Re: [PATCH] Fix stale relation close in sequence synchronization - Mailing list pgsql-hackers

From Ayush Tiwari
Subject Re: [PATCH] Fix stale relation close in sequence synchronization
Date
Msg-id CAJTYsWVjgJZP1F8j7+Bn=nEMDQJGBanwkpWozWm0s8CPceJWsQ@mail.gmail.com
Whole thread
In response to Re: [PATCH] Fix stale relation close in sequence synchronization  (vignesh C <vignesh21@gmail.com>)
Responses Re: [PATCH] Fix stale relation close in sequence synchronization
List pgsql-hackers
Hi,

On Thu, 30 Apr 2026 at 11:27, vignesh C <vignesh21@gmail.com> wrote:

Consider updating the error reported in report_sequence_errors to
convey that the failure could be due to either a missing sequence or
insufficient privileges (e.g., "missing sequence or insufficient
privilege on publisher"). While making this change, also update the
missing_seqs_idx variable name and the associated comments to
accurately reflect the broader scope of the condition.

Agreed. I have updated the warning to say "missing or inaccessible
sequence on publisher", so it covers both cases where the sequence is
actually missing and cases where the publisher-side role cannot access
its sequence data.

I also renamed the related variables from missing_* to unavailable_*,
and updated the associated comments and DEBUG message to avoid implying
that the sequence is necessarily absent on the publisher.

The TAP expectations for both the dropped-sequence case and the
insufficient-privileges case have been updated to match the new wording.

Attached is v5 with these changes.

Thoughts?

Regards,
Ayush
Attachment

pgsql-hackers by date:

Previous
From: Ashutosh Bapat
Date:
Subject: Re: [Patch]Add Graph* node support to expression_tree_mutator
Next
From: vignesh C
Date:
Subject: Re: [PATCH] Preserve replication origin OIDs in pg_upgrade