public inbox for [email protected]  

help / color / mirror / Atom feed

row from index tuple
3+ messages / 2 participants
[nested] [flat]

* row from index tuple
@ 2026-02-04 16:09 Scott Ribe <[email protected]>
  2026-02-04 17:23 ` Re: row from index tuple Ron Johnson <[email protected]>
  0 siblings, 1 reply; 3+ messages in thread

From: Scott Ribe @ 2026-02-04 16:09 UTC (permalink / raw)
  To: Pgsql-admin <[email protected]>

How can I get from an index tuple to the row?

The situation is that we're seeing deadlocks on inserts, because two different sources are sending rows that have the same unique constraint, when they should not be doing that. Would love to get to an example row to hunt this down further.

--
Scott Ribe
[email protected]
https://www.linkedin.com/in/scottribe/








^ permalink  raw  reply  [nested|flat] 3+ messages in thread

* Re: row from index tuple
  2026-02-04 16:09 row from index tuple Scott Ribe <[email protected]>
@ 2026-02-04 17:23 ` Ron Johnson <[email protected]>
  2026-02-04 17:42   ` Re: row from index tuple Scott Ribe <[email protected]>
  0 siblings, 1 reply; 3+ messages in thread

From: Ron Johnson @ 2026-02-04 17:23 UTC (permalink / raw)
  To: Pgsql-admin <[email protected]>

On Wed, Feb 4, 2026 at 11:09 AM Scott Ribe <[email protected]>
wrote:

> How can I get from an index tuple to the row?
>
> The situation is that we're seeing deadlocks on inserts, because two
> different sources are sending rows that have the same unique constraint,
> when they should not be doing that. Would love to get to an example row to
> hunt this down further.
>

When I see deadlocks, both records are in the log file.  You don't see
them, there are "?" placeholders. something else?

-- 
Death to <Redacted>, and butter sauce.
Don't boil me, I'm still alive.
<Redacted> lobster!


^ permalink  raw  reply  [nested|flat] 3+ messages in thread

* Re: row from index tuple
  2026-02-04 16:09 row from index tuple Scott Ribe <[email protected]>
  2026-02-04 17:23 ` Re: row from index tuple Ron Johnson <[email protected]>
@ 2026-02-04 17:42   ` Scott Ribe <[email protected]>
  0 siblings, 0 replies; 3+ messages in thread

From: Scott Ribe @ 2026-02-04 17:42 UTC (permalink / raw)
  To: Ron Johnson <[email protected]>; +Cc: Pgsql-admin <[email protected]>

> On Feb 4, 2026, at 10:23 AM, Ron Johnson <[email protected]> wrote:
> 
> When I see deadlocks, both records are in the log file.  You don't see them, there are "?" placeholders. something else?

I thought so, but it had been a long time, and at a different job.

So, these logs are piped into Splunk, and are not directly available. Splunk treats each log line as a separate entry, so searching on "deadlock" was not showing me what I expected. We found the information by catching one in pg_locks, and by searching on timestamp in splunk--at the same time actually, different people on the call...

Anyway, we figured out the upstream misconfiguration which was giving us duplicate records on different streams (coming from RabbitMQ).







^ permalink  raw  reply  [nested|flat] 3+ messages in thread

end of thread, other threads:[~2026-02-04 17:42 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2026-02-04 16:09 row from index tuple Scott Ribe <[email protected]>
2026-02-04 17:23 ` