-
Hi Nikee, thanks for the reply. Unfortunately, this logic doesn’t actually make sense. If the database is significantly scaled down and you are failing over to it, one would assume that your traffic levels would also return to normal levels. Given the degraded database state you would be creating a situation where you will quickly have another outage. While the suggested correct action will accomplish an RTO of 5 minutes on paper, in reality it will just start to fail shortly after failover as traffic increases and web servers scale up. No one who actually wants to gracefully recover from a outage to another region would accept this. A slightly longer RTO and scaling up the database would be more appropriate. Or a hot standby. The increased RTO time would have to be balanced against the cost of running a hot standby vs warm which would be a decision for the business to make.
What good is an RTO of 5 minutes if it just starts failing shortly after recovery in the new region?
