Simplified Disaster Recovery: Enable Disaster Recovery

Use this procedure if you have a pre-existing standalone deployment of FTL (no DR), and you want to add DR capabilities. You must be using a non-default persistence cluster for your applications. DR is not allowed for the default persistence cluster.

This one-time procedure requires a rolling restart of the FTL servers.

1. Create FTL server yaml configuration files for the DR site. Refer to Simplified Disaster Recovery: Setup. Be sure to add standby persistence services; they wait for you to update the FTL realm configuration, later in this procedure.

2. Start FTL servers at the DR site. These FTL servers are idle until DR has been enabled at the primary site.

3. Update FTL server yaml configuration files for the primary site. Refer to Simplified Disaster Recovery: Setup.

   You must add the dr parameter. The auto.init.primary.on.no.contact parameter is optional. The auto.init.primary.on.first.startup flag is irrelevant because FTL server already has state on disk.

4. Roll the FTL servers at the primary site.

   Restart each FTL server, one-by-one. After each restart, ensure that all persistence services at the primary site are operational and that client applications continue to function.

   Note that persistence services at the DR site are not online yet.

5. Verify the operating mode of FTL servers at the primary and DR sites. See Monitoring for Disaster Recovery.

6. Update the FTL realm configuration. For each persistence cluster used by your applications, you must add a standby server set (containing the persistence services you specified in the DR site yaml configuration file). You must also enable DR replication for the persistence cluster. See Persistence Configuration for Disaster Recovery.

7. Verify the state of persistence services at the DR site. See Monitoring for Disaster Recovery.