Re: Change ereport level for QueuePartitionConstraintValidation

On Mon, 15 Jul 2019 at 11:46, Thomas Munro <thomas(dot)munro(at)gmail(dot)com> wrote:
>
> On Tue, Jul 2, 2019 at 12:17 AM Sergei Kornilov <sk(at)zsrv(dot)org> wrote:
> > This change is discussed as open item for pg12. Seems we have nor objections nor agreement. I attached updated version due merge conflict.
> >
> > > Per discussion started here: https://www.postgresql.org/message-id/CA%2BTgmoZWSLUjVcc9KBSVvbn%3DU5QRgW1O-MgUX0y5CnLZOA2qyQ%40mail.gmail.com
>
> I took the liberty of setting this to "Ready for Committer" to see if
> we can get a decision one way or another and clear both a Commitfest
> item and a PG12 Open Item. No committer is signed up, but it looks
> like Amit L wrote the messages in question, Robert committed them, and
> David made arguments for AND against on the referenced thread, so I'm
> CCing them, and retreating to a safe distance.

I think the only argument against it was around lack of ability to
test if the constraint was used to verify no row breaks the partition
bound during the ATTACH PARTITION.

Does anyone feel strongly that we need to the test to confirm that the
constraint was used for this?

If nobody feels so strongly about that then I say we can just push
this. It seems something that's unlikely to get broken, but then you
could probably say that for most things our tests test for.

--
David Rowley http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services