Due to some infra set up, master broker goes down and come up withing a fraction of seconds i.e 10 seconds - In that fraction of seconds, the slave broker takes the control as live broker when master goes down and gave the control to master again when it comes up.
It's an extensive operation,having 300+ connections which shift from master to slave and again within a fraction of seconds to master, which we can avoid
The ask here is to implement a retry mechanism in between master goes down and slave becomes live
There should be a retry mechanism with retry interval and retry attempts, so after exhausting that retry attempts(let's say total retrial time T) if the master has not come up then slave should become live and if in that retries slaves gets ack from master that it has come up then slave should not become live rather should give the control to master again
Do we have any such feature?? or Can we implement that?
Story
Major
