RE: Failed transaction statistics to measure the logical replication progress

On Tuesday, February 22, 2022 10:15 AM Tang, Haiying/唐 海英 <tanghy(dot)fnst(at)fujitsu(dot)com> wrote:
> On Mon, Feb 21, 2022 11:46 AM osumi(dot)takamichi(at)fujitsu(dot)com
> <osumi(dot)takamichi(at)fujitsu(dot)com> wrote:
> >
> > On Saturday, February 19, 2022 12:00 AM osumi(dot)takamichi(at)fujitsu(dot)com
> > <osumi(dot)takamichi(at)fujitsu(dot)com> wrote:
> > > On Friday, February 18, 2022 3:34 PM Tang, Haiying/唐 海英
> > > <tanghy(dot)fnst(at)fujitsu(dot)com> wrote:
> > > > On Wed, Jan 12, 2022 8:35 PM osumi(dot)takamichi(at)fujitsu(dot)com
> > > > <osumi(dot)takamichi(at)fujitsu(dot)com> wrote:
> > > > 4) I noticed that the abort_count doesn't include aborted
> > > > streaming transactions.
> > > > Should we take this case into consideration?
> > > Hmm, we can add this into this column, when there's no objection.
> > > I'm not sure but someone might say those should be separate columns.
> > I've addressed this point in a new v23 patch, since there was no
> > opinion on this so far.
> >
> > Kindly have a look at the attached one.
> >
>
> Thanks for updating the patch.
>
> I found a problem when using it. When a replication workers exits, the
> transaction stats should be sent to stats collector if they were not sent before
> because it didn't reach PGSTAT_STAT_INTERVAL. But I saw that the stats
> weren't updated as expected.
>
> I looked into it and found that the replication worker would send the transaction
> stats (if any) before it exits. But it got invalid subid in
> pgstat_send_subworker_xact_stats(), which led to the following result:
>
> postgres=# select pg_stat_get_subscription_worker(0, null);
> pg_stat_get_subscription_worker
> ---------------------------------
> (0,,2,0,0,,,,0,"",)
> (1 row)
>
> I think that's because subid has already been cleaned when trying to send the
> stats. I printed the value of before_shmem_exit_list, the functions in this list
> would be called in shmem_exit() when the worker exits.
> logicalrep_worker_onexit() would clean up the worker info (including subid),
> and
> pgstat_shutdown_hook() would send stats if any. logicalrep_worker_onexit()
> was called before calling pgstat_shutdown_hook().
>
> (gdb) p before_shmem_exit_list
> $1 = {{function = 0xa88f1e <pgstat_shutdown_hook>, arg = 0}, {function =
> 0xb619e7 <BeforeShmemExit_Files>, arg = 0}, {function = 0xb07b5c
> <ReplicationSlotShmemExit>, arg = 0}, {
> function = 0xabdd93 <logicalrep_worker_onexit>, arg = 0}, {function =
> 0xe30c89 <ShutdownPostgres>, arg = 0}, {function = 0x0, arg = 0} <repeats 15
> times>}
>
> Maybe we should make some modification to fix it.
Thank you for letting me know this issue.
I'll investigate this and will report the result.

Best Regards,
Takamichi Osumi