=6 COPY-GROUP=Grp1,dev11 BUID=0000000001 DATETIME=20041201123456 COPY-GROUP=Grp1,dev12 BUID=0000000001 DATETIME=20041201123456 COPY-GROUP=Grp2,dev21 BUID=0000000002 DATETIME=20041201123456 COPY-GROUP=Grp2,dev22 BUID=0000000002 DATETIME=20041201123456 COPY-GROUP=Grp3,dev31 BUID=0000000001 DATETIME=20041201123456 COPY-GROUP=Grp3,dev32 BUID=0000000003 DATETIME=20041201123456 [DG_DEF] DISK-GROUP-NUM=2 BUID=0000000001 DATETIME=20041123225248 COPY
COPY-GROUP-NUM=2 COPY-GROUP=Grp3,dev31 COPY-GROUP=Grp3,dev32 P-DG=DG01 S-DG=DG01 FS-NUM=3 LVM-DEVICE=\Device\HarddiskDmVolumes\Dg01\Volume1 LABEL=Volume1 FS=X: LVM-DEVICE=\Device\HarddiskDmVolumes\Dg01\Volume2 LABEL=Volume2 FS=Y: LVM-DEVICE=\Device\HarddiskDmVolumes\Dg01\Volume3 LABEL=Volume3 FS=Z: 12. Execute the drmdevctl command to conceal all secondary volumes from the backup server. To perform a backup or restore, conceal all secondary volumes from the backup server.
2. Confirm that a backup catalog has been created. To confirm that the backup catalog has been created, execute a command (drmfscat, drmsqlcat, or drmexgcat) to display backup information on the file server or database server. When the backup is of a file system: FSServer > drmfscat mp_on_dg01.dat INSTANCE: mp_on_dg01.dat BACKUP-ID:0000000001 BACKUP-MODE: COLD INSTANCE: mp_on_dg01.
4. Transfer the backup catalog information exported from the file server or the database server to a backup server. Transfer the temporary file exported from the file server or the database server to the backup server. To transfer, execute the ftp (file transfer protocol) command. In this case, set the FTP route folder C:\FTP_ROOT. FSServer> ftp backup-server-name ftp> Username: (Enter the login name) ftp> password: (Enter the password) ftp> binary ftp> put C:\tmp\0000000001.drm ftp> quit FSServer> 5.
7. Create a copy group mount definition file. To create a copy group mount definition file, on the backup server execute the drmmount command with the backup ID and -conf option specified. If you do not create a copy group mount definition file, execute the drmmount command without the -conf option specified to confirm that a mount operation can be performed. BKServer > drmmount 0000000001 -conf KAVX0001-I The drmmount command will now start. KAVX0400-I Mounting will now start.
9. Check the contents of the created copy group mount definition file. Use a text editor to open the created copy group mount definition file, and check the contents of the file. Change the FS value (mount point of the mount destination) as necessary. If you do not create a copy group mount definition file, this step is not necessary.
Determination of a mount point The following table describes how Application Agent determines a mount point when executing the command for mounting a second volume. The following explains separately when the command is executed with a backup ID specified in the option, and when the command is executed with a copy group specified in the option.
The system uses a mount point specified in FS=mountpoint that corresponds to a copy group to be backed up, in the copy group mount definition file.
If you specify an absolute path that begins with the drive letter as a mount point directory name in the -mount_pt option, the mount point should be absolute-path-specified-as-a-mount-pointdirectory-name\mount-point-in-the-copy-group-mount-definition-file For example, when W:\ is specified in the -mount_pt option and FS=D:\ is specified in the copy group mount definition file, the system mounts the unoccupied drives from the W drive.
Tasks File server or database server Backup server Location in this manual Specify how the extended commands are to be started. Needed Needed Specifying how to start the extended commands, page 182 Create host environment settings files.
Execution permissions for extended commands To execute the extended commands, you need system administrator permissions. Make sure that you have these permissions. Automatic execution of extended commands A backup or restoration process that uses extended commands may be performed across more than one server. It is difficult to execute extended commands at each of physically remote servers. Therefore, you need to perform remote execution of the extended commands.
3. Delete the operation-ID directory created in the script_work directory in the same level of the directory that you checked in step 2. (Perform the following steps on the backup server:) 1. Confirm the name of the temporary directory for extended commands. Check whether FTP_HOME_DIR-value\FTP_SUB_DIR-value exists. 2. Delete the operation-ID directory created in the directory that you checked in step 1.
In the Windows default settings, wscript.exe starts an extended command and a pop-up dialog box appears. The dialog box displays the execution results of the command. Therefore, note that an extended command cannot execute automatically. Creating a host environment settings file Host environment settings files are created depending on the configuration of the system using extended commands. These files must be located on each server.
On the file or database server: EX_DRM_HOST_DEF_CHECK -db -f host-environment-settings-file On the backup server: EX_DRM_HOST_DEF_CHECK -bk -f host-environment-settings-file If an error is found, revise the contents of the file, and then re-execute the check tool. Creating operation IDs An operation ID is a unique character string given to each resource that is to be the processing target of an extended command.
After storing the operation definition file in the stipulated directory, execute the check tool for operation definition files to check the content of the operation definition file and automatically create a temporary directory for extended commands. Deployment of operation definition files Create two same operation definition files for each target mount point or file. Place one operation definition file on the file server and the other on the backup server.
The following table shows the entries of an operation definition file, their contents, and the maximum value length (in bytes). Specification of these entries is required. Table 56 Example: contents and maximum value length of each entry in an operation definition file (when the backup is of a file system) Specification item Information to be specified Maximum number of characters BACKUP_OBJECT Specify a character string identifying the type of the backup object. Specify FILESYSTEM.
The following prerequisites apply to this example: • There are two file servers and a backup server to which a tape device is connected. • A virtual server VServer_C is running on the file server A (server name: FSServer_A) and the file server B (server name: FSServer_B), which make up the cluster. • DRM_DB_PATH=directory-name-on-shared-disk;virtual-server-name has been set in the Application Agent configuration definition file (init.conf).
1. Create an operation definition file on the active server (file server A), and then set the required items. Create the following two files. Application-Agent-installation-directory\DRM\script\conf\_Operation_A.dat Application-Agent-installation-directory\DRM\script\conf\_Operation_B.dat The following shows an example of entries in the file Application-Agent-installation-directory\ DRM\script\conf\_Operation_A.dat.
4. On the backup server, execute EX_DRM_FS_DEF_CHECK, which is a check tool for operation definition files. EX_DRM_FS_DEF_CHECK operation-ID -bk The operation definition file is checked. In addition, a temporary directory for extended commands is automatically created. If an error occurs, change the backup server configuration depending on the contents of the operation definition file, or change the contents of the operation definition file depending on the backup server configuration.
Figure 58 Example of locations of operation definition files (when the backup is of an SQL Server database) In this example, the virtual server (server name: VServer_C) is running on the database server A (server name: DBServer_A) and the database server B (server name: DBServer_B), which make up a cluster. The SQL Server instance INSTANCE_1 exists on the databases A and B, and the virtual server C has been defined as a cluster resource.
Specification item Information to be specified Maximum number of characters DB_SERVER_NAME Specify the database server name. In a cluster configuration, specify the name of the virtual server that corresponds to the cluster resource. This virtual server name must be defined in DRM_DB_PATH=directory-on-the-shared-disk;virtual-server-name in the Application Agent configuration definition file (init.conf). 128 This entry specifies an SQL Server instance name.
Example of creating operation definition files The following gives an example of creating operation definition files for a system that has a cluster configuration. This example assumes the following conditions: • The virtual server (server name: VServer_C) is running on the database server A (server name: DBServer_A) and the database server B (server name: DBServer_B), which make up the cluster.
In this example, you create an operation definition file for each of the operation IDs listed in the following table. Table 62 Operation IDs for which to create operation definition files (when the backup is of an SQL Server and in a cluster configuration) Operation ID Target database server Target instance and database Operation_A VServer_C Entire INSTANCE_1 To create operation definition files (in a cluster configuration): 1.
4. On the backup server, execute EX_DRM_SQL_DEF_CHECK, which is a check tool for operation definition files. EX_DRM_SQL_DEF_CHECK operation-ID -bk The operation definition file is checked. In addition, a temporary directory for extended commands or VDI metafile storage directory is automatically created.
Figure 59 Example of locations of operation definition files (when the backup is of an Exchange database) In this example, the virtual server (server name: VServer_C) is running on the database server A (server name: DBServer_A) and the database server B (server name: DBServer_B), which make up the cluster. The storage group STORAGE_1 exists on the databases A and B, and the virtual server C has been defined as a cluster resource.
Specification item Information to be specified Maximum number of characters DB_SERVER_NAME Specify the database server name. In a cluster configuration, specify the name of the virtual server that corresponds to the cluster resource. This virtual server name must be defined in DRM_DB_PATH=directory-on-the-shared-disk;virtual-server-name in the Application Agent configuration definition file (init.conf)#1. 63 INSTANCE_NAME Enter a hyphen (-).
#4 If you specify this item, do not specify a recovery database. #5 The specified subdirectory must satisfy the following: • The subdirectory name must not contain a delimiter character (\). • The subdirectory name must not be specified as only one period (.) or two periods (..). • The root directory must not be specified. #6 Means 1,024 characters (either 1-byte or 2-byte characters), not 1,024 bytes. The character count does not include the characters in the header TARGET_NAME=.
This example creates operation definition files corresponding to the operation IDs listed in the following table. Table 65 Operation IDs for which an operation definition file is created (when the backup is of an Exchange database and in a cluster configuration) Operation ID Virtual server Target storage group Operation_A VServer_C STORAGE_1 To create an operation definition file for a cluster configuration: 1.
4. On the backup server, execute EX_DRM_EXG_DEF_CHECK, which is a check tool for operation definition files. EX_DRM_EXG_DEF_CHECK operation-ID -bk The operation definition file is checked. In addition, a temporary directory for extended commands is automatically created. If an error occurs, change the backup server configuration depending on the contents of the operation definition file, or change the contents of the operation definition file depending on the backup server configuration.
For example, if the dictionary map file directory is L:\PTM, the temporary directory name is L:\ script_work\operation-ID\DB. Temporary files created during execution of the extended command: • Backup ID record file • Backup information file • SQL Server's metafile (When the backup is of an SQL Server database) Checking the copy-group list file Certain extended commands allow you to specify a copy-group list file as an option value.
Creating a user script By specifying a user script when executing the backup command, you can simplify the backup operation because the commands specified in the user script will execute during the backup operation. You can enter the desired commands in a user script according to the coding rules for scripts. A user script file is a file that contains user scripts.
Applicable items Rules • The value of an item must be specified after the one-byte equal sign (=) that follows the item name. Do not enter a linefeed code between an item's name and value. Item values • The item value begins after the one-byte equal sign (=) that follows the item name and ends at the linefeed code. • If an item name is specified alone (without a value), an error results. • A line that begins with a one-byte hash mark (#) is regarded as a comment line.
Entry name Meaning and specifiable values Multiple specification#1 [SPLIT_PROC]#2, User processing section: Indicates the beginning of the "user processing for pair splitting". In this section, define commands to be executed as part of user processing for pair splitting. Not allowed Allowed#3 User processing section: Indicates the beginning of the "user processing for termination". In this section, define commands to be executed as part of user processing for termination.
Entry name Meaning and specifiable values Multiple specification#1 Omission Not allowed Not allowed Specifies a command line to be executed. • If you want to specify an option, separate the command name and the option with a one-byte space character. • One command definition section must contain only one command line. • A maximum of 2048 characters can be specified on a command line.#6 • A path or file name containing one or more spaces must be enclosed in double quotation marks (").
Entry name Meaning and specifiable values Multiple specification#1 Omission Allowed#4 Allowed Not allowed Allowed Specifies the environment variables to be applied for execution of the specified command. • Only one environment variable is allowed per line. • The name and value of an environment variable must be separated by a one-byte equal sign (=). • One environment variable definition can consist of up to 2048 characters.
Entry name Meaning and specifiable values Multiple specification#1 Omission Not allowed Allowed Not allowed Allowed Not allowed Allowed Specifies the command timeout value. Specifiable values are as follows: Timeout value (in seconds)#8 • The range of specifiable values is from 0 to 86400. • Use one-byte numeric characters only. • No timeout will occur if 0 is specified here. TIMEOUT#5 NOWAIT Once a command starts successfully, there is no wait for the command to finish.
#1 "Multiple specification is not allowed" means that multiple entries cannot be specified in the command definition section (CMD). #2 The [PRE_PROC], [RESYNC_PROC], [SPLIT_PROC], [FINISH_PROC], and [POST_PROC] can be defined in any order. #3 This item can be omitted if there is no command to be executed within the user processing section. #4 This item can be used more than once within a user script file.
Script environment variables The table below shows the Application Agent script environment variables that can be referenced by commands in the user post-processing section. These script environment variables can be referenced from the local server and the backup server. Table 72 Script environment variables Environment variable Description Remarks This environment variable is valid in the following cases: DRMENV_L_BACKUPID Backup ID of local server • The command is being executed in the local server.
Do not specify a command that is not listed. Doing so might prevent the system from running normally.
Examples of user script usage The following table provides references to the sections in which examples of user script usage are explained.
Parameter SMTP_SERVER Default (value after installation) Description Specifies the host name or IP address of the SMTP server to be used for sending the e-mail. The IP address can be specified in either IPv4 format or IPv6 format. mailserver Specify as a string of up to 64 one-byte characters. Specifies an e-mail send destination address. TO Specify as a string of up to 255 one-byte characters.
Protection Manager message: Hostname=flora1 DRM_HOSTNAME=ExchangeServer1 2005/10/08 19:22:32(00003896) drmexgbackup -mode vss 2005/10/08 19:22:32(00003896) KAVX0001-I The drmexgbackup command will now start. 2005/10/08 19:22:33(00003896) KAVX0256-I A connection to the Protection Manager service will now be established. Host name = 10.106.136.125 2005/10/08 19:22:33(00003896) KAVX1600-I Backup processing will now start. 2005/10/08 19:22:58(00003896) KAVX5108-I Resynchronizing a copy group.
• Copy-group list file For details on the file name and location of the copy group list file, see “Creating a copy-group list file” on page 159. • Database list file or mount point directory list file For details on the file names and locations of the database list files and mount point directory list files, see “Creating a database list file or mount point directory list file” on page 160.
Execute EX_DRM_FS_BACKUP, and then carry out the following procedure on the file server: 1. Execute EX_DRM_DB_EXPORT. Backup information is exported and recorded in the backup information file, which is in the temporary directory for extended commands. Back up the files in the temporary directory for extended commands# to a location such as tape.
Figure 61 Protecting backup information for tape backups Back up the tape, and then carry out the following procedure: Execute EX_DRM_TAPE_BACKUP, and then carry out the following procedure on the backup server: • Use a tape backup product to back up all the files in the dictionary map file directory#. # For details on the dictionary map file directory on the backup server, see “Creation in a noncluster configuration or on a backup server” on page 129.
Recovering backup information (file server) This section explains how to recover backup information and restore the file system after restoring the operating environment when the backup information on the file server is damaged. To recover backup information (file server): 1. From the backup data that was backed up in “Application Agent backup information” on page 213, decide the generation and backup information to be restored. 2.
4. Carry out the procedure described in this manual for restoration. Creating the Exchange environment configuration file The Exchange environment configuration file is used to specify the information required for Exchange Server linkage. If you use the automatic seed function in a CCR configuration, you must create the Exchange environment configuration file.
Parameter Description Default Maximum Character String (bytes) OFF 3 OFF 3 In Exchange Server 2007: If you have specified ON in the EXG_SCR_SEED parameter, specify whether seed processing is to be executed even if a connection cannot be established with some of the targets set for the storage group. Specify ON to execute seed processing even if connections with some of the targets fail. Specify OFF if an error is to result when a connection with even one of the targets fails.
4 Application Agent operations Application Agent operations differ depending on the object to be backed up. This chapter provides an outline of and precautions on extended commands and basic commands used when operating Application Agent. It also gives precautions common to all objects subject to backup. The descriptions in this chapter apply to the Application Agent CLI.
Basic commands Basic commands provide data protection functions such as backup and restore. Extended commands execute basic commands internally for processing. Therefore, you need to have an understanding of basic commands before you use extended commands. If an error occurs during extended command operation, log information of the basic command being executed in the extended command is output.
Corresponding basic command Function -- Collects a backup information file from the backup server. If the backup is of an SQL Server database, the metafile is also collected. EX_DRM_FTP_PUT -- Transfers a backup information file to the backup server. If the backup is of an SQL Server database, the metafile is also transferred. EX_DRM_HOST_DEF_CHECK -- Checks the contents of the host environment setup file. EX_DRM_RESYNC drmresync Resynchronizes a copy group.
Extended command Corresponding basic command Function EX_DRM_UMOUNT drmumount Unmounts a secondary volume. -- drmtapecat Displays a listing of backup information for a tape. -- drmtapeinit Registers the parameters for a tape backup management software. Legend: -- The corresponding command does not exist.
Extended command Corresponding basic command Function • Displays a listing of SQL Server database information. -- drmsqldisplay • Updates the SQL Server database information in the dictionary map files. -- drmsqlinit Registers SQL Server parameters. -- drmsqlrecover Recovers a restored SQL Server database. -- drmsqlrecovertool Recovers a restored SQL Server database via a GUI. Legend: -- The corresponding command does not exist.
