Their existing RP replicates from cluster A1 to Cluster B1.
We have configured a new cluster A2 at site A and connected to the RP system. The production CG is now replicating from A1 to both A2 & B1 cluster.
The idea is to use the new leg A1 => A2 for migration. However options we have in admin guide doesn't meet the customer's requirements. Customer wants A2 copy to be production, but keep A1 isolated & intact. But in case of issue, they can still fail back to A1 copy without full sweep.
Test copy -- the A2 copy will be in Virtual Access mode. I don't think we want that.
test a copy & recover -- we don't need recovery.
test a copy & fail over -- this will automatically change the direction of replication and may change A1 image.
Any advise what the proper way to do the migration? Thanks.
Have they looked into direct image access on A2.?, once direct IA is disabled, short init is done between prod and replica. If they don’t want to lose the changes performed while in direct IA on the replica, they can failover to it (or delete the CG if they don’t want to maintain the replication relationship).
Note that this can be done with any replica (local or remote) so having a separate cluster (A2) is not a requirement here.
Hope that helps,
Idan Kentor Tech Staff Engineering Technologist - Data Protection