Is there a new way around the requirement to rebuild your entire replication topology after upgrading versions? (say 9.4 to 9.5) You get a new master ID when ru

Is there a new way around the requirement to rebuild your entire replication topology after upgrading versions? (say 9.4 to 9.5) You get a new master ID when running the initdb step, and doing this throws everything else in the topology off. TIA