Backing up and restoring a file system, page 243 • When the backup is of an SQL Server database: Backing up and restoring an SQL Server database, page 275 • When the backup is of an Exchange database: Backing up and restoring a storage group, page 399 Precautions on operations • When using Application Agent in a cluster configuration, specify the same instance in the RAID Manager linkage definition file (DEFAULT.dat) as the instance for the active server and standby server.
Figure 62 Correspondence of backup catalog and copy group When you back up database A only (as in step 2), the backup ID 0000000002 is created, and the old backup information (backup ID: 0000000001) about the copy group VG01,obj01 on database A is deleted from the backup catalog.
• If there are multiple mount points specified for a single logical volume, as shown below, any backup will fail and an error message (KAVX0006-E and DRM-10064) will be output. • Two mount points (M:\, M:\MNT) are specified for Harddisk1. • Two mount points (M:\, L:\MNT) are specified for Harddisk1.
Precautions on restoration • Make sure that the pair statuses of the copy group to be restored are PSUS (for primary volume) and SSUS (for secondary volume). • Do not use thedrmresync command while the drmmediabackup command is backing up data from tape to the secondary volume, while the drmmediarestore command is restoring data from the secondary volume to tape, or while the drmmount command is mounting the secondary volume.
Precautions for executing commands in a cluster environment In a cluster environment, before executing a Application Agent command, you must set the environment variable DRM_HOSTNAME as shown below.
Precautions regarding the drive when mounting data to a backup server When Application Agent performs tape backup, it mounts the secondary volume to the backup server. If you specify a mount point in the command arguments when performing tape backup, Application Agent automatically uses the next available drive (in alphabetical order) after the specified drive excluding drive A and drive B. You cannot specify a drive other than the first drive for mounting data.
For details on the procedure for updating the dictionary map files, see “Updating a dictionary map file” on page 163. Unavailable tools when executing commands During execution of a Application Agent command, if the primary or secondary volume is being used by another application, a problem might occur. For example, RAID Manager might not operate normally, a mount or unmount operation might fail, or a data inconsistency might occur.
Figure 63 Disk group configuration that does not allow backup In such a case, change the configuration to one of the following volume configurations: • Delete the logical volume that has not been mounted, from the disk group. • Mount the logical volume onto the file system so that the file system is eligible to be backed up.
message appears, indicating that the user has insufficient permissions to execute the command, and command execution fails. • Database access permissions • When the backup target is an SQL Server database When executing a Application Agent command, the user accesses the SQL Server using Windows authentication. Therefore, users who execute Application Agent commands should be registered as a member of the sysadmin fixed server role of the SQL Server.
Extended command name OS administrator permissions#1 Database access permissions drmcgctl Needed Not Needed drmclusinit Needed Not Needed drmdbexport Needed Not Needed drmdbimport Needed Not Needed drmdevctl Needed Not Needed drmhostinfo Needed Not Needed drmresync Needed Not Needed #1 In a Windows Server 2008 or Windows Server 2012 operating environment, this also includes OS administrator privileges. #2 An FTP user requires OS administrator permissions on the backup server.
In a Windows Server 2008 or Windows Server 2012 operating environment, this also includes OS administrator privileges. Table 86 Required permissions for each command (utility commands) Extended command name OS administrator permissions# Needed drmdbsetup Database access permissions Not Needed # In a Windows Server 2008 or Windows Server 2012 operating environment, this also includes OS administrator privileges.
Extended command name OS administrator permissions# Database access permissions EX_DRM_EXG_RESTORE Needed Needed EX_DRM_EXG_VERIFY Needed Not Needed drmexgbackup Needed Needed drmexgcat Needed Not Needed drmexgdisplay Needed Needed drmexgrestore Needed Needed drmexgverify Needed Not Needed # In a Windows Server 2008 or Windows Server 2012 operating environment, this also includes OS administrator privileges.
This command cannot be executed with other commands when the -refresh option is specified. This command can be executed with other commands when other options are specified. Table 90 Possibility of concurrent execution with other commands (common commands) Command Function summary drmappcat Displays catalog information on a host. Possible to run with other commands? Yes#1 • Locks copy groups. drmcgctl • Unlocks copy groups. No#2 • Displays a list of copy groups.
Legend: No: These commands cannot execute concurrently with other commands marked by No. Yes: These commands can execute concurrently with any other command (marked by Yes or No). #1 These commands can execute concurrently with any other command only when the tape backup management software is running. #2 When these commands are executed with the -delete option, these commands cannot execute concurrently with other commands.
#2 This command can be executed with the drmresync command only when all of the following conditions are satisfied: • In Application Agent, the drmsqlbackup command has been executed for the target instance. • The -no_cat option is not specified. • The -v option is not specified. • The -lsn option is not specified. • The -d option is not specified. • The instance name is specified.
indefinitely until the former command terminates. To prevent the system from attempting to execute such a command, you need to edit the Application Agent configuration definition file (init.conf), which is stored in the following location: Application-Agent-installation-directory\DRM\conf\init.conf init.
Specify a period longer than the following: (the period from the time EX_DRM_TAPE_BACKUP or EX_DRM_TAPE_RESTORE is executed to the time the tape backup management software starts) x (the number of commands executed concurrently) Configurations that requires RAID Manager user authentication Application Agent commands support RAID Manager user authentication when the RAID Manager version is 01-25-03/01 or later.
Restarting the Protection Manager service When the contents of the Application Agent init.conf and DEFAULT.dat definition files are changed, you must use the Services pane in the Windows to restart the Protection Manager service. Stopping the Protection Manager service When Application Agent is removed, the Protection Manager service stops and is deleted from Windows.
Application Agent operations
5 Operation for file systems This chapter describes the procedures for running Application Agent when backing up a file system. The descriptions in this chapter apply to the Application Agent CLI. For details on setting up commands, see the manual HP P9000 Replication Manager Software Application Agent CLI Reference Guide. Backing up and restoring a file system This section provides notes on backing up and restoring a file system.
• Do not use the drmfsrestore command in the following situations: • When you back up a file system from a secondary volume to tape device by using the drmmediabackup command. • When you restore a file system from the tape device to a secondary volume by using the drmmediarestore command. • When you mount a secondary volume by using the drmmount command.
Figure 64 System configuration for backing up and restoring data between volumes (when the backup is of a file system) Processing flow The processing flow for backing up a file system to a secondary volume and restoring it to the primary volume on a single-server system is shown below. Execute commands on the file server.
Figure 65 Procedure for backing up a file system to the secondary volume Figure 66 Procedure for restoring a file system to the primary volume Backing up a file system to a secondary volume This section provides an example of backing up the mount point E: on the file server FSServer to a secondary volume in an online backup operation. Use Operation_A as an operation ID to identify the object to be backed up. To back up the file system: 1. Resynchronize the copy groups.
Restoring a file system to a primary volume This section gives an example of restoring a file system when the backup data is stored on a secondary volume. This example restores the data from the secondary volume to the primary volume by resynchronizing the volumes. This example uses Operation_A as the operation ID that identifies the file system to be restored. To restore the file system: 1. Check the backup ID of the backup data.
Figure 67 System configuration for backing up a file system to tape and restoring it from tape With Application Agent, file servers can be clustered. With file servers configured in a failover cluster, if the active server fails, a standby server is available to take over operations from the active server. This example assumes the following conditions: • The system is equipped with a file server (server name: FSServer), and a backup server (server name: BKServer) with a tape device connected to it.
Flow of processing This section describes the processing flow of backing up a file system to tape and restoring it to the primary volume on a multi-server system. A single-file-server system is used in this example, but a multi-file-server system may be used as appropriate. • When usual operation is in the split state, resynchronize the copy group before backing up the data. • When usual operation is in the pair status, there is no need to resynchronize the copy group before backing up data.
Figure 68 Procedure for backing up a file system to tape Figure 69 Procedure for restoring a file system from tape 250 Operation for file systems
Backing up a file system to tape This section gives an example of backup of a file system to a tape. In this example, we back up the file server FSServer to a secondary volume in an online backup operation before backing it up from a secondary volume to tape. As a mount point, we use E:. The example uses Operation_A as the operation ID that identifies the file system to be backed up.
2. Check that the file system has been backed up correctly. To check that backup information bearing the date on which the file system had been backed up from the primary volume to the secondary volume is available, execute the drmfscat command on the file server.
7. Back up the data from the secondary volume to a tape. To back up the data, execute EX_DRM_TAPE_BACKUP on the backup server. For the tape backup mount point, specify the E drive (drive letter E:). BKServer > EX_DRM_TAPE_BACKUP Operation_A -mount_pt E: Executing a backup in this example will load information about the backup operation in the backup catalog with new backup ID 0000000002. 8. Check that the file system has been backed up correctly.
3. Export the backup information to temporary files. To perform backup from the secondary volume to tape, you need to copy information, about the backup operations from the primary volume to the secondary volume, to the backup server. To export information about backup to temporary files, execute EX_DRM_DB_EXPORT. The temporary files are stored in the temporary directory for extended commands, on the file server. FSServer > EX_DRM_DB_EXPORT Operation_A 4.
To restore the file system: 1. Check the backup ID of the backup data. Check the backup ID of the backup data used to restore it from tape to the secondary volume. To check the backup ID, execute the drmtapecat command on the backup server.
6. Import the temporary files, which were transferred from the backup server, into the backup catalog on the file server. Import the temporary files, which were transferred from the backup server, into the backup catalog on the file server. To import the temporary files, execute EX_DRM_DB_IMPORT on the file server. FSServer >EX_DRM_DB_IMPORT Operation_A 7. Check that the temporary files have been imported correctly.
Figure 70 System configuration for backup of a file system to tape Processing flow In this example, the processing flow of the drmfsbackup command that specifies a user script is as follows: • All application programs that are using the F drive are stopped (this is a step in the user pre-processing section). • The cache of the secondary volume is cleared (this is a step in the user pre-processing section).
Figure 71 Processing flow User script example The table below shows the user script that is created for the example. Table 96 User script creation example Script text Explanation LOCAL_BACKUP=YES ...(1) (1) YES is required. #Pre-processing section [PRE_PROC] ...
Script text Explanation #Stop the application programs that are using the file system CMDLINE=C:\tmp\stop_app.bat ENV=VAL=ApplicationName ...(3) (3) Pass the environment variable VAL= ApplicationName to the command. TIMEOUT=60 ...(4) (4) Take a 60-second timeout. END_CODE=100 ...(5) (5) Treat a command return value of 100 or greater as an error. LOCATION=LOCAL ...(6) #Clearing of the secondary volume cache (mounting and unmounting) (6) Execute on the local server.
Script text Explanation [POST_PROC] ...(8) #Tape backup of the secondary volume (8) Start of the user post-processing section [CMD] CMDLINE=C:\tmp\tapebackup.bat TIMEOUT=NOWAIT ...(9) (9) Execute the next command without waiting for the command to terminate. END_CODE=TERMINATE_NZ ...(10) (10) Treat a command return value of a value other than 0 as an error. LOCATION=REMOTE PARENT_STAT=NORMAL ...(11) [CMD] #Application recovery processing if there is a backup error CMDLINE=C:\tmp\recovery.
Script text Explanation #Re-start the application programs that use the file system CMDLINE=C:\tmp\restart_app.bat ENV=VAL=ApplicationName TIMEOUT=60 END_CODE=100 LOCATION=LOCAL The following is an example of backing up the secondary volume to tape by using tapebackup.bat: rem Use NTBACKUP to execute the job Job1,and perform a copy backup of G:\ to tape Tape1 rem Set the backup comment passed to the DRMENV_COMMENT environment variable as the backup job explanation "C:\Windows\system32\ntbackup.
PROMPT> drmfsbackup F: -mode cold -script C:\tmp\script.txt -s BKHOST -auto_import -auto_mount G: -comment TEST1 KAVX0001-I The drmfsbackup command will now start. KAVX0256-I A connection to the Application Agent will now be established. Host name = BKHOST KAVX0210-I The user script will now start. Procedure section = [PRE_PROC] KAVX0263-I A command in the user script will now be executed. Host name = LOCAL Command line = C:\tmp\stop_app.bat KAVX0213-I A command from the user script has ended.
Backing up and restoring between local sites and remote sites This section shows examples for using volume replication functionality between storage systems, such as Continuous Access Software or Continuous Access Journal, to back up a file system from the primary volume in the local site to the secondary volume in the remote site, and for performing restoration from the secondary volume in the remote site to the primary volume in the local site.
• The system is equipped with a file server in the local site (server name: FSServer), and has a backup server (server name: BKServer) with a tape device connected to it. • An FTP service is running on the backup server of the remote site, and is set such that users who can log on to the OS on each file server can be used to log on to the FTP server and transfer files. The FTP user ID must be admin, and the FTP user password must be password.
3. Check that backup was performed correctly. On the file server, execute the drmfscat command, and check the backup information.
2. Check that the temporary files have been imported correctly. To check that backup information bearing the date on which the temporary files had been backed up from the primary volume to the secondary volume has been imported on the backup server, execute the drmfscat command on the backup server.
Restoring a file system to the local site This subsection explains an example for restoring a file system to a local site, when the local site is running normally, without failure. Procedures at the local site and remote site are explained separately. Operations at the remote site 1. Split the Continuous Access Software pair (usual operation is in the pair status). BKServer > pairsplit -g TC01 -d FS11 -rw BKServer > 2. Check the backup ID of the backup data.
5. Export the backup information to temporary files. To perform restoration from the secondary volume to the primary volume, you need to copy backup information about the restoration operations from tape to the secondary volume, to the file server. To export the backup information to temporary files in the temporary directory for extended commands, execute EX_DRM_DB_EXPORT. BKServer > EX_DRM_DB_EXPORT Operation_A Operations at the local site 1.
5. Resynchronize the copy group. To resynchronize the copy group, execute EX_DRM_RESYNC on the file server. FSServer > EX_DRM_RESYNC Operation_A Restoring a file system to a remote site This subsection explains an example for restoring a file system in a remote site when failure occurs in the local site. Operations at the remote site 1. Split the Continuous Access Software pair (usual operation is in the pair status). BKServer > pairsplit -g TC01 -d FS11 -S BKServer > 2.
5. Use the disk management function to mount the restored volume. The mount is to the G: drive. Recovering a file system from a remote site to the local site This subsection explains how to recover restored data in the remote site to the local site, after the file system is restored in the remote site and recovery has been performed for the failure in the local site. Procedures at the local site and remote site are explained separately.
3. Execute EX_DRM_FS_BACKUP to back up the file system to the secondary volume. 4. Execute EX_DRM_DB_EXPORT to export the backup information to temporary files. 5. Execute EX_DRM_FTP_PUT to transfer the temporary files to the backup server. 6. Execute EX_DRM_DB_IMPORT to import the temporary files transferred from the file server to the backup catalog on the backup server. 7. Check the backup ID. To check the backup ID, open the backup ID record file (operation-ID.
2. When running Explorer or any other tool, put the temporary directory for extended commands that is on the backup server in a shared state. 3. From the file server, mount the temporary directory for extended commands that is on the backup server. When running Explorer or any other tool on the file server, assign the shared temporary directory for extended commands to the backup server as a network drive. For now, assign it to X: on the file server. 4.
2. Execute EX_DRM_TAPE_RESTORE to restore the backup data from the tape to the secondary volume. 3. Execute EX_DRM_DB_EXPORT to export the backup information to temporary files in the temporary directory for extended commands. 4. Copy the temporary files from the backup server to the file server.
Operation for file systems
6 Operation for SQL Server databases This chapter describes how to operate Application Agent when backing up an SQL Server database, focusing on actual procedures. The descriptions in this chapter apply to the Application Agent CLI. For details on setting up commands, see the .HP P9000 Replication Manager Software Application Agent CLI Reference Guide Backing up and restoring an SQL Server database This section provides notes on backing up and restoring an SQL Server database.
Precautions when restoring a database • When you attempt to restore an SQL Server system database (master, model, msdb, or distribution), the SQL Server service to be restored needs to be stopped to recover the system database. The database to be restored cannot be accessed temporarily. • During restoration of the SQL Server system databases (master, model, and msdb), do not connect to the SQL Server.
For example, when ODBC sections cannot be established, if an application that performs retries is running, since an ODBC section establishment request is issued during the roll-forward after metafile application, the roll-forward will fail. • When using Enterprise Manager to view a database to be backed up, before performing restoration, either cut the connection to the database in Enterprise Manager, or terminate Enterprise Manager.
Figure 73 Backing up and restoring multiple databases (same volume configuration) When multiple databases are stored in multiple volumes You can specify database names to back up or restore particular user databases within an instance#. The following figure shows backup and restoration for databases stored on the multiple volume.
Figure 75 Tape restoring multiple databases (multiple volume configuration) In the operations shown in the above figure, the data restored from tape is stored on secondary volume 2, but when data on secondary volume 1 is restored to primary volume 1, Application Agent treats this as a completed restoration. As such, EX_DRM_SQL_RESTORE cannot be used to restore DB2 on secondary volume 2 to primary volume 2.
Figure 76 Database statuses during restoration and recovery Note that when a database in the standby status is viewed by Enterprise Manager, before executing the drmsqlrecover command or drmsqlrecovertool command, either cut the connection to the database in Enterprise Manager, or terminate Enterprise Manager.
