Re: [PATCH] Release replication slot on error in SQL-callable slot functions

From: Kyotaro Horiguchi <horikyota(dot)ntt(at)gmail(dot)com>
To: bharath(dot)rupireddyforpostgres(at)gmail(dot)com
Cc: sawada(dot)mshk(at)gmail(dot)com, shveta(dot)malik(at)gmail(dot)com, ashu(dot)coek88(at)gmail(dot)com, houzj(dot)fnst(at)fujitsu(dot)com, satyanarlapuram(at)gmail(dot)com, masao(dot)fujii(at)gmail(dot)com, vignesh21(at)gmail(dot)com, pgsql-hackers(at)lists(dot)postgresql(dot)org
Subject: Re: [PATCH] Release replication slot on error in SQL-callable slot functions
Date: 2026-08-27 05:09:24
Message-ID: 20260827.140924.1018328804092095493.horikyota.ntt@gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hello,

At Wed, 26 Aug 2026 18:36:00 -0700, Bharath Rupireddy <bharath(dot)rupireddyforpostgres(at)gmail(dot)com> wrote in
> > In short, having just the slot release in the subxact path gives the
> > same error behavior, is simple to reason about, and fixes the crash
> > reported in this thread.
> >
> > Thoughts?
>
> I removed the temporary slot handling in the code and tests and
> attached the simplified v14 patch. I also adjusted the docs to mention
> that temporary slots are not automatically dropped when an error is
> raised and caught in a subtransaction.

Sorry for chiming in this late, but looking at the patch, I wonder
whether this should be handled by ResourceOwner rather than
AtEO(Sub)Xact. In particular, the fact that the ownership is simply
transferred to the parent subtransaction on commit makes ResourceOwner
seem like a natural fit here. What do you think?

Regards,

--
Kyotaro Horiguchi
NTT Open Source Software Center

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message vignesh C 2026-08-27 05:11:26 Re: Proposal: Conflict log history table for Logical Replication
Previous Message Tom Lane 2026-08-27 05:07:22 Re: Tracking role modification timestamps in pg_authid / pg_roles