Reid Thompson <Reid.Thompson@omnicell.com> writes:
> We have a planned upgrade that would permanentlty remedy this.
> regarding the below errors on our PG 9.6.x instance.
> 2020-09-28 09:08:15.741 EDT,,,26212,,5f71e03f.6664,1,,2020-09-28 09:08:15 EDT,250/9136136,0,ERROR,XX001,"found xmin
2675436435from before relfrozenxid 321165377",,,,,"automatic vacuum of table ""postgres.pg_catalog.pg_authid""",,,,""
> 2020-09-28 09:08:15.751 EDT,,,26212,,5f71e03f.6664,2,,2020-09-28 09:08:15 EDT,250/9136138,0,ERROR,XX001,"found xmin
2675019557from before relfrozenxid 321165377",,,,,"automatic vacuum of table
""postgres.pg_catalog.pg_auth_members""",,,,""
> I know that a db restart will reset the 'counter' (reset the
> pg_internal.init files??) on this issue. We would rather avoid a
> restart if possible. Some research on the internet found a page
> suggesting that removal of pg_internal.init file(s) on a running system
> would provide the same temporary resolution as a restart??? Is this a
> valid temporary work around for this issue? If it is, is there a
> particular pg_internal.init file that needs to be removed, or just
> remove all pg_internal.init that can be found?
I'm a bit dubious that that'd actually help, but it's perfectly safe
if you want to try it. pg_internal.init is just a cache file that
will be rebuilt if it's missing.
regards, tom lane