The following figure shows the example when LOCAL_BACKUP is set to YES and the secondary volume on the local site is backed up. In this example, the point A is used as the start point and log backup 1 and 2 are obtained. Subsequently, the secondary volume on the remote site is backed up when LOCAL_BACKUP is set to NO. The log backup 3 is obtained, with the backup ID 0000000001. When the drmsqllogbackup -v command is executed, log backup 1, 2, and 3 are displayed for the backup ID 0000000001.
Figure 77 Transaction log backups after roll-forward (1) • After performing roll-forward, use Application Agent to perform backup from the primary volume to the secondary volume, and then obtain a backup of the transaction log files. Figure 78 Transaction log backups after roll-forward (2) In this case, when the second backup from the primary volume to the secondary volume is performed, the contents of the secondary volume are updated.
Figure 79 Transaction log backup error after roll-forward Precautions regarding application of transaction logs Application of transaction logs enables a database to be recovered from a backup at a certain point, or when the metafile is damaged. For such a recovery, the transaction logs must be backed up continuously. Also, a recovery cannot be performed if any transaction logs are missing. In addition, transaction log files must be applied in the order they were backed up.
2. Transaction log backups at point C can be applied after the previous transaction log backup (point B) is applied. Recovering the SQL Server database from backup data existing prior to the last backup data You can recover the SQL Server database from backup data existing prior to the last backup data when there is a problem in the latest backup data.
Service status Commands Running Stopping drmsqldisplay (when the -refresh option is specified) Executable (Not applicable) drmsqldisplay (when the -refresh option is not specified) Executable Executable drmsqlinit Executable (Not applicable) drmsqllogbackup Executable (Not applicable) drmsqlrecover Executable (Not applicable) drmsqlrecovertool Executable (Not applicable) Executable Executable Executable (Not applicable) drmsqlrestore (when the system database# is included) drmsqlrest
SQL Server database status Commands Online Offline Reading Read-only Unconfirmed drmsqllogbackup (when the -no_truncate option is not specified) Yes (DB) No No Yes No (becomes read-only) No Yes (DB) Yes (DB) Yes (DB) Yes No (becomes online) (becomes online) No No Yes (DB) (becomes reading) No No Yes (DB) (becomes read-only) Yes (DB) Yes (DB) (becomes online) No Yes (DB)#2 (becomes read-only) drmsqlrecovertool No No (becomes online) Yes (DB) Yes (DB) Conditional#1 Yes (D
If you restore the user database only, delete the database in a state where a restoration command cannot be executed, and then re-execute the restoration. #2 If the database is in the unconfirmed status, it will be deleted and then restored automatically.
SQL Server database status Offline and unconfirmed Reading and unconfirmed Read-only and offline Read-only and unconfirmed Read-only, offline, and unconfirmed No No No No No No No No No No drmsqlrestore Yes (DB)# Yes (DB)# Yes (DB) Yes (DB)# Yes (DB)# (when the -undo option is specified) (becomes read-only) (becomes read-only) (becomes read-only) (becomes read-only) (becomes read-only) drmsqlrestore Yes (DB)# Yes (DB)# Yes (DB) Yes (DB)# Yes (DB)# (when the -undo option is n
SQL Server database type Commands System database User database master drmsqlrecover drmsqlrecovertool drmsqlrestore msdb model No No No (No operation is performed) (No operation is performed) (No operation is performed) tempdb No (terminates with an error) Yes (DB) No No No (No operation is performed) (No operation is performed) (No operation is performed) (terminates with an error) Yes (DB) Yes (DB) Yes (DB) Yes (DB) Yes Yes (DB) No Legend: Yes (DB): The command can be execute
Figure 82 System configuration for backing up and restoring data between volumes (when the backup is of an SQL server database) Processing flow The processing flow for backing up an SQL Server database to the secondary volume and restoring it to the primary volume on a single-server system is shown below. Commands are executed on the database server.
Figure 83 Procedure for backing up an SQL Server database to the secondary volume Figure 84 Procedure for restoring an SQL Server database to the primary volume Backing up an SQL Server database to a secondary volume This section provides an example of performing an online backup of an SQL Server database between volumes. To back up the SQL Server database: 1. Resynchronize the copy group. Execute EX_DRM_RESYNC on the database server to resynchronize the copy groups.
Restoring an SQL Server database to a primary volume This section provides an example of recovering an SQL Server database when the backup data is stored on the secondary volume. This example restores the data from the secondary volume to the primary volume by resynchronizing the volumes. The operation ID is Operation_A. Restoring an SQL Server database To restore the SQL Server database: 1. Check the backup ID of the backup you want to use for restoration.
Using the drmsqlrecover command to recover an SQL Server database The drmsqlrecover command is useful for recovering particular databases and recovering multiple volumes at once. To use the drmsqlrecover command to recover an SQL server database: • Specify an instance name to execute the drmsqlrecover command. The following command illustrates execution of the drmsqlrecover command to recover an SQL Server database: DBServer > drmsqlrecover default KAVX0001-I The drmsqlrecover command will now start.
To use the drmsqlrecovertool dialog box to recover an SQL Server database: 1. Specify an instance name to execute the drmsqlrecovertool command. DBServer > drmsqlrecovertool default When you execute the command, the drmsqlrecovertool dialog box is displayed. 2. Select a recovery destination database in Target database. 3. Check the status of the database after recovery. For Recovery mode, select the status of the recovered database.
System configuration This section uses examples to describe how to use extended commands. The example here assumes that the system is configured to use one database server, but you can also configure the system to use multiple database servers. Figure 85 System configuration for backing up an SQL Server database to tape and restoring it from tape Application Agent allows you to use a cluster configuration for database servers.
• The FTP service on the backup server is started, and the system is set up so a user who has logged onto the OS of the database server can log onto the FTP server and transfer files to the FTP server. The FTP user ID is admin, and the FTP user password is password. • Data for which an online back up was performed from the primary volume to the secondary volume was backed up to tape. • The SQL Server instance (INSTANCE_1) exists, and services are running on database server A (DBServer).
Figure 86 Procedure for restoring an SQL Server database to the primary volume Application Agent CLI User Guide 297
Figure 87 Procedure for restoring an SQL Server database from its backup tape Backing up an SQL Server database to tape This section provides an example of performing an online backup of an SQL Server database onto tape. In this example, an online backup of the INSTANCE_1 instance of the DBServer database server is first performed to place a backup on the secondary volume, and then a backup is performed from the secondary volume to tape. The operation ID is Operation_A.
Resynchronize the copy group. 1. For multi-generation operations, check the copy group to be used next. To check the copy group to be used next, execute the drmsqlcat command on the backup server. If the copy group is not locked, the least recently backed up copy group is overwritten. You can execute the drmcgctl command to check the lock status of a copy group. 2. Clear the cache of the secondary volume. Before backing up SQL Server, clear the system cache on the backup server.
3. Export the backup information to temporary files To back up data from the secondary volume to tape, copy the backup information to the backup server. To export the backup information to temporary files, execute EX_DRM_DB_EXPORT. These temporary files are stored in the temporary folder for extended commands, on the database server. DBServer > EX_DRM_DB_EXPORT Operation_A After EX_DRM_DB_EXPORT is executed, a operation-ID.
8. Back up the data of the secondary volume to tape. To back up the data of the secondary volume, execute EX_DRM_TAPE_BACKUP on the backup server. In this example, the drive letter of the secondary volume is assumed to be E: in this example. BKServer > EX_DRM_TAPE_BACKUP Operation_A -mount_pt E: Note that you can omit the -mount_pt option if the mount point defined in the configuration definition file for tape backup (such as NBU.DAT) matches a mount point that is actually free.
3. Export the backup information to temporary files. To restore data from the secondary volume to the primary volume, the backup information about the restore operation from tape to the secondary volume needs to be copied to the database server. To export the backup information to temporary files, execute EX_DRM_DB_EXPORT. These temporary files are stored in the temporary folder for extended commands. BKServer > EX_DRM_DB_EXPORT Operation_A 4.
9. Recover the SQL Server database. To recover the SQL Server database, execute the drmsqlrecover or drmsqlrecovertool command. The following shows an example of the drmsqlrecover command and drmsqlrecovertool dialog box. For details on the drmsqlrecover command or the drmsqlrecovertool dialog box, see the HP P9000 Replication Manager Software Application Agent CLI Reference Guide.
To use the drmsqlrecovertool dialog box to recover an SQL Server database: 1. Specify an instance name to execute the drmsqlrecovertool command. DBServer > drmsqlrecovertool default When you execute the command, the drmsqlrecovertool dialog box is displayed. 2. Select a recovery destination database in Target database. 3. Check the status of the database after recovery. For Recovery mode, select the status of the recovered database.
System configuration The example is based on the system configuration shown in the figure below. Figure 88 System configuration for backup of an SQL Server database to tape Processing flow In this example, the processing flow of the drmsqlbackup command that specifies a user script is as follows: • The cache of the secondary volume is cleared (this is a step in the user pre-processing section). • An online backup of SQL Server's DEFAULT instance is performed.
By not specifying the VDI metafile storage location in drmsqlinit, the VDI metafile is written to the primary volume and can be backed up at the same time as the data files. Figure 89 Processing flow User script example The table below shows the user script that is created for the example. Table 101 User script creation example Script text Explanation LOCAL_BACKUP=YES ...(1) (1) YES is required. #Pre-processing section [PRE_PROC] ...
Script text Explanation [CMD] ...(3) (3) Mounting of the first volume CMDLINE="C:\program files\CVXPAE\drm\bin \drmmount.exe" -copy_group vg01,vol01 LOCATION=REMOTE ...(4) (4) Execute on a remote server. [CMD] ...(5) CMDLINE="C:\program files\CVXPAE\drm\bin \drmmount.exe" -copy_group vg01,vol02 (5) Mounting of the second volume LOCATION=REMOTE [CMD] ...(6) CMDLINE="C:\program files\CVXPAE\drm\bin \drmumount.exe" -copy_group vg01,vol01 (6) Unmounting of the first volume LOCATION=REMOTE [CMD] ...
Script text Explanation #Tape backup of the secondary volume [CMD] CMDLINE=C:\tmp\tapebackup.bat TIMEOUT=NOWAIT ...(9) (9) Execute the next command without waiting for the command to terminate. END_CODE=TERMINATE_NZ ...(10) (10) Treat a command return value that is a value other than 0 as an error. LOCATION=REMOTE PARENT_STAT=NORMAL ...(11) (11) Execute only when the backup command is normal. The following is an example of backing up the secondary volume to tape using tapebackup.
Operation using SQL Server transaction logs Backing up an SQL Server database entails either backing up all databases, or backing up transaction log files. When formulating a plan for backing up an SQL Server database, if the database is fairly large so that the time required to perform backup is significant, or if the database is updated regularly, we recommend a combination of database backup and transaction log backup. There are several precautions to keep in mind when backing up transaction logs.
• The primary volume and the secondary volume are pair defined in the database server and the backup server. • The tape backup management software is installed on the backup server. • The parameters of the tape backup management software have been registered by using the drmtapeinit command. • The FTP service on the backup server is started, and the system is set up so a user who has logged onto the OS of the database server can log onto the FTP server and transfer files to the FTP server.
Procedures for restoration and recovery are as follows after failure occurs at point D. 1. Restore the data in the SQL Server database backed up at point A. The database is recovered to point A. For details about how to restore the SQL Server database backed up at point A, see “Restoring SQL Server databases” on page 313. Also, for the order to apply the transaction logs, see “Precautions regarding application of transaction logs” on page 283. 2.
3. Resynchronize the copy group. Execute EX_DRM_RESYNC on the database server to resynchronize the copy groups. If there is a big difference between the primary volume data and the secondary volume data when data is backed up from the primary volume to the secondary volume, the backup might take a while. You can facilitate backup processing by resynchronizing the copy group in advance. DBServer > EX_DRM_RESYNC Operation_A -cg_file C:\temp\CGLIST.
To back up the transaction logs: • Back up the transaction logs at point B (point C). To back up the transaction logs, execute EX_DRM_SQL_TLOG_BACKUP. DBServer > EX_DRM_SQL_TLOG_BACKUP Operation_A Restoring SQL Server databases Restore the SQL Server database backed up at "point A". To restore the SQL Server database: • Restore the SQL Server database. To restore the SQL Server database, execute EX_DRM_SQL_RESTORE.
4. Select Standby for Recovery mode and Yes for Roll forward?, and click the Recovery button. The transaction log obtained at point B and point C is applied, and the database is restored to the status at point C. At this point, the status of the database after recovery changes based on the selection for Recovery mode.
Example of operation using transaction log backup If a failure occurs, you can restore the database up to the point of the last backup by using the data and transaction logs backed up during regular operation before the failure occurred. However, when you use only backups under normal operation, since operations performed after the last backup is truncated, such operations cannot be restored.
• Obtain transaction logs, without truncating those subsequent to point C. To obtain transaction logs without truncating, execute the EX_DRM_SQL_TLOG_BACKUP with the -no_truncate option specified. DBServer > EX_DRM_SQL_TLOG_BACKUP Operation_A -no_truncate When the -no_truncate option is specified, transaction logs are obtained from after transaction log backup acquisition at point B, until failure occurred.
System configuration The examples in this section assume the system configuration shown in the following figure. Figure 93 Backing up SQL Server transaction log backup files The system prerequisites for this example are as follows: • A pair is defined for each of the volumes on which the E, F, and G drives are mounted. • SQL Server data files are placed on the E drive and SQL Server transaction log files are placed on the F drive.
FTP_HOME_DIR=C:\FTP_ROOT FTP_SUB_DIR=script SET_DRM_HOSTNAME=1 • The operation definition file used to back up the file system is named FS1. The backup target is the G drive that contains the transaction log backup file storage directory. Contents of _FS1.dat: BACKUP_OBJECT=FILESYSTEM DB_SERVER_NAME=FILESERV1 INSTANCE_NAME=G: TARGET_NAME= FTP_HOME_DIR=C:\FTP_ROOT FTP_SUB_DIR=script SET_DRM_HOSTNAME=1 Backing up the transaction log backup file 1. Back up the transaction logs.
4. Back up the backed up transaction log file to the secondary volume. This step backs up the transaction log file online. To do this, execute EX_DRM_FS_BACKUP. DBServer > EX_DRM_FS_BACKUP FS1 DBServer > When executes this command, places backup information with a new backup ID (the next unused ID) in the backup catalog on the database server. It is assumed here that this new backup ID is 0000000003. 5. Check the backup execution results.
10. Back up the data of the secondary volume to tape. To back up the data of the secondary volume, execute EX_DRM_TAPE_BACKUP on the backup server. Specify the G: drive as the mount point. BKServer > EX_DRM_TAPE_BACKUP FS1 -mount_pt G: BKServer > When this backup operation is performed, information on this operation is entered into the backup catalog, with a new backup ID 0000000004. 11. Check the backup to tape execution results.
4. Export the backup information to temporary files. To restore data from the secondary volume to the primary volume, the backup information about the restore operation from tape to the secondary volume needs to be copied to the database server. To export the backup information to temporary files in the temporary directory for extended commands, execute EX_DRM_DB_EXPORT. BKServer > EX_DRM_DB_EXPORT FS1 BKServer > 5. On the database server, receive the temporary files.
• Backups obtained for a volume are kept, while the usual backups are obtained for the other two volumes. Example of operation using lock functionality and generation management The following figure shows an example of operation in which a particular copy group is locked. Figure 94 Example of operation in which a particular copy group is locked 1. Back up the primary volume to secondary volume 1 (generation 1), and then lock copy group 1 (generation 1).
For the following, Operation_A is used as the operation ID. To lock the copy group: 1. Checking the secondary volume status. Display the list of copy groups and make sure that none are locked. To check the lock status of a copy group, execute the drmcgctl command with no parameters. If UNLOCKED is displayed in the LOCK STATUS column, the copy group is not locked. DBServer > COPY GROUP VG01,VOL11 VG01,VOL12 VG01,VOL13 VG02,VOL21 VG02,VOL22 VG02,VOL23 VG03,VOL31 VG03,VOL32 VG03,VOL33 DBServer > 2.
4. Lock secondary volume 1. To keep the backup data (generation 1), lock the copy group (generation 1) by executing the drmcgctl command with the -mode lock option specified. You can specify the item to be locked by backup ID or copy group. At this point, execute the drmcgctl command, specifying 0000000004 as the backup ID. DBServer > KAVX0001-I KAVX0302-I KAVX0002-I drmcgctl -backup_id 0000000004 -mode lock The drmcgctl command will now start.
2. Check the copy group status. Display the copy group list and check that the copy group has been unlocked.
Figure 95 System configuration for backing up and restoring an SQL Server database between a local site and a remote site The system prerequisites for this example are as follows: • The system includes a local-site database server (server name: DBServer) and a remote-site backup server (server name: BKServer) to which a tape device is attached.
• The SQL Server data files and transaction log files are stored in the same directory structures on the local and remote sites. • Use Operation_A as the operation ID. Backing up an SQL Server database to a remote site The example provided below explains the operation used to back up a file system to a remote site. Procedures at the local site and remote site are explained separately. Operation at the local site 1. Check the volume to be backed up and its generation identifier.
