Re: Proposal: Conflict log history table for Logical Replication

From: Zhijie Hou <houzhijie22(at)gmail(dot)com>
To: vignesh C <vignesh21(at)gmail(dot)com>, Nisha Moond <nisha(dot)moond412(at)gmail(dot)com>
Cc: Amit Kapila <amit(dot)kapila16(at)gmail(dot)com>, Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com>, shveta malik <shveta(dot)malik(at)gmail(dot)com>, shihao zhong <zhong950419(at)gmail(dot)com>, Dilip Kumar <dilipbalaut(at)gmail(dot)com>, saurabh singh <saurabh(dot)singh214(at)gmail(dot)com>, Robert Haas <robertmhaas(at)gmail(dot)com>, Peter Smith <smithpb2250(at)gmail(dot)com>, Bharath Rupireddy <bharath(dot)rupireddyforpostgres(at)gmail(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>, shveta malik <shvetamalik(at)gmail(dot)com>
Subject: Re: Proposal: Conflict log history table for Logical Replication
Date: 2026-11-19 14:21:27
Message-ID: 27405573-1bb8-4944-a82b-8f01584c2e81@gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi,

On Thu, Nov19, 2026 at 06:06 PM vignesh C wrote:
> On Wed, 18 Nov 2026 at 15:56, vignesh C <vignesh21(at)gmail(dot)com> wrote:
> >
> > Thanks for the comments and changes, I have merged the changes with
> > slight changes in the 002 patch, there are no changes in 001 and 003
> > patch, it is the same as the v94 version patch. The attached v95
> > version patch has the changes for the same.
>
> Few comments:
> 1. catalogs.sgml is missing the new pg_subscription columns:
> subconflictlogrelid and subconflictlogdest

I think this is existing issue on HEAD, so we can have a separate patch
to fix it.

Other comments look good to me and have been addressed in the new
version patch.

Here is the V96 version which:

1. Addressed above comments.

2. Updated the local_conflicts column doc to drop "existing", as it also
records deleted tuple metadata for update_deleted conflicts (per
Kuroda-San's off-list suggestion).

3. Clarified the replica_identity doc to note that recorded key text
matches a column SELECT for most columns (index opclass input types such
as oid_ops on regclass can format values differently, such as recording
an OID, though casting back to the column's type still resolves to the
target row and does not affect conflict inspection or resolution) (per
Shveta's off-list suggestion).

4. Added a code comment in insert_conflict_log_tuple() explaining why
heap_insert() is used directly instead of table_tuple_insert().

Only 0001 patch is changed.

--
Best Regards,
Zhijie Hou

Attachment Content-Type Size
v96-0001-Record-logical-replication-conflicts-in-the-conf.patch text/plain 133.7 KB
v96-0002-Preserve-conflict-log-destination-and-subscripti.patch text/plain 25.9 KB
v96-0003-Add-conflict-log-table-information-to-describe-s.patch text/plain 80.3 KB

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Zsolt Parragi 2026-11-19 14:26:35 Re: test: avoid redundant standby catchup in 049_wait_for_lsn
Previous Message Heikki Linnakangas 2026-11-19 14:19:52 Re: Fix background writer processing locking in README