> For example, with PG I have to reseed the original master if it comes back online after an outage pg_rewind[1] is supposed to fix this. It's moving into core

> For example, with PG I have to reseed the original master if it comes back online after an outage pg_rewind[1] is supposed to fix this. It's moving into core for 9.5, by the way. [1] https://github.com/vmware/pg_rewind https://github.com/vmware/pg_rewind