4. Export the backup information to temporary files. To perform backup from the secondary volume to tape, you need to copy information about the backup operations from the primary volume to the secondary volume, to the backup server. To export the information pertaining to the backup operations to temporary files, execute EX_DRM_DB_EXPORT. The temporary files are stored in the temporary directory for extended commands, on the database server. DBServer > EX_DRM_DB_EXPORT Operation_A 5.
4. Confirm that data has been correctly backed up to tape. To confirm that backup information with the date of the backup from the primary volume to the secondary volume exists, execute drmtapecat on the backup server. BKServer > drmtapecat BACKUP-ID BACKUP-OBJECT 0000000002 MSSQL BKServer > 5. SNAPSHOT TIME 2004/08/05 10:00:00 EXPIRATION TIME 2004/11/05 10:00:00 BACKUP-MEDIA MEDIA1 Clear the cache of the secondary volume.
3. Restore data from its backup tape to the secondary volume. To restore data, execute EX_DRM_TAPE_RESTORE on the backup server. BKServer > EX_DRM_TAPE_RESTORE Operation_A -backup_id 0000000002 When restore is performed, information about the restore operation is registered in the backup catalog on the backup server with new backup ID 0000000003. 4. Check that correct data is restored from the tape to the secondary volume.
3. Check that the import operation has been performed correctly. To confirm that backup information with the date of the backup from the primary volume to the secondary volume has been imported to the database server, execute the drmsqlcat command on the database server.
3. Restore data from its backup tape to the secondary volume. To restore data, execute EX_DRM_TAPE_RESTORE on the backup server. BKServer > EX_DRM_TAPE_RESTORE Operation_A -backup_id 0000000002 When restore is performed, information about the restore operation is registered in the backup catalog on the backup server with new backup ID 0000000003. 4. Check that correct data is restored from the tape to the secondary volume.
Operation at the local site 1. If SQL Server is running on the database server, stop SQL Server. 2. Use a disk management function to unmount the volume mounted by using the disk management function. Operation at the remote site 1. Stop SQL Server. 2. Use the disk management function to unmount the volume that was mounted using the disk management function. The G: drive is unmounted. 3. To create a backup catalog, create a pair the primary volume of which should reside in the local site.
Figure 96 System configuration for backing up an SQL Server database to tape and restoring it from tape in a multi-target configuration This example assumes the following conditions: • The primary volume is NTFS-formatted. • Both of the primary volumes in the local site and in the remote site are defined as a pair by Continuous Access Software. In each site, the primary volume and the secondary volume are also defined as a pair by Business Copy.
• The FTP service on the backup server is started, and the system is set up so a user who has logged onto the OS of the database server can log onto the FTP server and transfer files to the FTP server. • Data for which an online back up was performed from the primary volume to the secondary volume was backed up to tape. • Two SQL Server instances (INSTANCE_1 and INSTANCE_2) exist, and services are running on database server (hostB). • The secondary volume is not always mounted.
END_CODE=TERMINATE_NZ TIMEOUT=0 [SPLIT_PROC] # Target pair volume: SI01,dev01 [CMD] CMDLINE=C:\HORCM\etc\pairsplit.exe -g TC01 -d sql01 -FBC 0 ENV=HORCMINST=0 ENV=HORCC_MRCF= END_CODE=TERMINATE_NZ TIMEOUT=10 [CMD] CMDLINE=C:\HORCM\etc\pairevtwait.
4. Export the backup information at the local site to temporary files. To back up data from the secondary volume to tape, copy the backup information to the backup server. To export the backup information to temporary files, execute EX_DRM_DB_EXPORT. These temporary files are stored in the temporary directory for extended commands, on the database server. hostB > EX_DRM_DB_EXPORT opid_SQL 5. Transmit the temporary files to the backup server at the local site.
2. Restore data from its backup tape to the secondary volume. To restore data, execute EX_DRM_TAPE_RESTORE on the backup server at the local site. hostA > EX_DRM_TAPE_RESTORE opid_SQL -backup_id 0000000002 3. Export the backup information to temporary files. To restore data from the secondary volume to the primary volume, the backup information about the restore operation from tape to the secondary volume needs to be copied to the database server.
7. Recover the SQL Server database. To recover the SQL Server database, execute the drmsqlrecover or drmsqlrecovertool command on the database server at the local site. The following command illustrates execution of the drmsqlrecover command to recover the SQL Server database: hostB > drmsqlrecover DEFAULT KAVX0001-I The drmsqlrecover command will now start. KAVX0002-I The drmsqlrecover command will now end.
Figure 97 System configuration for restoring and recovering from a host not used for backup (when backing up) 340 Operation for SQL Server databases
Figure 98 System configuration for restoring and recovering from a host not used for backup (when restoring) The following conditions are assumed: • The primary volumes are formatted with NTFS, and are mounted on the D and E drives. Both volumes are used for the SQL Server database. A volume used for storing data files cannot also be used to store logs. • The primary and secondary volumes are mirrored (as set by the volume replication function and RAID Manager).
1. Back up the SQL Server database to the secondary volumes. To perform an online backup of the SQL Server database, execute the drmsqlbackup command. PROMPT> drmsqlbackup SQL1 KAVX0001-I The drmsqlbackup command will now start. KAVX1017-I SQL Server information is being collected now. KAVX1207-I Backing up the database... KAVX0040-I The backup ID has been assigned as follows: Backup ID = 0000000001 KAVX0002-I The drmsqlbackup command will now end.
4. Transfer the exported temporary file from database server 1 to database server 2. Use the ftp (file transfer protocol) command to transfer the exported temporary file C:\temp\ EX-FILE1.drm from database server 1 to database server 2. PROMPT> ftp database-server2-name ftp> Username: (Enter the login name) ftp> password: (Enter the password) ftp> binary ftp> put C:\temp\EX-FILE1.drm ftp> quit PROMPT> 5.
7. Restore the data from the secondary volume to the primary volume. Restore the data by resynchronizing the primary and secondary volumes. To restore the data, execute the drmsqlrestore command on database server 2. PROMPT> drmsqlrestore 0000000001 -resync -nochk_host KAVX0001-I The drmsqlrestore command will now start. KAVX1316-I Performing restore pre-processing... KAVX1000-I Logged on to the SQL Server system. KAVX1300-I Restore processing will now start. KAVX1000-I Logged on to the SQL Server system.
Figure 99 System configuration for log-shipping operations The following conditions are assumed:: • The primary volumes on the source server and the volumes on the distribution target server are formatted with NTFS, and are mounted on the D and E drives, respectively. • The primary and secondary volumes on the source server are mirrored (as set by the volume replication function and RAID Manager). • The secondary volumes on the source server are formatted with NTFS, and are not mounted.
1. Back up the SQL Server database to the secondary volumes. To perform an online backup of the SQL Server database, execute the drmsqlbackup command on the source server. PROMPT> drmsqlbackup SQL1 KAVX0001-I The drmsqlbackup command will now start. KAVX1017-I SQL Server information is being collected now. KAVX1207-I Backing up the database... KAVX0040-I The backup ID has been assigned as follows: Backup ID = 0000000001 KAVX0002-I The drmsqlbackup command will now end.
4. Transfer the metafile and the temporary file from the source server to the distribution target server. Use the ftp (file transfer protocol) command to transfer the metafile C:\METADIR\Meta1 and the exported temporary file C:\temp\EX-FILE1.drm from the source server to the distribution target server. PROMPT> ftp distribution-destination-server-name ftp> Username: (Enter the login name) ftp> password: (Enter the password) ftp> binary ftp> put C:\temp\EX-FILE1.
8. Restore the backup data to the distribution target server volumes. Check that the backup data has been copied to the distribution target server volumes, and then restore the backup data from the distribution target server. When restoring, apply the metafile to the backup data without recovering the backup data from the secondary volumes to the primary volumes, and place the database into standby status.
Figure 100 System configuration for setting a distribution target server to be an operating server The following conditions are assumed: • The distribution target server volumes are formatted with NTFS, and are mounted on the D and E drives. • The drmsqlinit command has been executed on the distribution target server. • The system databases (master, model, and msdb) are not specified when restoring. • The database on the distribution target server is in the standby status.
2. Place the standby database online. To use a Application Agent command after restoring the database, execute the drmsqldisplay command with the -refresh option. Backing up and restoring in a cascade configuration or multi-target configuration Application Agent supports backups to and restorations from a secondary volume at a remote site (in a cascade configuration).
• The primary volume at the local site and the primary volume at the remote site must be defined as a pair in Continuous Access Software. The primary and secondary volumes at each site must be defined as a pair in Business Copy. • The same RAID Manager instance number is defined at both the local and remote sites. • The SQL Server data files and transaction log files must have the same directory structure at both the local and remote sites.
4. After the pair status of the Continuous Access Software pair volume is changed to PAIR, perform takeover. To perform takeover, execute the RAID Manager command at the remote site. PROMPT> paircreate -g TC01 -vr ( or pairresync -g TC01 ) PROMPT> pairevtwait -g TC01 -t 600 -s pair PROMPT> horctakeover -g TC01 5. Mount the Continuous Access Software pair volume at the remote site and restart SQL Server. The mount point name to mount must be same as the mount point name at the local site.
12. Create a user script file. To obtain the information required to create a user script file, execute the drmsqldisplay command on the database server at the local site. For details on how to create a user script file, see “Creating a user script to perform a backup in a cascade configuration” on page 355. PROMPT> drmsqldisplay default -remote INSTANCE: default T DB OBJECT FILE TID# LUN# MU# LDEV# P/S SERIAL# D UserDB1 DATAFILE Z:\UserDB1_Data.
18. Create a template catalog at the remote site. To create a template catalog, execute the drmsqlbackup command on the database server at the remote site. The drmsqlbackup command does not perform backup when it is used to create a template catalog. PROMPT> drmsqlbackup default -template KAVX0001-I The drmsqlbackup command will now start. KAVX1017-I SQL Server information is being collected now. KAVX0085-I Now making backup information...
Creating a user script to perform a backup in a cascade configuration To perform backup in a cascade configuration, you need to create a user script file.
• If you set TIMEOUT to 0, the drmsqlbackup command does not finish unless the command executed by the user script file finishes. Keep this in mind when you specify the timeout setting. • When you want to stop an executing command (for example, when a command executed by a user script file does not respond), enter Ctrl+C from the keyboard, or use the Windows Task Manager to terminate the command processing. (The command processing is executed by the user script file, and not by the drmsqlbackup command.
- C D UserDB P - - - - C T UserDB P - - - - C PROMPT> - 55555 DATAFILE 55014 - - - - 55555 TRANSACT 55014 - - - - 55555 TC01,sql01 P:\UserDB_Data.NDF - SI01,sql02 SI02,sql02 SI03,sql02 TC01,sql02 O:\UserDB_Log.
Figure 104 Flow of script-specified operations performed when local backup is not involved Table 102 Sample script for when a local backup is not performed Script code LOCAL_BACKUP=NO ...(1) Description (1) Specifies that the database should not be backed up to the local site secondary volume (backed up only to the remote site secondary volume). [RESYNC_PROC] ...
Script code Description CMDLINE=C:\HORCM\etc\pairresync.exe -g TC01 -FBC 0 ...(3) (3) Batch resynchronization with a pair group specified ENV=HORCC_MRCF= ...(4) (4) Specifies a remote copy. ENV=HORCMINST=0 ...(5) (5) Specifies an instance number. END_CODE=TERMINATE_NZ ...(6) TIMEOUT=10 ...(7) [CMD] CMDLINE=C:\HORCM\etc\pairevtwait.exe -g TC01 -d sql01 -s pair -t 600 -FBC 0 ...(8) ENV=HORCMINST=0 ...(9) END_CODE=TERMINATE_NZ ...
Script code TIMEOUT=0 ...(15) Description (15) Waits for command termination. [SPLIT_PROC] ...(16) # Target pair volumes: SI11 dev group (16) User processing for pair splitting. [CMD] CMDLINE=C:\HORCM\etc\pairsplit.exe -g TC01 -FBC 0 ...(17) (17) Batch splitting with a pair group specified ENV=HORCC_MRCF= ...(18) (18) Specifies a remote copy. ENV=HORCMINST=0 ...(19) (19) Specifies an instance number. END_CODE=TERMINATE_NZ ...
Script code Description CMDLINE=C:\HORCM\etc\pairevtwait.exe -g TC01 -d sql02 -s psus -t 600 -FBC 0 ...(26) ENV=HORCMINST=0 ...(27) END_CODE=TERMINATE_NZ ...(28) TIMEOUT=0 ...(29) (26) Checks the sql02 pair status. Waits until the sql02 status becomes PSUS. (27) Specifies an instance number. (28) If a non-zero return value returns, the script will terminate with an error. (29) Waits for command termination. [FINISH_PROC] ...
Figure 105 Flow of script-specified operations performed when local backup is involved Table 103 Sample script for when a local backup is performed Script code LOCAL_BACKUP=YES ...(1) Description (1) Specifies that the database should also be backed up at the local site secondary volume (backup in a multi-target configuration). [RESYNC_PROC] ...
Script code Description CMDLINE=C:\HORCM\etc\pairresync.exe -g TC01 -FBC 0 ...(3) (3) Batch resynchronization with a pair group specified ENV=HORCC_MRCF= ...(4) (4) Specifies a remote copy. ENV=HORCMINST=0 ...(5) (5) Specifies an instance number. END_CODE=TERMINATE_NZ ...(6) TIMEOUT=10 ...(7) [CMD] CMDLINE=C:\HORCM\etc\pairevtwait.exe -g TC01 -d sql01 -s pair -t 600 -FBC 0 ...(8) ENV=HORCMINST=0 ...(9) END_CODE=TERMINATE_NZ ...
Script code TIMEOUT=0 ...(15) Description (15) Waits for command termination. [SPLIT_PROC] ...(16) # Target pair volumes: SI11 dev group (16) User processing for pair splitting. [CMD] CMDLINE=C:\HORCM\etc\pairsplit.exe -g TC01 -FBC 0 ...(17) (17) Batch splitting with a pair group specified ENV=HORCC_MRCF= ...(18) (18) Specifies a remote copy. ENV=HORCMINST=0 ...(19) (19) Specifies an instance number. END_CODE=TERMINATE_NZ ...
Script code Description (26) Checks the sql02 pair status. CMDLINE=C:\HORCM\etc\pairevtwait.exe -g TC01 -d sql02 -s psus -t 600 -FBC 0 ...(26) Waits until the sql02 status becomes PSUS. (27) Specifies an instance number. ENV=HORCMINST=0 ...(27) (28) If a non-zero return value returns, the script will terminate with an error. END_CODE=TERMINATE_NZ ...(28) (29) Waits for command termination. TIMEOUT=0 ...(29) [FINISH_PROC] ...
1. Prepare the user script file. This sample procedure uses the following user script file: LOCAL_BACKUP=NO [RESYNC_PROC] [CMD] CMDLINE=C:\HORCM\etc\pairresync.exe -g TC01 -d sql01 -FBC 0 ENV=HORCMINST=0 END_CODE=TERMINATE_NZ TIMEOUT=0 [SPLIT_PROC] [CMD] CMDLINE=C:\HORCM\etc\pairsplit.exe -g TC01 -d sql01 -FBC 0 ENV=HORCMINST=0 END_CODE=TERMINATE_NZ TIMEOUT=0 [FINISH_PROC] 2. Back up the SQL Server database. PROMPT> drmsqlbackup default -script C:\Uscrip.
4. Export the backup catalog to a temporary file. Export the backup catalog to a temporary file. To export the backup catalog to the temporary file C:\tmp\EX-FILE1.drm, execute the drmdbexport command. This step must be performed on the database server at the remote site. PROMPT> drmdbexport 0000000001 -f C:\tmp\EX-FILE1.drm KAVX0001-I The drmdbexport command will now start. KAVX0002-I The drmdbexport command will now end. 5.
7. Check the import results. Confirm that the import finished successfully. To confirm this, execute the drmsqlcat command on the backup server at the remote site.
10. Unmount the secondary volume. To unmount the secondary volume, specify the backup ID used when mounting 0000000002. This step must be performed on the backup server at the remote site. PROMPT> drmumount 0000000002 KAVX0001-I The drmumount command will now start. KAVX0406-I Unmounting will now start. Mount point = E: KAVX0407-I Unmounting has finished. Mount point = E: 11. Check the backup to tape execution results. Confirm that backup to tape finished successfully.
3. Perform takeover. If the local site primary volume can communicate with the remote site secondary volume (for example when the current and standby systems are intentionally switched), use the horctakeover command to perform takeover. This step must be performed at the remote site. PROMPT> horctakeover -g TC01 -d sql01 4. Split the Continuous Access Software pair (usual operation is in the pair status).
7. Unmount the secondary volume. To unmount the secondary volume, specify the backup ID that was used when mounting 0000000002. This step must be performed on the backup server at the remote site. PROMPT> drmumount 0000000002 KAVX0001-I The drmumount command will now start. KAVX0406-I Unmounting will now start. Mount point = E: KAVX0407-I Unmounting has finished. Mount point = E: KAVX0002-I The drmumount command will now end. PROMPT> 8.
10. From the database server, obtain the temporary file that was exported on the backup server. Obtain the temporary file C:\FTP_ROOT\EX-FILE1.drm, which was exported on the backup server, from the database server at the remote site. Use the ftp (file transfer protocol) command to obtain the temporary file C:\FTP_ROOT\EX-FILE1.drm (created in step 9) to the database server. The temporary file will be stored in the current directory. The following assumes that the current directory is C:\.
