This message was deleted.

I've been a part of organizations which handled DR and BCPs in this manner both as a legacy pattern to be migrated away from and also one in which this was the primary solution for the long-term. In the situation where it was a legacy environment we did no regular DR planning or exercise to stress the system - it was simply an accepted short-term risk. We relied on parallel automation and cluster automation in this situation. In the environment where this was our primary DR/BC strategy we enforced strict codebase parity through system level automation (think - rsync and the like) and log shipping for datastores. Here we led a well orchestrated and documentation driven BC/DR Exercise twice annually. On my platform team we would typically involve 4 - 6 engineers according to on-call schedule and it was a weekend long process including a scripted failover of a percentage of our callcenter to exercise the telecom aspects of the BCP. It was typically run across the entire BU so there were probably about 100 employees engaged across the Enterprise during this timeframe.