On Fri, Jun 28, 2019 at 07:43:30AM -0400, Brad Nicholson wrote:
> So if all the checksums are being recalculated on the replica, this
> approach should be relatively safe, should it not?
Yep.
> Assuming pg_checksums is doing the right thing (and it looks to me like it
> should be).
The constraint that a cluster needs to be cleanly shut down to be able
to enable checksums with pg_checksums is the actual deal here. After
that of course comes the WAL retention on the primary or in the WAL
archives that a standby would need again to catch up while it was
offline.
--
Michael