15. Restore the data from the secondary volume to the primary volume. Restore the data by resynchronizing the primary and secondary volumes. To restore the data, execute the drmsqlrestore command on the database server at the remote site. If you want to execute the command with the template catalog backup ID specified, specify the -template option. PROMPT> drmsqlrestore 0000000003 -resync -template KAVX0001-I The drmsqlrestore command will now start. KAVX1316-I Performing restore pre-processing...
21. Perform takeover at the local site to change the local site to the primary volume. Execute the horctakeover command at the local site. PROMPT> horctakeover -g TC01 -d sql01 22. At the local site, mount the primary volume and start the SQL Server service. When performing this step, note the following point. A transaction log backup file is associated with the backup catalog that was created for the last complete backup of the SQL Server database.
10. Check the storage location of the data files and log files. To check the location, execute the drmmount command using the backup ID checked in step 9 as the argument. The command results show the mounted drive name. The mounted drive stores data files and transaction log files for the SQL Server database. If necessary, execute the drmsqlcat command to check the contents of the file to be backed up. 11. Check the storage location of the metafile.
Figure 106 System configuration using file sharing to back up and restore an SQL Server database Preparing to back up or restore by using a file share This subsection describes the preparations for backing up or restoring an SQL Server database by using file share. To prepare for backing up or restoring an SQL Server database by using a file share: 1. Check the temporary directory for extended commands that was created automatically by the check tool for operation definition files.
3. Use Explorer on the database server to allocate the temporary directory for extended commands as a network drive. In this case, C:\FTP_ROOT\script\Operation_A\BK is allocated to network drive X:. 4. On the backup server, set the permissions for the automatically generated temporary directory for extended commands, so that users executing extended commands from the database server can read and write to the directory.
To restore an SQL Server database on the backup server by using a file share: 1. Execute the drmtapecat command to check the backup ID of the backup data. 2. Execute EX_DRM_TAPE_RESTORE to restore the backed up data from tape to the secondary volume. 3. Execute EX_DRM_DB_EXPORT to export the backup information to a temporary file. 4. Execute EX_DRM_SQLFILE_PACK to back up the SQL Server metafile created at the restoration to the temporary folder. 5.
Figure 107 System configuration for restoring data to an SQL Server instance different from the backup source For example, if only the user database backed up from the SQL Server instance instA of database server 1 is restored and recovered to SQL Server instance instB of database server 2, the user database from database server 2 becomes available. System databases such as master, model, and msdb cannot be restored to a different SQL Server instance.
Additionally, execute the drmsqlinit command on database server 2 to create a database configuration definition file. In this case, do not specify the VDI metafile storage directory. Operations for restoration (recovery) To check the backup ID of the backup data before performing restoration, specify the name of the backed up SQL Server instance name isntA as an argument of the drmsqlcat command. At the time of restoration, the database is restored to SQL Server instance instB on database server 2.
To restore (recover) backup data from tape to the secondary volume: 1. Execute EX_DRM_TAPE_RESTORE to restore the SQL Server database from tape to the secondary volume. 2. Execute the drmsqlcat command with instA specified to check the backup ID. 3. Execute the drmsqlrestore command with instB specified as the -instance argument to restore the SQL Server database on the secondary volume. 4.
• Execute the drmresync command, specifying the backup ID of the data to be deleted. PROMPT> drmresync -backup_id backup-ID KAVX0001-I The drmresync command will now start. KAVX0002-I The drmresync command will now end. PROMPT> When you execute this command, the primary volume and the secondary volume are resynchronized and mirrored. Now the backup information of the specified backup ID is deleted from the backup catalog.
Figure 108 System configuration for restoring to an SQL Server instance that is different from the backup source (using a user script) The following system prerequisites apply to this example: Server configuration • The same version of SQL Server version is installed in Database Server 1 and Database Server 2. • The database file path configurations are the same in Database Server 1 and Database Server 2.
• The copy groups are composed of the following two generations: Generation 1 (VG01, VOL11 and VG01, VOL12): Used for data mining Generation 2 (VG02, VOL21 and VG02, VOL22): Used for backup User script example The following two user scripts are used in this example: script1.txt (used for data mining) LOCAL_BACKUP=YES #Pre-backup user script section [PRE_PROC] #The Protection Manager service will be started. [CMD] CMDLINE=C:\user\detach_databases.
[POST_PROC]: Executes the remote_sqlrestore.bat batch file. The contents of remote_sqlrestore.bat mount the secondary volume and execute the restore command with the -no_resync specification and the recovery command. • remote_sqlrestore.
@echo off @bpbackup -k "%DRMENV_COMMENT%" -p ptm_plc -s ptm_schd -i -S BackupServer -w -t 0 -L C:\temp \nbu.log if NOT "%ERRORLEVEL%"=="0" GOTO ABEND @drmumount %DRMENV_R_BACKUPID% if NOT "%ERRORLEVEL%"=="0" GOTO ABEND GOTO END :ABEND exit 1 :END exit 0 Example This section shows the examples used for data mining and backup. • Execute the drmsqlbackup command on the DBServer1 server to make a backup (generation 1) on the secondary volume on the DBServer2 server.
Figure 109 Configuration for using the SQL Server replication function Requirements for the publication database and the distribution database The publication database and the distribution database used in the processing of a single replication must be allocated to satisfy the following requirements: • Because backups of these two databases must be made at the same time, they must be allocated to the same SQL Server instance.
Figure 110 Configuration of backup server for the publisher/distributor server Requirements for the subscription database Because the subscription database must be restored separately from the system databases (master, model, and msdb), it must be allocated to a different volume from the system databases.
PROMPT> drmsqlbackup instA -target PubUserDB1,distribution PROMPT> • Back up the subscription database Execute the drmsqlbackup command on the subscriber server. PROMPT> drmsqlbackup instC PROMPT> • Back up the databases, including the system databases Execute the drmsqlbackup command with the -system option specified in the publisher/distributor server or the subscriber server.
Restore the databases • Restore the publication database Specify the backup ID used when the publication database and the distribution database were backed up, and execute the drmsqlrestore command on the publisher/distributor server.
Specify the backup ID used when the databases, including the system databases, were backed up, and execute the drmsqlrestore command on the publisher/distributor server or the subscriber server. PROMPT> drmsqlrestore 0000000005 -resync -target master,model,msdb PROMPT> • Restore the subscription database On the subscriber server, execute the drmsqlrestore command.
sp_replicationdboption 'database-name', 'sync with backup', 'true' For the database-name argument, specify the name of the publication database or distribution database. How to check the 'sync with backup' option To ensure that the option has been set, execute the following SQL statement: select databasepropertyex('database-name', 'IsSyncWithBackup') For the database-name in the SQL statement, specify the name of the publication database or distribution database.
Figure 111 Operation when using Application Agent on an SQL Server in an AlwaysOn availability group configuration The following conditions must be satisfied beforehand. • An SQL Server system database must be installed on the active server, standby server, and secondary-use server. • The SQL Server parameters must be registered by executing the drmsqlinit command on the active server, standby server, and secondary-use server.
1. Back up the user database. On the database server (active server), execute the drmsqlbackup command with the -target or -f option specified. To execute the drmsqlbackup command with the -target option specified: PROMPT> drmsqlbackup instance-name -target user-database-name To execute the drmsqlbackup command with the -f option specified: PROMPT> drmsqlbackup instance-name -f target-definition-file-name 2. Back up the user database transaction logs.
1. Back up the user database to the secondary volume. On the database server (active server), execute the drmsqlbackup command with the -target or -f option specified.
7. Confirm that the backup information file has been imported correctly. On the database server (standby server), execute the drmsqlcat command. PROMPT> drmsqlcat instance-name 8. Restore the user database. On the database server (standby server), execute the drmsqlrestore command. PROMPT> drmsqlrestore backup-ID -no_resync -nochk_host -target user-database-name 9. Recover the user database. On the database server (standby server), execute the drmsqlrecovertool command.
4. Back up the user database to the secondary volume. On the database server (active server), execute the drmsqlbackup command with the -target or -f option specified.
10. Confirm that the backup information file has been imported correctly. On the secondary-use server (active server), execute the drmsqlcat command. PROMPT> drmsqlcat instance-name 11. Restore the user database. On the secondary-use server (active server), execute the drmsqlrestore command. PROMPT> drmsqlrestore backup-ID -no_resync -nochk_host -target user-database-name 12. Recover the user database. On the secondary-use server (active server), execute the drmsqlrecovertool command.
7 Operation for Exchange databases This chapter explains the procedures for running Application Agent when backing up Exchange databases. The descriptions in this chapter apply to the Application Agent CLI. For details on setting up commands, see the HP P9000 Replication Manager Software Application Agent CLI Reference Guide. The term storage group refers to a database management unit created on Exchange Server.
Event source: ftdisk Event ID: 57 or Event type: warning Event source: disk Event ID: 51 This Windows event information is logged during a VSS backup when the secondary volume becomes temporarily hidden. However, this will not influence the backup operation.
transactions log files, a roll-forward will be completed when the drmexgrestore command is executed, but the Information Store will not be mounted. In this case, the KAVX1510-E message is output. For details on this message, contact your HP support representative for assistance. Notes on restoring data obtained with a VSS backup • While restoring data obtained in a VSS backup, you must not open the physical disk resource properties on the cluster administrator window.
Precautions on recovery storage groups and recovery databases If you are using Exchange Server 2003 or Exchange Server 2007 as the database management system, do not specify a recovery storage group when specifying a storage group. If you are using Exchange Server 2010 as the database management system, do not specify a recovery database when specifying a database.
Figure 112 System configuration for backing up and restoring data between volumes without using VSS (when an Exchange database is backed up) Processing flow The processing flow for backing up storage groups to a secondary volume and restoring them to the primary volume on a single-server system is shown below. Execute commands on the database server.
Figure 113 Processing flow for backing up storage groups to a secondary volume Figure 114 Processing flow for restoring storage groups on the primary volume Backing up storage groups to a secondary volume This section describes the processing flow for backing up storage groups to a secondary volume on a single-server system. To back up a storage group: 1. Resynchronize the copy groups. Execute EX_DRM_RESYNC on the database server to resynchronize the copy groups.
Restoring storage groups to a primary volume This section explains how to restore a storage group when backup data has been saved on a secondary volume. This example restores the data from the secondary volume to the primary volume by resynchronizing the volumes. To restore a storage group: 1. Check the backup ID of the backup data. To check the backup ID of the backup data used to restore it from a secondary volume to the primary volume, execute the drmexgcat command on the database server.
Figure 115 System configuration for backing up and restoring data between volumes by using VSS (when an Exchange database is backed up) Processing flow The processing flow for backing up storage groups to a secondary volume and restoring them to the primary volume on a single-server system is shown below. Execute commands on the database server.
Figure 116 Processing flow for backing up storage groups to a secondary volume Figure 117 Processing flow for restoring storage groups on the primary volume Backing up storage groups to a secondary volume This section describes the processing flow for backing up a storage group to a secondary volume on a single-server system. To back up a storage group: 1. Resynchronize the copy groups. Execute EX_DRM_RESYNC on the database server to resynchronize the copy groups.
Restoring storage groups to a primary volume This section explains how to restore a storage group when backup data has been saved on a secondary volume. This example restores the data from the secondary volume to the primary volume by resynchronizing the volumes. To restore a storage group: 1. Check the backup ID of the backup data. To check the backup ID that will be used to restore backup data from a secondary volume to the primary volume, execute the drmexgcat command on the database server.
Figure 118 System configuration for backing up and restoring storage groups to and from tape With Application Agent, database servers can be clustered. With database servers configured in a failover cluster, if the active server fails, a standby server is available to take over operations from the active server. The examples in this section assume the following conditions: • The primary volume is NTFS-formatted.
• The primary volume configured as a pair with the secondary volume to be backed up will not be backed up until the secondary volume is backed up onto tape. • The cscript //H:Cscript command is executed from the command prompt and the host parameters are updated. Processing flow This section describes the processing flow for backing up storage groups to tape and restoring them on a multi-server system. Use extended commands to execute cold backups and backups with VSS.
Figure 119 Processing flow for backing up storage groups to tape Application Agent CLI User Guide 411
Figure 120 Processing flow for restoring storage groups from tape Backing up storage groups to tape Resynchronizing copy groups When usual operation is in the split state, resynchronize the copy group before backing up the data. When usual operation is in the pair status, there is no need to resynchronize the copy group before backing up data. Instead, when the tape backup is complete, resynchronize the copy group to initialize it. To resynchronize a copy group: 1. Clear the cache of the secondary volume.
Cold backup An example of backing up storage groups to tape in a cold backup is described below. In a cold backup, the Information Store to be backed up is unmounted and remains inaccessible until the backup completes. This is an example of performing a cold backup onto tape of an Exchange database. In this example, an cold backup of storage group STORAGE_1 for the database server DBServer is performed, and the backup is placed on the secondary volume.
5. Import the temporary files, which were transmitted from the database server, into the backup catalog on the backup server. To import the temporary files, which were transmitted from the database server, into the backup catalog on the backup server, execute EX_DRM_DB_IMPORT on the backup server. BKServer > EX_DRM_DB_IMPORT Operation_A 6. Back up the data of the secondary volume to a tape. To back up the data, execute EX_DRM_TAPE_BACKUP on the backup server.
3. Export the backup information to temporary files. To back up from the secondary volume onto tape, you must copy onto the backup server all information relating to the backup operation from the primary volume to the secondary volume. To export into temporary files the information relating to the backup operation, execute EX_DRM_DB_EXPORT. The temporary files are stored in the temporary folder for extended commands, on the database server. DBServer > EX_DRM_DB_EXPORT Operation_A 4.
1. Back up the storage group to the secondary volume. Using VSS, back up the storage group in an online backup operation. To perform backup, execute EX_DRM_EXG_BACKUP. Specify the operation ID Operation_A as the argument. DBServer > EX_DRM_EXG_BACKUP Operation_A -mode -vss 2. Check that the storage group has been backed up correctly.
6. Back up the data of the secondary volume to a tape. To back up the data, execute EX_DRM_TAPE_BACKUP on the backup server. In this example, the drive containing the secondary volume is E:. BKServer > EX_DRM_TAPE_BACKUP Operation_A -mount_pt E: When you execute backup, information about this backup is registered in the backup catalog with the new backup ID 0000000002.
4. Receive the temporary files on the database server. To batch-receive the temporary files held on the backup server onto the database server, execute EX_DRM_FTP_GET on the database server. In this example, the user ID used for logon to the FTP server is admin and the password is password. Temporary files are stored in the temporary folder for extended commands, on the database server. DBServer > EX_DRM_FTP_GET Operation_A -server BKServer -user admin -password password 5.
Backing up a storage group by using a user script By using a backup command that specifies a user script, you can use a single operation to back up a storage group to tape from the primary volume via the secondary volume. The explanations in this section use operation examples for Exchange Server 2007. System configuration The example is based on the system configuration shown in the figure below.
Figure 122 Processing flow User script example The table below shows the user script that is created for the example. Table 104 User script creation example Script text Explanation LOCAL_BACKUP=YES ...(1) #Pre-processing section (1) YES is required. #None #Post-processing section [POST_PROC] ...
Script text Explanation [CMD] CMDLINE=C:\tmp\tapebackup.bat TIMEOUT=NOWAIT ...(3) (3) Execute the next command without waiting for the command to terminate. END_CODE=TERMINATE_NZ ...(4) (4) Treat a command return value that is a value other than 0 as an error. LOCATION=REMOTE ...(5) (5) Execute on the remote server. The -s option is required for the backup command. PARENT_STAT=NORMAL ...(6) (6) Execute only when the backup command is normal.
rolled forward by specifying the -recovery option. In this case, by applying the transaction logs after backup, the database can be recovered to the state it was in immediately before the error occurred. The procedure for using the transaction logs of Exchange Server to restore backup data by using roll-forward is described below. The explanations in this section use operation examples for Exchange Server 2007. The figure below shows a flow of operation using transaction logs.
This example describes the procedure for restoring only the data in the mailbox store (Information Store) out of the backup data obtained by VSS from the secondary volume to the primary volume, and rolling forward transaction logs. (If restoration is executed in units of Information Stores, you must roll forward transaction logs at the time of restoration.
To perform restoration of data from a VSS backup, the Protection Manager service must be running on the backup server. To roll forward transaction logs at the time of restoration: 1. Check the backup ID and Information Store name of the backup data. To check the backup ID and Information Store name, execute the drmexgcat command. The backup ID is displayed in BACKUP-ID and the Information Store name is displayed in INFORMATIONSTORE.
The following examples are provided: • Backing up storage groups to a remote site. • Restoring the storage groups backed up in the remote site to the local site (if the local site is operating properly) The examples in this section assume the system configuration shown in the following figure.
• A storage group STR1 exists on the database server (DBServer) in the local site, and services are running. • The volume (usual operation is in the pair status) TC01,EXG11 for Continuous Access Software or Continuous Access Journal contains the data files for the storage group STR1. • The volume (usual operation is in the pair status) TC02,EXG12 for Continuous Access Software or Continuous Access Journal contains the transaction log files and checkpoint files for the storage group STR1.
