Re: Proposal: Conflict log history table for Logical Replication

On Thu, Jul 9, 2026 at 5:11 PM Kedar Anavardekar
<kedar(dot)anwardekar3(at)gmail(dot)com> wrote:
>
> > I rebased the remaining patches on top of HEAD. So far, I have run
> > pgindent and completed the doc merge for 0001. The 0002 and 0003 are
> > just rebased.
> >
>
> Hi Dilip,
>
> I was testing on the master branch (commit
> a5918fddf10d297c70f7ec9067e9177e0be6d520 -
> "Allow logical replication conflicts to be logged to a table") for
> basic CLT operations,
> and ran into an issue I'd like to check with you.
> I ran the SQL from the shell script and expected an insert conflict to
> get logged in the CLT. I created the subscription as follows:
>
> create subscription sub connection '...' publication pub
> WITH(conflict_log_destination='table');
> NOTICE: created conflict log table
> "pg_conflict.pg_conflict_log_16414" for subscription "sub"
> NOTICE: created replication slot "sub" on publisher
> CREATE SUBSCRIPTION
>
> However, I noticed that the conflicts are still being logged to the
> log file rather than to pg_conflict.pg_conflict_log_16414.
>
> Would you be able to let me know if my test setup was correct, or if
> I'm missing a step somewhere? Any pointers would be much appreciated.
>

The concerned commit only creates the infrastructure required for
conflict logging, it does not enable conflict logging yet. Please see
the last paragraph of the commit message:

"This commit only establishes the conflict log table along with its
creation, cleanup, and protection; recording the conflicts detected
during apply into the table will be handled in a follow-up commit."

The patch for conflict insertion is in progress in this thread. See
v64 attached in [1].

[1]: https://www.postgresql.org/message-id/CAFiTN-vLKRRSDMwPGMvcBM1CRntXbsHARU-0e76PDk%3D-k5RO6A%40mail.gmail.com

thanks
Shveta