Re: Collect statistics about conflicts in logical replication - Mailing list pgsql-hackers

On Thu, Aug 29, 2024 at 11:06 AM Zhijie Hou (Fujitsu)
<houzj.fnst@fujitsu.com> wrote:
>
>
> Agreed. Here is new version patch which change the names as suggested. I also
> rebased the patch based on another renaming commit 640178c9.
>

Thanks for the patch. Few minor things:

1)
conflict.h:
 * This enum is used in statistics collection (see
 * PgStat_StatSubEntry::conflict_count) as well, therefore, when adding new
 * values or reordering existing ones, ensure to review and potentially adjust
 * the corresponding statistics collection codes.

--We shall mention PgStat_BackendSubEntry as well.

026_stats.pl:
2)
# Now that the table is empty, the
# update conflict (update_existing) ERRORs will stop happening.

--Shall it be update_exists instead of update_existing here:

3)
This is an existing comment above insert_exists conflict capture:
# Wait for the apply error to be reported.

--Shall we change to:
# Wait for the subscriber to report apply error and insert_exists conflict.

thanks
Shveta