Horton

Re: Improve pg_sync_replication_slots() to wait for primary to advance

shveta malik <shveta.malik@gmail.com>

View thread
From: shveta malik <shveta.malik@gmail.com>
To: Ajin Cherian <itsajin@gmail.com>
Cc: Chao Li <li.evan.chao@gmail.com>, Amit Kapila <amit.kapila16@gmail.com>, Ashutosh Bapat <ashutosh.bapat.oss@gmail.com>, Japin Li <japinli@hotmail.com>, Ashutosh Sharma <ashu.coek88@gmail.com>, PostgreSQL mailing lists <pgsql-hackers@postgresql.org>, shveta malik <shveta.malik@gmail.com>
Date: 2025-12-10T04:05:01Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Enhance slot synchronization API to respect promotion signal.

  2. Fix inconsistent elevel in pg_sync_replication_slots() retry logic.

  3. Refactor slot synchronization logic in slotsync.c.

  4. Fix intermittent BF failure in 040_standby_failover_slots_sync.

  5. Add retry logic to pg_sync_replication_slots().

  6. Fix LOCK_TIMEOUT handling in slotsync worker.

  7. Add slotsync skip statistics.

On Wed, Dec 10, 2025 at 8:10 AM Ajin Cherian <itsajin@gmail.com> wrote:
>
> On Wed, Dec 10, 2025 at 1:29 PM Chao Li <li.evan.chao@gmail.com> wrote:
> >
> > Hi Ajin,
> >
> > I’d like to revisit this patch, but looks like 04396eacd3faeaa4fa3d084a6749e4e384bdf0db has some conflicts to this patch. So can you please rebase this patch?
> >
> > Best regards,
> > --
>
> It's been rebased. Have a look at the latest version.
>

Few comments on 001:

1)
/*
* Emit an error if a promotion or a concurrent sync call is in progress.
* Otherwise, advertise that a sync is in progress.
*/
static void
check_and_set_sync_info

We need to change this comment because now this function does not
handle promotion case.

2)
+  if (sync_process_pid!= InvalidPid)
+    kill(sync_process_pid, SIGUSR1);

We need to have space between sync_process_pid and '!='

3)
+ * Exit or throw errors if relevant GUCs have changed depending on whether

errors->error

4)
In slotsync_reread_config(), even when we mark parameter_changed=true
in the first if-block, we still go to the second if-block which was
not needed.  So shall we make second if-block as else-if to avoid
this? Thoughts?

5)
As discussed in [1], we can make this change in ProcessSlotSyncInterrupts():

'replication slot synchronization worker is shutting down because
promotion is triggered'
to
'replication slot synchronization worker will stop because promotion
is triggered'

[1]: https://www.postgresql.org/message-id/6AE56C64-F760-4CBD-BABF-72633D3F7B5E%40gmail.com

thanks
Shveta