Restoring clusters to a consistent point

You can restore the primary server in a high-availability cluster to a consistent point after a failed upgrade.

Prerequisites:

Before you began the upgrade, you must have enabled quick reversion, according to information in Configuring for recovery of restore point data in case an upgrade fails.
The server must be offline.

To restore the primary server in a cluster to a consistent point after a failed upgrade:

Run the onrestorept utility according to these instructions.

Alternatively, if you backed up your primary server or you prepared for using the RSS or HDR secondary as a backup server before you upgraded and the upgrade fails, you can take other steps to restore the cluster.