Once a failover has occurred, Data Mover does not automatically fail back to the old active daemon if it is restarted. To fail back to the old-restored active daemon after a failover has occurred, do the following:
- Log on to the old active daemon server and run dmcluster setmaster.This starts up the sync service associated with this daemon in active mode and the sync service on the old daemon server in standby mode. Once the sync service is started in standby mode on the old active daemon, it receives all the updates that occurred on the new active repository while the old active was down and catches up with the new active.
- Wait a few minutes to allow the old active to catch up with the new active, and then check the dmSync.log on the standby (old active) to determine when the old active has caught up with the new active.
- When the old active repository is in sync with the new active repository, set the old active back as the new active by running the dmcluster setmaster command on the old active.
- Run the dmcluster status command to verify the old active has been restored successfully.Trying to start the daemon on the old-restored active using the dmdaemon service script (/etc/init.d/dmdaemon start) causes the monitoring service to assume that the daemon has incorrectly started on the old active and shut it down.