Re: More issues with pg_verify_checksums and checksum verification in base backups

On Tue, Aug 6, 2019 at 6:07 PM Stephen Frost <sfrost@snowman.net> wrote:

> Greetings,
>
> * Andres Freund (andres@anarazel.de) wrote:
> > On 2019-08-06 10:58:15 -0400, Stephen Frost wrote:
> > > * Michael Banck (michael.banck@credativ.de) wrote:
> > > > Independently of the whitelist/blacklist question, I believe
> > > > pg_checksums should not error out as soon as it encounters a weird
> looking
> > > > file, but either (i) still checksum it or (ii) skip it? Or is that
> to be
> > > > considered a pilot error and it's fine for pg_checksums to fold?
> > >
> > > imv, random files that we don't know about are exactly 'pilot error' to
> > > be complained about..  This is exactly why the whitelist idea falls
> > > over.
> >
> > I still think this whole assumption is bad, and that you're fixing
> > non-problems, and creating serious usability issues with zero benefits.
>
> I doubt we're going to get to agreement on this, unfortunately.
>

When agreement cannot be found, perhaps a parameter is in order?

That is, have the tool complain about such files by default but with a HINT
that it may or may not be a problem, and a switch that makes it stop
complaining?

-- 
 Magnus Hagander
 Me: https://www.hagander.net/ <http://www.hagander.net/>
 Work: https://www.redpill-linpro.com/ <http://www.redpill-linpro.com/>