Redis old-master starts as master before accepting CONFIG REWRITE changes

In a 3 node setup (3 redis/3 sentinel) if the master node process is stopped, a new master gets promoted, all good so far, but if the old-master is started it starts as a master for a short time before syncing/receiving the CONFIG REWRITE adding the replicaof <IP> <PORT>, so for some seconds there are more than 1 masters.

Is there a way to start the node and wait for possible CONFIG REWRITE changes before bringing the service up?