Frequently Asked Questions - DB2 MultiNode iDataAgent

Do we need to restart DB2 instances on UNIX after upgrading a Commvault DB2 DataAgent?

The DB2 Services must be restarted to clear the API information cached by the db2logmgr process.

This will prevent the following error from occurring during an online full backup job.

SQL2428N The BACKUP did not complete because one or more of the requested log files could not be retrieved

Can we protect data on non-global zones if the File System iDataAgent is installed only on the global zone?

Yes. We can protect file system data on non-global zones if the File System iDataAgent is installed only on the global zone. However, in order to enable consistent backups of application specific data on the non-global zones, you will need to install the corresponding application specific iDataAgent on the non-global zone.

How do we protect DB2 data on a non-global zone?

In order to enable consistent backups of DB2 data on a non-global zone, you need to install the DB2 iDataAgent on the non-global zone.

How are command line backup images displayed in the CommCell Console?

When you run DB2 command line offline/online backup jobs, the backup image is always displayed as online in the CommCell Console.

What are the folders or files created and changes done during install of this agent?

Note: The Solaris and HP-UX platform information only apply to DB2 DPF configurations. The are not supported for DB2 pureScale configurations. For a complete list of supported platforms, see

• System Requirements - DB2 MultiNode iDataAgents (DB2 pureScale)

• System Requirements - DB2 MultiNode iDataAgents (DB2 DB2 DPF)

  Platforms	Folders Created	Files Created	Other Changes
  AIX	None	None	Symbolic links to Galaxy libraries will be created under the $DB2_HOME/sqllib/lib directory
  HP-UX	None	None	Symbolic links to Galaxy libraries will be created under the $DB2_HOME/sqllib/lib directory
  Linux	None	None	Symbolic links to Galaxy libraries will be created under the $DB2_HOME/sqllib/lib directory
  Solaris	None	None	Symbolic links to Galaxy libraries will be created under the $DB2_HOME/sqllib/lib directory

How do we discover DB2 instances from the command line?

DB2 instances on the host machine can be found from the command line using the following command:

#>db2ilist
    

How can we prevent log backup failures due to different database and alias names?

To prevent transaction log backup failures, ensure that the name that you select is the actual name of the database and not the alias name for the database. If the database name and the corresponding alias name are different, both names are displayed under Available DB Names; if both names are the same, the name that is displayed is the database name. To find the database names and their aliases, run the following command as a DB2 instance owner on the DB2 client:

C:\Program Files\IBM\SQLLIB\BIN>db2 list db directory

System Database Directory

Number of entries in the directory = 2

Database alias = TESTDBS1

Database name = TDBS1

Local database directory = C:

Can we perform out-of-place restores of databases created in another directory?

If you defined your own directory for the database instead of using the DB2 default location when you created the database, you cannot restore this database to either a new database or another instance in the conventional manner.

Can we perform out-of-place restores of databases with non-default parameter values?

If the DBHEAP, UTIL_HEAP, and/or APP_CTL_HEAP_SZ configuration parameters for the (source) database that you want to restore contain a value other than the default value, you cannot restore this database to either a new database or another instance in the conventional manner.

Can we modify the archive, retrieve and audit error paths after the installation?

Yes, you can modify the archive, retrieve, and audit error paths after the DB2 Data Agent installation. You can use additional settings in the CommCell Console to specify the new directory paths.

These additional settings can also be set in the etc/CommVaultRegistry/Galaxy/Instance001/Db2Agent/.properties file. Add an entry for each directory path that you want to modify.

 
sDB2_ARCHIVE_PATH <new_archive_path>
 sDB2_AUDIT_ERROR_PATH <new_audit_error_path>
 sDB2_RETRIEVE_PATH <new_retrieve_path>
    

Example

sDB2_ARCHIVE_PATH /database/db2/GDB2ARCHIVE
 sDB2_AUDIT_ERROR_PATH /database/db2/GDB2ERROR
 sDB2_RETRIEVE_PATH /database/db2/GDB2RETRIEVE
    

Note: Set the permissions for the archive, audit error and retrieve paths with the following permissions

Before you create the new path, move all logs files in the old log directory to the new log directory.

A log at the following location:

/db2inst3/DBNOAUTO/NODE0000/C0000007/S0000011.LOG

is moved to

/db2inst3/DBNOAUTO/NODE0000/C0000036/S0000011.LOG

Why is a command line restore job submitted during an online full backup?

From DB2 V9.5 and higher, full ONLINE DB2 database backup requires the transactions logs to make it transitionally consistent backup image, even though you already have a log backup. If a transaction is spread across many log files,DB2 will try to include all the required logs in the FULL Backup. If the logs are already backed up and deleted, DB2 will submit a restore request to Media Manager.

This behavior is controlled by NUM_LOG_SPAN - a database configuration parameter. Make sure that this parameter is configured as per the DB2 guide lines (not infinite).

Can we restore a job that had started automatically and on waiting mode for available resource if that resource is used by data restore?

when you try to restore a database after updating the transaction logs with new changes in a database, the new restore job automatically starts and hangs displaying the status as waiting with the following error:

The media is already reserved by some other jobs

The main restore job is not yet completed, so the current restore job will wait for the main restore job to complete.

Configure the bDisableRestoreCaching additional setting's value as 1 on CommServe under the Resource Manager.

Can we use multiplexed backup jobs for restoring to a disk?

No. Currently, we do not support restoring a multiplexed backup jobs to a disk.

Is there a problem if there are more backup images than sessions?

If you run a DB2 version that is higher than 10.5.8, the DB2 software creates more images than sessions (streams) for data backups. For example, if you have a single session, there will be 2 backup images .

How is the number of streams used for a DB2 DPF snapshot, backup copy, or auxiliary copy restore determined?

Snapshot restores of the database may run with fewer streams than the number of streams you choose. This is because the number of streams for a snapshot restore is determined by the number of files to be restored. The snapshot restore procedure makes the best effort to increase the number of streams used for the restore and to therefore reduce the restore time. However, it is limited by the number of files to be restored. A restore of a log file snapshot is limited to one stream, since log files are relatively small.

The number of streams used for a restore from a backup copy or auxiliary copy is the lesser of the following criteria:

• The number of streams used to make the backup copy or auxiliary copy

• The number of streams you specify in the GUI used to start the restore