When your Oracle database DR strategy makes a difference

For on-premise customers, setting up a DR solution for their Oracle database can be seen as a complex and pricy effort. And sometimes, you might even never have to use it.

But what if you do?

At a longtime customer. we did setup Dataguard between 2 datacenters and configured TAF (Transparent Application Failover) for the application connectivity since years now.

We rehearse the switchover procedure every time we patch the database (Standby First).

Last week, as part of an infrastructure move, a piece of production hardware didn’t restart.

The production databases had been switched over to the Dataguard server before the move, and, thanks to TAF, the applications were running without interruption.

We had to rebuild from scratch the initial Oracle production infrastructure, clone the DB online to recreate the standby databases (RMAN). It took us a few days. We ensured that the DG server was coping with the load and that backups were copied to external storage.

Then, with a simple switchover, the DG broker brought back everything in sync.

The application switched back following the active service defined in TAF.

Applications suffered no downtime nor data losses.

Lessons learned :

-      Build a robust DR

-      Always think about the application DR, not only the database

-      Rehearse and document the switchover/failover procedures