This might have slipped under your radar – but there’s been a number of updates that affect SRM in the last couple...
Firstly, Dell has re-released their Site Recovery Adapter (SRA) for SRM 5.0 – officially, it's called EQL 2.1 SRA. It includes one vitally important update. With Dell Equallogic there was an issue with the failover and failback process. What would happen is that you would lose the “schedule” of replication – that meant after failback, you’d need to remember to go into Group Manager, and attach the schedule again to get replication occurring every 15mins or what ever you had configured for your RPO.
Secondly, SRM 5.01 introduces a new “fast failover” feature or what might better be described as “Forced Recovery” option. One of things that SRM 5.0 tries to do is synch the arrays at the Protected (Production) Site with those at the Recovery (DR) site – additionally, it tries to gracefully shutdown VMs in the Protected Site. Of course, if there is a smoking crater in the Protected Site – there really isn’t anything to synch or shutdown anymore. As a consequence this could lead to quite long time-out values – as SRM in the Recovery Site – tries and then repeatedly fails to communicate to the Protected Site – which like the parrot is no-more…
The new “Forced Recovery” option is not enabled by default. You must first enable the option in “Advanced Settings” held in the “Recovery” section called “Recovery.forceRecovery”:
and then the option will show itself when you select the “Disaster Recovery” option in the new SRM UI.
Technically, this only needs to be enabled at the Recovery Site, but if you want the option to be available in a failback situation as well (You’d be unlucky to have run SRM twice in two different locations affected by a different disaster!), then you could enable it on all your sites managed by SRM.