2. Back up the database to the secondary volume in the remote site. To back up the database to the secondary volume in the remote site, specify the generation identifier for the back up destination remote_0 on the database server, and then execute EX_DRM_EXG_BACKUP. DBServer > EX_DRM_EXG_BACKUP Operation_A -rc remote_0 3. Check that the storage group has been backed correctly.
Operations at the remote site 1. Import the temporary files, which were transmitted from the database server in the local site, into the backup catalog on the backup server. To import the temporary files, execute EX_DRM_DB_IMPORT on the backup server. BKServer > EX_DRM_DB_IMPORT Operation_A 2. Check that the temporary file has been imported correctly.
Operations at the local site • Resynchronize the copy groups. To resynchronize the copy groups, execute EX_DRM_RESYNC on the database server. DBServer > EX_DRM_RESYNC Operation_A Restoring a storage group to a local site This section explains how to restore a storage group to a local site when the local site is operating normally without any failures. The procedures for restoring the data to a local site and to a remote site are explained separately.
4. Check that the storage group has been stored from the tape to the second volume correctly. To check that backup information bearing the date on which the storage group had been backed up from the primary volume to the second volume has been restored, execute the drmexgcat command on the backup server.
3. Check that the temporary file has been imported correctly. To check that backup information bearing the date on which the storage groups had been backed up from the primary volume to the secondary volume has been imported into the database server, execute the drmexgcat command on the database server.
Figure 126 Example of a system for backing up and restoring data in a cascade configuration The examples in this section assume the following conditions: • The Protection Manager service is running on the backup server at the local site and the database server and backup server at the remote site. • Backup of another primary volume cannot be performed until the secondary volumes are backed up to the tape devices. • Initially, Exchange virtual server is running on database server A at the local site.
• Continuous Access Synchronous is configured between disks P1 and P3, and between disks P2 and P4. • Using the RAID Manager configuration definition file, the pair definitions on the database server and backup server have been completed for the primary and secondary volumes.
KAVX0256-I A connection to the Protection Manager service will now be established. Host name = name-of-backup-server-B KAVX0040-I The backup ID has been assigned as follows: Backup ID = 0000000011 KAVX1601-I Backup processing has finished. KAVX5156-I The backup catalog will now be exported. KAVX5157-I The backup catalog was exported. KAVX5158-I The backup catalog will now be imported. Host name = name-of-backup-server-B KAVX5159-I The backup catalog has been imported.
Operation at both the local and remote sites 1. Confirm that the backups were performed correctly. Execute the drmexgcat command on database server A to check the backup catalog information: DBServer A > drmexgcat STRAGEGROUP: STR1 BACKUP-ID:0000000001 BACKUP-MODE: VSS STRAGEGROUP: STR1 ORIGINAL-ID:0000000001 START-TIME:2007/11/01 10:00:00 END-TIME: 2007/11/01 10:05:00 HOSTNAME:EXG_SRV1 T OBJECT INFORMATIONSTORE FILE FS DG COPY-GROUP M MAILBOXSTORE MAILBOX1(STR1) D:\STR1\priv1.
2. Confirm that the files were imported correctly.
3. Back up the backup data on the secondary volumes to tape. Execute the drmmediabackup command on backup server A to back up the backup data on the secondary volume to tape: BKServer A > drmmediabackup 0000000001 KAVX0001-I The drmmediabackup command will now start. KAVX0402-I Backup processing will now start. To see more details on backup progress, use the monitoring method of the backup management product. Source = W: KAVX0402-I Backup processing will now start.
4. Confirm that the data was correctly backed up to tape.
5. Unmount the secondary volumes on which the backup data is stored. Execute the drmumount command on backup server A to unmount the secondary volume: BKServer A > drmumount 0000000001 KAVX0001-I The drmumount command will now start. KAVX0406-I Unmounting will now start. Mount point = W: KAVX0407-I Unmounting has finished. Mount point = W: KAVX0406-I Unmounting will now start. Mount point = X: KAVX0407-I Unmounting has finished. Mount point = X: KAVX0232-I Hid EXG001,B001. KAVX0232-I Hid EXG002,B001.
2. Mount the secondary volume to which the backup data will be restored. Execute the drmmount command on backup server A to mount the secondary volume: BKServer A > drmmount 0000000002 -mount_pt W: KAVX0001-I The drmmount command will now start. KAVX0233-I Made EXG001,B001 open. KAVX0233-I Made EXG002,B001 open. KAVX0234-I Rerecognizes the physical volume. KAVX0235-I Rerecognized the physical volume. KAVX0400-I Mounting will now start. Mount point = W: KAVX0401-I Mounting has finished.
4. Confirm that the data was restored correctly from the tape to the secondary volume.
7. Transfer the temporary file to the database server. Using ftp on database server A, transfer the temporary file from backup server A to database server A: BKServer A > ftp name-of-database-server-A ftp> Username: (Enter the login name) ftp> password: (Enter the password) ftp> get 0000000003.drm ftp> quit BKServer A > 8. Import the temporary file transferred from the backup server into the backup catalog on the database server.
10. Separate the Continuous Access Software pair.
2. Check the backup ID of the backup data. To check the backup ID, execute the drmtapecat command on backup server B: BKServer B > drmtapecat BACKUP-ID BACKUP-OBJECT 0000000012 MSEXCHANGE BKServer B > 3. SNAPSHOT TIME 2007/11/01 11:00:00 EXPIRATION TIME 2007/11/01 11:30:00 BACKUP-MEDIA MEDIA1 Mount the secondary volume to which the backup data will be restored.
5. Confirm that the data was restored correctly from the tape to the secondary volume.
8. Transfer the temporary file to the database server. Using ftp on database server B, transfer the temporary file from backup server B to database server B: BKServer B > ftp name-of-database-server-A ftp> Username: (Enter the login name) ftp> password: (Enter the password) ftp> get 0000000013.drm ftp> quit BKServer B > 9. Import the temporary file transferred from the backup server into the backup catalog on the database server.
11. Separate the Continuous Access Software pair.
• Performing a VSS backup on the standby server (backup data is created on a secondary volume). In this example, the backup data on the secondary volume is then backed up to a tape device. • Restoring on the active server backup data that was acquired on the active server • Performing a fail-over of the active server, and switching the standby server to the active server.
• Backup of another primary volume cannot be performed until the secondary volume is backed up to the tape device. • Initially, database server A is the active server (the one on which the Exchange virtual server is running). Database server B is the standby server. • NetBackup is installed as the tape backup management software on the backup servers, and the configuration definitions for the tape library devices are defined.
Performing a backup on the active server 1. Clear the system cache on the backup server. To clear the system cache, execute the drmmount command on backup server A to temporarily mount the secondary volume. Then, use the drmumount command to unmount the volume. BKServer A > drmmount -copy_group EXG001,B001 -mount_pt W: KAVX0001-I The drmmount command will now start. KAVX0400-I Mounting will now start. Mount point = W: KAVX0401-I Mounting has finished.
3. Confirm that the data was backed up correctly. Execute the drmexgcat command on database server A, and check the backup information. DBServer A > drmexgcat -target STR1 STRAGEGROUP: STR1 BACKUP-ID:0000000001 BACKUP-MODE: VSS STRAGEGROUP: STR1 ORIGINAL-ID:0000000001 START-TIME:2007/11/01 10:00:00 END-TIME: 2007/11/01 10:05:00 HOSTNAME:EXG_SRV1 T OBJECT INFORMATIONSTORE FILE FS DG COPY-GROUP M MAILBOXSTORE MAILBOX1(STR1) D:\STR1\priv1.
7. Confirm that the file was imported correctly. Execute the drmexgcat command on backup server A to check whether the backup information for the date on which backup from the primary volume to the secondary volume was performed was imported to backup server A.
9. Back up the backup data on the secondary volume to tape. Execute the drmmediabackup command on backup sever A to back up the backup data on the secondary volume to tape. An unused backup ID of 0000000002 is assigned. BKServer A > drmmediabackup 0000000001 KAVX0001-I The drmmediabackup command will now start. KAVX0402-I Backup processing will now start. To see more details on backup progress, use the monitoring method of the backup management product.
Performing a backup on the standby server 1. Clear the system cache on the backup server. To clear the system cache, execute the drmmount command on backup server B to temporarily mount the secondary volume. Then, use the drmumount command to unmount the volume. BKServer B > drmmount -copy_group EXG003,B001 -mount_pt W: KAVX0001-I The drmmount command will now start. KAVX0400-I Mounting will now start. Mount point = W: KAVX0401-I Mounting has finished.
3. Confirm that the data was backed up correctly. Execute the drmexgcat command on database server B, and check the backup information. DBServer B > drmexgcat STRAGEGROUP: STR1 BACKUP-ID:0000000001 BACKUP-MODE: VSS STRAGEGROUP: STR1 ORIGINAL-ID:0000000001 START-TIME:2007/11/01 10:00:00 END-TIME: 2007/11/01 10:05:00 HOSTNAME:EXG_SRV1 T OBJECT INFORMATIONSTORE FILE FS DG COPY-GROUP M MAILBOXSTORE MAILBOX1(STR1) D:\STR1\priv1.edb D: EXG003,B001 - EXG003,B001 P PUBLICSTORE PUBLIC1(STR1) D:\STR1\pub1.
7. Confirm that the file was imported correctly. Execute the drmexgcat command on backup server B to check whether the backup information for the date on which backup from the primary volume to the secondary volume was performed was imported to backup server B.
9. Back up the backup data on the secondary volume to tape. Execute the drmmediabackup command on backup server B to back up the backup data on the secondary volume to tape. An unused backup ID of 0000000002 is assigned. BKServer B > drmmediabackup 0000000001 KAVX0001-I The drmmediabackup command will now start. KAVX0402-I Backup processing will now start. To see more details on backup progress, use the monitoring method of the backup management product.
Restoring data when using the automatic seed function The following subsections provide examples of restoring data using different backup servers for the active and standby servers. This example uses the automatic seed function during the restore operation. Restoring backup data acquired on the active server (when the automatic seed function is used) This subsection gives an example of performing a point-in-time restore operation on a storage group, using backup data that was acquired on the active server.
3. Restore the backup data from the tape to the secondary volume. Execute the drmmediarestore command on backup server A to restore the backup data on tape to the secondary volume. BKServer A > drmmediarestore 0000000002 KAVX0001-I The drmmediarestore command will now start. KAVX0404-I Restoration will now start. To see more details on restoration progress, use the monitoring method of the backup management product. Source = W: Target = W: KAVX0404-I Restoration will now start.
5. Unmount the secondary volume after restoration of the backup data is finished. Execute the drmumount command on backup server A to unmount the secondary volume. BKServer A > drmumount 0000000002 KAVX0001-I The drmumount command will now start. KAVX0406-I Unmounting will now start. Mount point = W: KAVX0407-I Unmounting has finished. Mount point = W: KAVX0406-I Unmounting will now start. Mount point = X: KAVX0407-I Unmounting has finished. Mount point = X: KAVX0002-I The drmumount command will now end.
9. Confirm that the file was imported correctly. Execute the drmexgcat command on database server A to check whether the backup information for the date on which backup from the primary volume to the secondary volume was performed was imported to database server A.
Restoring backup data acquired on the standby server (when the automatic seed function is used) This subsection gives an example of performing a roll-forward restore operation on a storage group, using backup data that was acquired on the standby server. 1. Perform a failover operation so that the Exchange virtual server running on database server A fails over to database server B. 2. Check the backup ID of the backup data. To check the backup ID, execute the drmtapecat command on backup server B.
4. Restore the backup data from the tape to the secondary volume. Execute the drmmediarestore command on backup server B to restore the backup data on tape to the secondary volume. BKServer B > drmmediarestore 0000000002 KAVX0001-I The drmmediarestore command will now start. KAVX0404-I Restoration will now start. To see more details on restoration progress, use the monitoring method of the backup management product. Source = W: Target = W: KAVX0404-I Restoration will now start.
6. Unmount the secondary volume after restoration of the backup data is finished. Execute the drmumount command on backup server B to unmount the secondary volume. BKServer B > drmumount 0000000002 KAVX0001-I The drmumount command will now start. KAVX0406-I Unmounting will now start. Mount point = W: KAVX0407-I Unmounting has finished. Mount point = W: KAVX0406-I Unmounting will now start. Mount point = X: KAVX0407-I Unmounting has finished. Mount point = X: KAVX0002-I The drmumount command will now end.
10. Confirm that the file was imported correctly. Execute the drmexgcat command on database server B to check whether the backup information for the date on which backup from the primary volume to the secondary volume was performed was imported to database server B.
Restoring data when not using the automatic seed function The following subsections provide examples of restoring data using different backup servers for the active and standby servers. Restoring Data Using Different Backup Servers for the Active and Standby Servers.
3. Restore the backup data from the tape to the secondary volume. Execute the drmmediarestore command on backup server A to restore the backup data on tape to the secondary volume. BKServer A > drmmediarestore 0000000002 KAVX0001-I The drmmediarestore command will now start. KAVX0404-I Restoration will now start. To see more details on restoration progress, use the monitoring method of the backup management product. Source = W: Target = W: KAVX0404-I Restoration will now start.
5. Unmount the secondary volume after restoration of the backup data is finished. Execute the drmumount command on backup server A to unmount the secondary volume. BKServer A > drmumount 0000000002 KAVX0001-I The drmumount command will now start. KAVX0406-I Unmounting will now start. Mount point = W: KAVX0407-I Unmounting has finished. Mount point = W: KAVX0406-I Unmounting will now start. Mount point = X: KAVX0407-I Unmounting has finished. Mount point = X: KAVX0002-I The drmumount command will now end.
9. Confirm that the file was imported correctly. Execute the drmexgcat command on database server A to check whether the backup information for the date on which backup from the primary volume to the secondary volume was performed was imported to database server A.
Restoring backup data acquired on the standby server (when the automatic seed function is not used) This subsection gives an example of performing a roll-forward restore operation on a storage group, using backup data that was acquired on the standby server. 1. Perform a failover operation so that the Exchange virtual server running on database server A fails over to database server B. 2. Check the backup ID of the backup data. To check the backup ID, execute the drmtapecat command on backup server B.
4. Restore the backup data from the tape to the secondary volume. Execute the drmmediarestore command on backup server B to restore the backup data on tape to the secondary volume. BKServer B > drmmediarestore 0000000002 KAVX0001-I The drmmediarestore command will now start. KAVX0404-I Restoration will now start. To see more details on restoration progress, use the monitoring method of the backup management product. Source = W: Target = W: KAVX0404-I Restoration will now start.
6. Unmount the secondary volume after restoration of the backup data is finished. Execute the drmumount command on backup server B to unmount the secondary volume. BKServer B > drmumount 0000000002 KAVX0001-I The drmumount command will now start. KAVX0406-I Unmounting will now start. Mount point = W: KAVX0407-I Unmounting has finished. Mount point = W: KAVX0406-I Unmounting will now start. Mount point = X: KAVX0407-I Unmounting has finished. Mount point = X: KAVX0002-I The drmumount command will now end.
10. Confirm that the file was imported correctly. Execute the drmexgcat command on database server B to check whether the backup information for the date on which backup from the primary volume to the secondary volume was performed was imported to database server B.
Backing up and restoring in SCR configurations This section provides examples of how to back up and restore data in SCR configurations. The explanations in this section use operation examples for Exchange Server 2007. System configuration A system configuration example is shown below.
Disk number Drive letter Disk P1,Disk P3 D: Disk P2,Disk P4 E: • Each disks contains the following. Disk number Object Disk P1 The information storage configuration file for the source storage group (STR1). Disk P2 The transaction log file and checkpoint file for the source storage group (STR1). Disk P3 The information storage configuration file for the target storage group (STR1). Disk P4 The transaction log file and checkpoint file for the target storage group (STR1).
2. Clear the system cache right before performing a backup. To clear the system cache, temporarily mount the secondary volume on the backup server. PROMPT> drmmount -copy_group EXG001,B001 -mount_pt W: KAVX0001-I The drmmount command will now start. KAVX0400-I Mounting will now start. Mount point = W: KAVX0401-I Mounting has finished. Mount point = W: KAVX0002-I The drmmount will now end. PROMPT> drmumount -copy_group EXG001,B001 KAVX0001-I The drmumount command will now start.
4. From the source server, confirm that the backup was successful. PROMPT> drmexgcat STRAGEGROUP: STR1 BACKUP-ID:0000000001 BACKUP-MODE: VSS STRAGEGROUP: STR1 ORIGINAL-ID:0000000001 START-TIME:2007/11/01 10:00:00 END-TIME: 2007/11/01 10:05:00 HOSTNAME:EXG_SRV1 T OBJECT INFORMATIONSTORE FILE FS DG COPY-GROUP M MAILBOXSTORE MAILBOX1(STR1) D:\STR1\priv1.edb D: EXG001,B001 - EXG001,B001 P PUBLICSTORE PUBLIC1(STR1) D:\STR1\pub1.edb D: EXG001,B001 - EXG001,B001 T TRANSACT E:\STR1\E00*.
8. Check the backup catalog information that was imported into the backup server. PROMPT> drmexgcat STRAGEGROUP: STR1 BACKUP-ID:0000000001 BACKUP-MODE: VSS STRAGEGROUP: STR1 ORIGINAL-ID:0000000001 START-TIME:2007/11/01 10:00:00 END-TIME: 2007/11/01 10:05:00 HOSTNAME:EXG_SRV1 T OBJECT INFORMATIONSTORE FILE FS DG COPY-GROUP M MAILBOXSTORE MAILBOX1(STR1) D:\STR1\priv1.edb D: EXG001,B001 - EXG001,B001 P PUBLICSTORE PUBLIC1(STR1) D:\STR1\pub1.edb D: EXG001,B001 - EXG001,B001 T TRANSACT E:\STR1\E00*.
