Re: pg15b2: large objects lost on upgrade

From: Peter Geoghegan <pg(at)bowt(dot)ie>
To: Robert Haas <robertmhaas(at)gmail(dot)com>
Cc: Andres Freund <andres(at)anarazel(dot)de>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, "Jonathan S(dot) Katz" <jkatz(at)postgresql(dot)org>, Noah Misch <noah(at)leadboat(dot)com>, Andrew Dunstan <andrew(at)dunslane(dot)net>, Bruce Momjian <bruce(at)momjian(dot)us>, Michael Paquier <michael(at)paquier(dot)xyz>, Justin Pryzby <pryzby(at)telsasoft(dot)com>, "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org>, Shruthi Gowda <gowdashru(at)gmail(dot)com>
Subject: Re: pg15b2: large objects lost on upgrade
Date: 2022-08-04 19:52:34
Message-ID: CAH2-WzmmSc1EATof89wBnjwuv8FbKEx6zMK7emMs5=my=p_E4g@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Thu, Aug 4, 2022 at 12:31 PM Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> > What about autoanalyze?
>
> What about it?

It has a tendency to consume an XID, here or there, quite
unpredictably. I've noticed that this often involves an analyze of
pg_statistic. Have you accounted for that?

You said upthread that you don't like "fuzzy" tests, because it's too
easy for things to look like they're working when they really aren't.
I suppose that there may be some truth to that, but ISTM that there is
also a lot to be said for a test that can catch failures that weren't
specifically anticipated. Users won't be running pg_upgrade with
autovacuum disabled. And so ISTM that just testing that relfrozenxid
has been carried forward is more precise about one particular detail
(more precise than alternative approaches to testing), but less
precise about the thing that we actually care about.

--
Peter Geoghegan

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Robert Haas 2022-08-04 20:07:01 Re: [Proposal] Fully WAL logged CREATE DATABASE - No Checkpoints
Previous Message Robert Haas 2022-08-04 19:32:00 Re: pg15b2: large objects lost on upgrade