On Fri, Jun 11, 2021 at 7:55 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
>
> [...]
>
> If you can afford to wait, it'd be interesting to see what happens.
It's been running for a few days now and heap_blks_vacuumed /
heal_blks_total is around 74% at present. If it continues at the
current rate, we will most probably let it complete. In any case, we
will keep it running over the weekend at least.
> Looking at the code, it looks to me like what's probably happening is
> that those errors are coming from auto-analyze attempting to update
> pg_statistic. That should not, however, prevent auto-vacuum from
> cleaning things up and advancing the transaction threshold limit,
> because the catalog updates that are needed for that are nontransactional
> and so don't need to consume XIDs.
>
> I wonder whether we should disable auto-analyze once we get into
> vacuum-for-wraparound mode. But in any case, it'd be useful to
> see whether there's an actual bug in there.
>
> It'd also be useful to know exactly which PG version you are running.
PG version 9.6.22. Missed mentioning it earlier, sorry about that.
>
> regards, tom lane
Thanks,
Vineet