10. From the backup server, back up the data on the secondary volume to a tape device. An unused backup ID (0000000002) is assigned. PROMPT> drmmediabackup 0000000001 KAVX0001-I The drmmediabackup command will now start. KAVX0402-I Backup processing will now start. To see more details on backup progress, use the monitoring method of the backup management product. Source = W: KAVX0402-I Backup processing will now start.
2. From the backup server, check the backup ID of the tape device. PROMPT> drmtapecat BACKUP-ID BACKUP-OBJECT 0000000002 MSEXCHANGE PROMPT> 3. SNAPSHOT TIME 2002/06/01 11:00:00 EXPIRATION TIME 2002/06/02 11:00:00 BACKUP-MEDIA MEDIA1 From the backup server, mount the secondary volume to restore data from the tape device. Specify the backup ID 0000000002, and then execute the drmmount command. PROMPT> drmmount 0000000002 -mount_pt W: KAVX0001-I The drmmount command will now start.
5. From the backup server, confirm that the restoration from the tape device to the secondary volume was successful. PROMPT> drmexgcat STRAGEGROUP: STR1 BACKUP-ID:0000000003 BACKUP-MODE: VSS STRAGEGROUP: STR1 ORIGINAL-ID:0000000001 START-TIME:2007/11/01 10:00:00 END-TIME: 2007/11/01 10:05:00 HOSTNAME:EXG_SRV1 T OBJECT INFORMATIONSTORE FILE FS DG COPY-GROUP M MAILBOXSTORE MAILBOX1(STR1) D:\STR1\priv1.edb D: EXG001,B001 - EXG001,B001 P PUBLICSTORE PUBLIC1(STR1) D:\STR1\pub1.
9. Import the transferred temporary file on the source server into the backup catalog. A new backup ID is registered into the backup catalog. An unused backup ID (0000000002) is assigned. PROMPT> drmdbimport -f C:\0000000005.drm KAVX0001-I The drmdbimport command will now start. KAVX0418-I Backup information from the specified file was imported into the backup catalog. Imported backup ID = 0000000002 KAVX0002-I The drmdbimport command will now end. PROMPT> 10.
Point-in-time restoration when not using the SCR seed function This section gives an example of how to perform a point-in-time restoration on a storage group, using backed up data that was acquired in an SCR configuration, and without using the SCR seed function. This example uses the backup data acquired in “Backing up in an SCR configuration” on page 475. 1. From the backup server, start the Protection Manager service. 2. From the backup server, check the backup ID of the tape device.
5. From the backup server, confirm that the restoration from the tape device to the secondary volume was successful. PROMPT> drmexgcat STRAGEGROUP: STR1 BACKUP-ID:0000000003 BACKUP-MODE: VSS STRAGEGROUP: STR1 ORIGINAL-ID:0000000001 START-TIME:2007/11/01 10:00:00 END-TIME: 2007/11/01 10:05:00 HOSTNAME:EXG_SRV1 T OBJECT INFORMATIONSTORE FILE FS DG COPY-GROUP M MAILBOXSTORE MAILBOX1(STR1) D:\STR1\priv1.edb D: EXG001,B001 - EXG001,B001 P PUBLICSTORE PUBLIC1(STR1) D:\STR1\pub1.
9. Import the transferred temporary file on the source server into the backup catalog. A new backup ID is registered into the backup catalog. An unused backup ID (0000000002) is assigned. PROMPT> drmdbimport -f C:\0000000005.drm KAVX0001-I The drmdbimport command will now start. KAVX0418-I Backup information from the specified file was imported into the backup catalog. Imported backup ID = 0000000002 KAVX0002-I The drmdbimport command will now end. PROMPT> 10.
Point-in-time restoration when using the SCR seed function This section gives an example of how to perform a point-in-time restoration on a storage group, using backed up data that was acquired in an SCR configuration, and using the SCR seed function. This example uses the backup data acquired in “Backing up in an SCR configuration” on page 475. 1. From the backup server or target server, start the Protection Manager service. 2. From the backup server, check the backup ID of the tape device.
5. From the backup server, confirm that the restoration from the tape device to the secondary volume was successful. PROMPT> drmexgcat STRAGEGROUP: STR1 BACKUP-ID:0000000003 BACKUP-MODE: VSS STRAGEGROUP: STR1 ORIGINAL-ID:0000000001 START-TIME:2007/11/01 10:00:00 END-TIME: 2007/11/01 10:05:00 HOSTNAME:EXG_SRV1 T OBJECT INFORMATIONSTORE FILE FS DG COPY-GROUP M MAILBOXSTORE MAILBOX1(STR1) D:\STR1\priv1.edb D: EXG001,B001 - EXG001,B001 P PUBLICSTORE PUBLIC1(STR1) D:\STR1\pub1.
9. Import the transferred temporary file on the source server into the backup catalog. A new backup ID is registered into the backup catalog. An unused backup ID (0000000002) is assigned. PROMPT> drmdbimport -f C:\0000000005.drm KAVX0001-I The drmdbimport command will now start. KAVX0418-I Backup information from the specified file was imported into the backup catalog. Imported backup ID = 0000000002 KAVX0002-I The drmdbimport command will now end. PROMPT> 10.
11. While the SCR seed function is still enabled, perform a point-in-time restoration on the backup data from the secondary volume to the primary volume. PROMPT> drmexgrestore 0000000002 -resync -recovery -ef exchange.conf KAVX0001-I The drmexgrestore command will now start. KAVX1526-I The Exchange environment configuration file has been read. Exchange environment configuration file = exchange.conf KAVX0256-I A connection to the Protection Manager service will now be established.
Figure 129 System configuration for backing up and restoring data in a DAG configuration For the backup and restoration examples below, assume the following: • If the backup operation is backing up a secondary volume to a tape device, another primary volume cannot be backed up until the current operation finishes. • In the initial status, disks P1 and P2 of database server 1 are the active mailbox database copies.
Disk number Object Disk P1 The database file for the active mailbox database copy Mail01 Disk P2 The transaction log file and checkpoint file for the active mailbox database copy Mail01 Disk P3 The database file for the passive mailbox database copy Mail01 Disk P4 The transaction log file and checkpoint file for the passive mailbox database copy Mail01 • Pair definitions have been defined for the primary and secondary volumes in a RAID Manager configuration definition file on the database server a
2. On database server 1, perform a VSS backup from the primary volume to the secondary volume. A new backup ID is registered in the backup catalog. The unused ID 0000000001 is set for the backup ID. The backup catalog is imported into backup server 1. PROMPT> drmexgbackup -mode vss -s backup server 1 -auto_import KAVX0001-I The drmexgbackup command will now start. KAVX0256-I A connection to the Protection Manager service will now be established.
4. Check the imported backup catalog information on backup server 1. PROMPT> drmexgcat STORAGEGROUP: /EXCHANGE BACKUP-ID: 0000000001 BACKUP-MODE: VSS STORAGEGROUP: /EXCHANGE ORIGINAL-ID: 0000000001 START-TIME: 2010/01/01 11:00:00 END-TIME: 2010/01/01 11:30:00 HOSTNAME: EXG_SRV1 T OBJECT INFORMATIONSTORE FILE FS DG DEVICE COPY-GROUP M MAILBOXSTORE EXG001,B001 - EXG001,B001 T TRANSACT EXG001,B002 - EXG001,B002 C CHECKPOINT EXG001,B002 - EXG001,B002 PROMPT> 5. Mail1 D:\Mail1\mail.
7. On backup server 1, verify that the backup to the tape device was completed successfully. PROMPT> drmtapecat -backup_id 0000000002 BACKUP-ID BACKUP-OBJECT SNAPSHOT TIME 0000000002 MSEXCHANGE 2010/01/01 12:00:00 PROMPT> 8. EXPIRATION TIME 2010/01/01 13:00:00 BACKUP-MEDIA MEDIA1 On backup server 1, unmount the secondary volume on which the backup data is stored. Unmount the secondary volume with backup ID 0000000001 (set when the temporary file was imported).
2. On database server 2, perform a VSS backup from the primary volume to the secondary volume. A new backup ID is registered in the backup catalog. The unused ID 0000000001 is set for the backup ID. The backup catalog is imported into backup server 2. PROMPT> drmexgbackup -mode vss -s backup server 2 -auto_import KAVX0001-I The drmexgbackup command will now start. KAVX0256-I A connection to the Protection Manager service will now be established.
4. Check the imported backup catalog information on backup server 2. PROMPT> drmexgcat STORAGEGROUP: /EXCHANGE BACKUP-ID: 0000000001 BACKUP-MODE: VSS STORAGEGROUP: /EXCHANGE ORIGINAL-ID: 0000000001 START-TIME: 2010/01/01 11:00:00 END-TIME: 2010/01/01 11:30:00 HOSTNAME: EXG_SRV2 T OBJECT INFORMATIONSTORE FILE FS DG DEVICE COPY-GROUP M MAILBOXSTORE EXG002,B001 - EXG002,B001 T TRANSACT EXG002,B002 - EXG002,B002 C CHECKPOINT EXG002,B002 - EXG002,B002 PROMPT> 5. Mail1 D:\Mail1\mail.
7. On backup server 2, verify that the backup to the tape device was completed successfully. PROMPT> drmtapecat -backup_id 0000000002 BACKUP-ID BACKUP-OBJECT SNAPSHOT TIME 0000000002 MSEXCHANGE 2010/01/01 12:00:00 PROMPT> 8. EXPIRATION TIME 2010/01/01 13:00:00 BACKUP-MEDIA MEDIA1 On backup server 2, unmount the secondary volume on which the backup data is stored. Unmount the secondary volume with backup ID 0000000001 (set when the temporary file was imported).
4. On backup server 1, restore the backup data from the tape device to the secondary volume. Execute the drmmediarestore command with backup ID 0000000002 specified. The backup ID 0000000001 is deleted, and the new backup ID 0000000003 is registered. PROMPT> drmmediarestore 0000000002 KAVX0001-I The drmmediarestore command will now start. KAVX0404-I Restoration will now start. To see more details on restoration progress, use the monitoring method of the backup management product.
7. On backup server 1, export the backup catalog information to the temporary file. PROMPT> drmdbexport 0000000003 -f C:\FTP_ROOT\0000000003.drm KAVX0001-I The drmdbexport command will now start. KAVX0002-I The drmdbexport command will now end. PROMPT> 8. On backup server 1, use FTP to transfer the exported temporary file to database server 1. PROMPT> ftp-database-server-1-name ftp> Username: (Enter the login name) ftp> password: (Enter the password) ftp> get 0000000003.drm ftp> quit PROMPT> 9.
11. On the backup server, perform a roll-forward restore operation for the data on the secondary volume to the primary volume. Do this with the seed function enabled. PROMPT> drmexgrestore 0000000002 -resync -recovery -ef exchange.conf KAVX0001-I The drmexgrestore command will now start. KAVX1526-I The Exchange environment configuration file has been read. Exchange environment configuration file = exchange.conf KAVX0256-I A connection to the Protection Manager service will now be established.
4. On backup server 2, mount the secondary volume to restore the backup data from the tape device. Execute the drmmount command with backup ID 0000000002 specified. PROMPT> drmmount 0000000002 -mount_pt W: KAVX0001-I The drmmount command will now start. KAVX0400-I Mounting will now start. Mount point = W: KAVX0401-I Mounting has finished. Mount point = W: KAVX0400-I Mounting will now start. Mount point = X: KAVX0401-I Mounting has finished. Mount point = X: KAVX0002-I The drmmount command will now end.
7. On backup server 2, unmount the secondary volume on which the restore operation from the tape device has been completed. Execute the drmumount command with backup ID 0000000002 specified. PROMPT> drmumount 0000000002 KAVX0001-I The drmumount command will now start. KAVX0406-I Unmounting will now start. Mount point = W: KAVX0407-I Unmounting has finished. Mount point = W: KAVX0406-I Unmounting will now start. Mount point = X: KAVX0407-I Unmounting has finished.
11. On database server 2, check the execution result of the import operation. PROMPT> drmexgcat STORAGEGROUP: /EXCHANGE BACKUP-ID: 0000000001 BACKUP-MODE: VSS STORAGEGROUP: /EXCHANGE ORIGINAL-ID: 0000000001 START-TIME: 2010/01/01 11:00:00 END-TIME: 2010/01/01 11:30:00 HOSTNAME: EXG_SRV2 T OBJECT INFORMATIONSTORE FILE FS DG DEVICE COPY-GROUP M MAILBOXSTORE Mail1 D:\Mail1\mail.edb D: Harddisk2 EXG002,B001 - EXG002,B001 T TRANSACT Mail1 E:\Mail1\E01*.
To back up storage groups to tape (where a tape backup management software that is unable to link with Application Agent is used): 1. When usual operation is in the split state, execute EX_DRM_CACHE_PURGE to clear the secondary volume cache. 2. When usual operation is in the split state, execute EX_DRM_RESYNC to resynchronize the copy group. 3. Execute EX_DRM_EXG_BACKUP to back up the file system to the secondary volume. 4. Execute EX_DRM_DB_EXPORT to export the backup information to temporary files.
Preparing to use a file share for storage group backup and restore To prepare to back up and restore storage groups by using a file share: 1. Check the temporary directory for extended commands that was created automatically by the check tool for operation definition files. For more information about where the temporary directory for extended commands is created automatically, see “Checking the temporary directory for extended commands” on page 199. 2.
Example of using a file share for storage group restoration This section explains how to restore storage groups on the backup server by using a file share. For more operational details on steps 1 to 3 and step 5 and subsequent steps, see “Restoring storage groups from tape” on page 417. To restore storage groups on the backup server by using a file share: 1. Execute the drmtapecat command to check the backup ID of the backup data. 2.
8 Troubleshooting This chapter explains how to deal with problems that may occur during the operation of Application Agent. The descriptions in this chapter apply to the Application Agent CLI. Troubleshooting procedure To deal with problems that may occur during the operation of Application Agent follow the steps below. If a problem occurs during operation using an extended command, see “Troubleshooting extended commands” on page 507. 1. Check the conditions in which the problem occurred.
2. Check the problem in the Application Agent log file. When an error occurs in a Application Agent command executed within an extended command, an error message is output to the Application Agent log file drm_output.log (or drm_output.log.old). See the content of the Application Agent log file regarding the time when the error message for the extended command was output, and check the message and the operation status from the time the problem with the Application Agent command occurred. 3.
Information type Summary Default file name Windows Server 2008 or Windows Server 2012 Problem report and problem resolution log file#2 Microsoft Cluster Service or Windows Server Failover Clustering information All files in the following folder: system-drive\Users\user-name#3\AppData\Local\ Microsoft\Windows\WER Definition file and log file CHKDSK execution logs -system-directory\cluster\ChkDsk_*.log Legend: -- indicates that there is no corresponding file name.
Information type Summary Default files Detailed trace logs Application Agent trace log files#2 Application-Agent-installation-directory\DRM\log\ drm_pp_trace[1-16].log Installation trace logs Application Agent installation trace log files Application-Agent-installation-directory\DRM\log\ drm_inst.log Application-Agent-installation-directory\DRM\log\ drm_nbu_backup.log Trace logs of NetBackup Application Agent log files#3 Application-Agent-installation-directory\DRM\log\ drm_nbu_backup.log.
size of the drm_nbu_restore.log file reaches 10 MB, the contents of the file are moved to drm_nbu_restore.log.old. If drm_nbu_restore.log.old already exists, it is overwritten. #4 If multiple directories of the same virtual server name are specified in the DRM_DB_PATH parameter, a number n will be appended to the name of the directories from which log data will be collected. This number is based on the order that the directories are specified in the DRM_DB_PATH parameter.
Database information (when the backup is of an SQL Server database or Exchange database) Collect the following information about the database. In cases where the problem occurred over a network connection, you will need to collect files from the connected machine.
• A command (including Application Agent commands (drmxxx)) inside the extended command is executed and then failed The log file output directory can store up to two generations of log files. The paths of the log files are as follows. Current log file: Application-Agent-installation-directory\DRM\script\log\drm_script.log Previous log file: Application-Agent-installation-directory\DRM\script\log\drm_script.log.
Item Explanation When the backup is of a file system: The name of the mount point directory or mount point directory list file that was backed up. Instance name When the backup is of an SQL Server database: The name of the SQL server instance that was backed up is output. When the backup is of an Exchange database: The character string defined in the specification item INSTANCE_NAME in the operation definition file is output.
Adjusting trace log settings This section explains how to adjust trace log settings. Adjusting the output level for detailed trace log information The level of information output to trace log files of Application Agent can be changed whenever necessary. To adjust the amount of log information to be collected, or to select a certain type of log information to be collected, edit the Application Agent configuration definition file (init.
Investigating standard output logs When execution of a Application Agent results in some problem, investigate the information in the standard output logs. If a problem occurs, the time when the problem occurred and the messages that were displayed in the standard output are recorded in the standard output logs. Refer to the contents of the standard output log, and check what messages were output when the problem occurred.
