Comment by sofuture

The only bump in service we had was ~30s of downtime when our multi-AZ RDS cut over to the failover instance. :) I think we lucked out a little bit...