>> Should replication cause corruption on the secondary when stopping/starting
>> the primary? (pg 8.3.12, windows 2008 R2 on both servers)
>
> No, it shouldn't. Any duplicate keys would represent a serious error.
>
> It sounds like you're using warm standby, but when you say run
> pg_start_backup() AFTER each nightly backup I admit to being confused.
>
Thanks for your response. Perhaps a quick process flow would help clarify:
Scenario 1 (no errors):
a) create warm standby and let run throughout the day (works great)
b) at the end of the day, trigger the secondary and run a reindex for
testing (no errors)
Scenario 2 (errors):
a) create warm standby and let run throughout the day (works great)
b) nightly backup: shutdown pg on primary, do a file system copy (for
backup later), start pg again on primary
c) the next morning, trigger the secondary and run a re-index for
testing (ERRORS as described in thread)
Side note: the data copied in 2.b is fine and also passes a full re-index.
Scenario 3 (work around - not a very good one):
a) create warm standby and let run throughout the day (works great)
b) nightly backup: shutdown pg on primary, do a file system copy (for
backup later), start pg again on primary
c) the next morning, re-create the warm standby (this is where I may
have confused you with doing a pg_start_backup after nightly backups)
Thanks!