The following shows the items output to the detailed trace log. Table 111 Items output to the detailed trace log Output item Description Number Log sequence number Date Date when the log was output, using the yyyy/mm/dd format Time Time when the log was output, using the hh:mm:ss.sss format Application name Application Agent command name (drmxxx) that triggered the output of the log. Process ID Process ID Thread ID Thread ID Message ID ID of the message Type of the message.
Source of messages (controlled objects) Message output is triggered by: Message ID Cluster Cluster operations using the Windows API KAVX9203-I, KAVX9204-W, KAVX9205-E Operations and instruction using commands KAVX9300-I, KAVX9301-W, KAVX9302-E Operations and instruction using ODBC KAVX9306-I, KAVX9307-W, KAVX9308-E Use of COM (MSSQL) interface KAVX9400-I, KAVX9401-W, KAVX9402-E Use of COM (ADO) interface KAVX9403-I, KAVX9404-W, KAVX9405-E Use of COM (CDOEXM) interface KAVX9406-I, KAVX9407-W,
KAVX0006-E An error that stopped processing occurred. Cause = DRM-10106: An internal error occurred. In such a case, if the following message is output to the detailed trace log, the dictionary map files are invalid: ISAM API error code = xxx (yyyyy) The dictionary map files consist of key definition files (.DEF), primary key files (.K01), secondary key files (.Knn, where nn is 02 - 99), and data files (.DRF or .DAT).
2. Take the following actions in accordance with the output results. When the message KAIU020-I is output: Make sure that there are no problems in the dictionary map file information and the backup information. Execute the commands to display the dictionary map file information and the backup information and by confirming that no error messages are displayed, you can confirm that there are no problems in the dictionary map files.
1. Create the parameter file for error recovery in the desired directory. fi-Application-Agent-installation-directory\DRM\db\table-to-be-recovered 2. Execute the drmiskeymnt command to restore missing key files (.K01 or .Knn). drmiskeymnt parameter-file-name 3. Execute the drmischk command to verify the restored dictionary map files. drmischk -l 3 Application-Agent-installation-directory\DRM\db\DRM* 4.
1. Execute the drmisext command to extract as many records as possible to a temporary file. drmischk Application-Agent-installation-directory\DRM\db\table-to-be-recovered temporary-file-name 2. Create dictionary map files from the temporary file containing the data extracted by the drmisconv command. drmisconv -t SI temporary-file-name Application-Agent-installation-directory\DRM\db\ table-to-be-recovered 3. Execute the drmischk command to verify the restored dictionary map files.
Recreating dictionary map files 1. Copy and save all files and directories in the directory storing the dictionary map files. 2. Delete all files and directories in the directory storing the dictionary map files. 3. When using the SQL Server database, delete all files and directories in the directory in which the VDI metafiles are stored. To identify the directory storing the VDI metafiles, execute the drmsqlinit command with the -v option. 4. Recreate the dictionary map files.
Troubleshooting
9 Support and other resources Numerous support options are available. Contacting HP HP technical support For worldwide technical support information, see the HP support website: http://www.hp.
Conventions This guide follows the conventions shown in Table 115 to refer to HP P9000 products.
Product reference Full name or meaning This abbreviation is used when it is not necessary to distinguish the following products: Continuous Access Asynchronous • XP Continuous Access Asynchronous • Continuous Access XP Asynchronous • Continuous Access XP Extension This abbreviation is used when it is not necessary to distinguish the following products: Data Retention Utility • Data Retention Utility • XP Data Retention Utility • LUN Security XP Extension This abbreviation is used when it is not necessa
Product reference Full name or meaning This abbreviation is used when it is not necessary to distinguish the following products: • Snapshot Snapshot • XP Snapshot • Snapshot XP • P9000 Snapshot This abbreviation is used when it is not necessary to distinguish the following products: Volume Shredder • Volume Shredder • XP Volume Shredder • P9000 Volume Shredder 528 Support and other resources
A Example of an Application Agent environment configuration This appendix explains an example configuration of the Application Agent environment. The descriptions in this appendix apply to the Application Agent CLI. You can use this example as a reference when setting up an actual environment. Note This example is based on operation in a specific configuration. For other configurations, you must adjust the settings to match operational requirements.
Figure 130 System configuration (for the example Application Agent environment configuration) The following example shows details regarding the above configuration. Server configuration Using two database servers, configure a virtual server by using an active-passive cluster configuration. The backup server is linked to a tape device. The following table shows the server configuration.
Information (Database server) Information (Backup server) • RAID Manager (instance number: 1) • RAID Manager (instance number: 1) • Device Manager Agent • Device Manager Agent • Application Agent • Application Agent Application Agent installation directory C:\Program Files\CVXPAE\DRM (default-installation-directory) C:\Program Files\CVXPAE\DRM (default-installation-directory) Cluster software Microsoft Cluster Service -- Virtual server name SQL001 -- DBMS SQL Server -- SQL Server instanc
Mount drive from the database server Mount drive from the backup server Copy group name Generation VG02,VOL23 2nd generation VG03,VOL33 3rd generation VG01,VOL14 1st generation VG02,VOL24 2nd generation VG03,VOL34 3rd generation Volume contents DB02 (LDF) K: L: Z: -- -- -- DB03 (MDF,LDF) Dictionary map file, VDI metafile, transaction log backup file, temporary directory for extended commands, etc.
(C:\Windows\horcm1.conf) HORCM_MON #ip_address service poll(10ms) localhost horcm1 6000 HORCM_CMD #dev_name dev_name \\.
HORCM_MON #ip_address service poll(10ms) localhost horcm2 6000 HORCM_CMD #dev_name dev_name \\.
• RAID Manager configuration definition file, for the backup server (C:\Windows\horcm2.conf) HORCM_MON #ip_address service poll(10ms) localhost horcm2 6000 HORCM_CMD #dev_name dev_name \\.
LOGLEVEL=3 PP_LOGFILE_NUM=2 COM_RETRY_TIME=0 COM_RETRY_WAIT=10 CLU_RETRY_TIME=6 CLU_RETRY_WAIT=10 SVC_RETRY_TIME=6 SVC_RETRY_WAIT=10 DRM_DB_PATH=L:\PTM;SQL001 In this environment configuration example, the dictionary map file directory is L:\PTM. Since the name of the virtual server is SQL001, specify L:\PTM;SQL001 for DRM_DB_PATH. • Application Agent configuration definition file, for the backup server (C:\Program Files\CVXPAE\DRM\conf\init.
HORCMINST=1 HORCMINST_AUX=2 RETRY_TIME=120 RETRY_WAIT=5 VENDER=HP PRODUCT=OPEN-3 INSTALLPATH=C:\HORCM • RAID Manager-linkage definition file, for the backup server (C:\Program Files\CVXPAE\DRM\conf\raid\DEFAULT.dat) HORCMINST=2 HORCMINST_AUX=1 RETRY_TIME=120 RETRY_WAIT=5 VENDER=HP PRODUCT=OPEN-3 INSTALLPATH=C:\HORCM For details about the RAID Manager-linkage definition file (DEFAULT.dat), see “Setting up Application Agent for linkage with RAID Manager” on page 115.
PROMPT> set DRM_HOSTNAME=SQL001 PROMPT> drmsqlinit DEFAULT PROMPT> KAVX1100-I Enter the name of the storage directory for VDI metafiles: L:\mssql\meta PROMPT> KAVX1101-I Enter the timeout interval for VDI generation (in seconds, between 0 and 3600): 1000 PROMPT> KAVX1102-I Enter the name of the storage directory for undo files: L:\mssql\undo PROMPT> KAVX1107-I Enter the name of the storage directory for backup-log files: L:\mssql\log Check that the SQL Server database configuration definition file was c
CLASS can be replaced with POLICY for this example. However, both POLICY and CLASS cannot be specified for the same MOUNT_POINT. Example registering information for linkage to the tape backup management software • The following shows an example registering information for linkage to the tape backup management software.
INST_PATH=C:\Program Files\VERITAS\NetBackup MOUNT_POINT=W: CLASS=Class1 SCHEDULE=Schedule1 MOUNT_POINT=X: CLASS=Class2 SCHEDULE=Schedule2 MOUNT_POINT=Y: CLASS=Class3 SCHEDULE=Schedule3 MOUNT_POINT=Z: CLASS=Class4 SCHEDULE=Schedule4 MOUNT_POINT=default CLASS=Class5 SCHEDULE=Schedule5 PARALLEL_COUNT=5 For details on the configuration definition file for tape backup, see “Creating a configuration definition file for tape backup” on page 154.
Operation definition files (_operation-ID.dat) with the same contents are placed on each backup server (both the active server and the standby server). In this example, the operation ID is OperationA.
VG03,VOL31 VG03,VOL32 VG03,VOL33 VG03,VOL34 For details about copy-group list files, see “Creating a copy-group list file” on page 159.
B List of files used in Application Agent This appendix explains the purposes and locations of the files used by Application Agent. The descriptions in this appendix apply to the Application Agent CLI. List of files used in Application Agent The following table lists the files used in Application Agent.
File name Description Storage location Backup catalog file Files that contain exported backup catalog information. These files are used to exchange backup catalog information between file servers (or database servers) and backup servers. -- Database configuration definition files (instancename.dat) Files that contain information (parameters) about target databases registered by using the drmsqlinit command. These files are used when backup is of an SQL Server database.
File name Description Storage location • When the backup is of an SQL Server database: Definition files containing a list of databases subject to backup Database batch definition files • When the backup is of an Exchange database: User creates the files in the desired location.
File name Backup ID files Description Storage location Temporary files created when extended commands are executed. These files are used to carry over backup IDs between extended commands. These files are used for extended commands. Temporary files Saved copies of files specific to the system subject to backup, created during disk backup. These files are used during database restoration. In a file server or database server: These files are used for extended commands.
File name Description Storage location Copy parameters definition file (any-filename) Create this file to change the retry time for confirming the pair statuses used when commands are executed in operation. Specify this file during command execution for backup, restoration, and resynchronization. Copy group mount file (CG_MP.conf) This file is necessary when copy group names are used for parameters, in a dynamic disk configuration.
List of files used in Application Agent
C Replacing primary and secondary volume disks while Application Agent is running This appendix explains how to replace a primary or secondary volume disk while Application Agent is running. The descriptions in this appendix apply to the Application Agent CLI. Replacing primary and secondary volume disks while Application Agent is running This section explains how to replace a primary volume disk and secondary volume disk while Application Agent is running.
Figure 131 Example of replacing a primary volume disk To replace the primary volume: 1. Execute the following RAID Manager command to delete the pair definition of the primary volume to be replaced. PROMPT> pairsplit -g SI01 -S 2. On the host that manages the primary volume, delete the volume. In Windows Server 2003: From the Start menu, choose Control Panel, Management Tool, and Computer Management.
3. 4. On the host that manages the primary volume, create a volume for the new primary volume. In Windows Server 2003: From the Start menu, choose Control Panel, Management Tool, and Computer Management. Select Disk Management from the Computer Management window and create partitions in disk management. In Windows Server 2008: From the Start menu, choose Administrative Tools, and then open Server Manager.
12. On the host that manages the primary volume, import the catalog. PROMPT> drmdbimport -f \tmp\expfile 13. Execute the drmfsrestore command with the -force option to restore data from the secondary volume to the primary volume. PROMPT> drmfsrestore 0000000155 -resync -force Procedure for replacing a secondary volume disk The following example shows how to replace a disk on which a secondary volume belongs: Figure 132 Example of replacing a secondary volume disk To replace the secondary volume: 1.
2. 3. On the host that manages the secondary volume, create the volume for a new secondary volume. In Windows Server 2003: From the Start menu, choose Control Panel, Management Tool, and Computer Management. Select Disk Management from the Computer Management window and create partitions by using the disk management functionality. In Windows Server 2008: From the Start menu, choose Administrative Tools, and then open Server Manager.
Replacing primary and secondary volume disks while Application Agent is running
D Restoring from a tape directly to a primary volume in a Fast Snap or Snapshot configuration This appendix describes an example of restoring data from tape directly to a primary volume in a Fast Snap or Snapshot configuration. The descriptions in this appendix apply to the Application Agent CLI. Note that the procedure for restoring data from tape directly to a primary volume is not supported if the configuration uses Exchange Server.
Procedure for restoring from tape directly to the primary volume Perform the following procedure on the database server. To restore data that was backed up to tape directly to the primary volume: 1. Execute the following command to import the file of backup information corresponding to the data that is to be restored.
E Correspondence to Replication Manager GUI terms Some terms used in the Application Agent command line interface have different meanings from those used in the Replication Manager GUI. Correspondence to Replication Manager GUI terms The following table describes the correspondence between these terms.
Correspondence to Replication Manager GUI terms
Glossary This glossary defines the special terms used in this document. Click the desired letter below to display the glossary entries that start with that letter. application map file One of the map files making up the dictionary map files. Application map files are used to keep track of the mapping information between the application data to be backed up, and the files on the file system.
copy group A copy group consists of a primary volume and secondary volume that are synchronized and split using volume replication functionality and RAID Manager functionality. A copy group can also be called a pair volume. copy group map file One of the map files making up the dictionary map files. Copy group map files are used to keep track of the mapping information between the primary volume and its corresponding secondary volumes.
If a database is frozen when a paired volume is synchronized, the primary and secondary volumes will be identical. Backing up the database in this state will ensure that its integrity is maintained during backup processing. HP StorageWorks P9000 Protection Manager software The predecessor product of Application Agent. Information Store A database that is used in Exchange Server. There are two types of Information Stores: Mailbox Stores and Public Folder Stores.
Glossary
Index A Active-Active, 44, 132 Active-Passive, 43, 130 adjusting number of log files for detailed trace log, 515 output level for detailed trace log, 515 allocating files configuring database, 54 metafiles, 55 multiple database configuration files to one volume, 55 AlwaysOn availability group configuration (SQL Server) backing up and restoring databases, 392 backing up and restoring user database on active server, 393 restoring user database (from active server to secondary-use server), 396 restoring user d
backing up storage group by using user script system configuration, 419 backing up storage group to secondary volume, 404 backing up storage group to tape, 412 backing up storage groups to a secondary volume, 407 backing up transaction logs, 312 Backing up the Application Agent operating environment, 212 transaction logs in a cascade configuration without a backup catalog, 369 Backing up and restoring a file system, 243 file share, 271 backup catalog, 29, 559 backup catalog file, 544 Backup Exec backup job
CONFIRM_GENERATION_IDENTICAL, 127 contents of list file, 161 contents of transaction log list file, 162 Continuous Access Journal, 559 Continuous Access Software, 559 copy group, 560 locking, 322 unlocking, 324 copy group batch definition file, 545 copy group map file, 543, 560 copy group mount file, 547 copy group names, 160 copy parameter definition file, 122 copy parameters definition file, 547 COPY_GROUP, 151 COPY_GROUP_NUM, 151 COPY_SIZE, 125 core map file, 543, 560 correspondence between extended comm
distribution target server setting, 348 DRM_DB_SYNC, 110 drm_output.log, 201, 508, 509 drm_output.log.old, 508 drm_script.log, 507 drm_script.log.
extended command trace log file, 507 extended commands, 219 troubleshooting, 507 F failover, 560 Fast Snap, 560 Fast Snap or Snapshot configuration example, 555 file name of list file, 161 file name of transaction log list file, 161 file share backing up and restoring storage group, 504 backing up or restoring SQL Server database, 375 backing up SQL Server database by using file share, 377 example of backing up file system, 272 example of backing up storage group, 505 example of restoring database by using
HP StorageWorks P9000 Protection Manager software, 561 I INCLUDE_EXEC, 155 incorrect data file restoring, 521 Information Store, 561 init.
period required to verify Exchange Server 2007 database for backup by using VSS, 72 period required to verify Exchange Server 2010 database for backup by using VSS, 73 permissions required for users executing commands, 231 POLICY, 156 possibility of concurrent execution with other command, 235 possibility of concurrent execution with other command (common command), 236 possibility of concurrent execution with other command (tape-related command), 236 possibility of concurrent execution with other command (u
R RAID Manager environment setup, 100 RAID Manager configuration definition file, 100, 544 RAID Manager information, 511 RAID Manager-linkage definition file, 115, 544 RAID Manager-linkage definition file (DEFAULT.
SCR configurations for backing up Exchange databases, 61 SCRIPT_FILE, 157 secondary volume disk replacing, 552 replacing, 552 SET_DRM_HOSTNAME, 186, 191, 196 setting as operating server distribution target server, 348 setting up Application Agent for linkage with RAID Manager, 115 setting up environment for Application Agent, 97 setting up linkage with tape backup management software creating configuration definition file for tape backup, 154 registering information for linkage with tape backup management s
SQL Server database requirement for executing command, 284 SQL Server database status for executing command, 285 SQL Server database type for executing command, 288 SQL Server replication function, 386 SQL Server service status for executing command, 284 SQL_AUTORECOVERY_TIME, 144 SQL_CHECKDB, 110 SQL_LOGIN_TIMEOUT, 144 SQL_QUICK_BACKUP, 109 SQL Server database Backing up and restoring an SQL Server database, 275 Backing up and restoring between local sites and remote sites, 325 Precautions when backing up
volume copy function between storage systems, 32 volume dynamic recognition, 561 volume replication functions, 561 VSS using, 70 VSS definition file, 145, 547 VSS used configurations, 45 vsscom.
