2.7

Table Of Contents
Database End of Life and Backups
When you decommission and delete a database, you decide whether to retain its backup files. The decision is
based on your site's policies and whether you might need the database in the future.
When you delete a database, you can retain all external backups. The backups expire at the end of the normal
retention period. It is good practice to take a final backup of a database and specify the final backup's retention
period before you delete a database. If you retain the external backups, the snapshots and the executable
instance of the database are deleted. If the deleted database had point-in-time recovery enabled, all the archived
write-ahead log (WAL) segments are deleted as well. This means that the only way to recover the database is
by using the external backups. You cannot recover the database using snapshots or point-in-time recovery.
If you do not retain the external backups, the database and its associated backups, snapshots, and WALs are
deleted. In addition, the database resources are released, and the database cannot be restored.
Perform Point-in-time Recovery of Management Server Database
You can perform a point-in-time recover of the Management Server database if you determine that the database
is corrupted.
Prerequisites
n
Verify that VMware Data Recovery (VDR) is deployed with vCenter and set up with the vSphere client.
n
Verify that regular backups are being performed on the Data Director vApp using VDR.
n
Before starting point-in-time recovery on the Management Server database, take a full backup of the Data
Director vApp. If the target time you choose for your point-in-time recovery results in recovery failure,
you can restore the management server to this backup and perform a point-in-time recovery with an earlier
target time.
Procedure
1 In the vSphere client, shut down the Management Server virtual machine.
2 Restore only the database disk (management_server_2.vmdk) from a backup using VDR.
3 Select Edit Settings > Options > vApp Options > Properties, and select the vApp
Management server recovery flag.
4 Turn on the Management server.
5 Open the console and log in as aurora, or log in from root as su aurora, and type the following command.
/opt/aurora/sbin/recover_cmsdb.sh
6 Enter a target time stamp when prompted.
The time must be after the time of the VDR backup that you use to restore the Central Management Server
(CMS).
7 When the recover finishes, shut down the vApp of Data Director.
8 Deselect the management server recovery flag on the Management Server.
9 Check the Management Server network cards connection in the virtual machine settings, and connect them
manually if necessary.
10 Turn on the Data Director vApp.
VMware vFabric Data Director Administrator and User Guide
160 VMware, Inc.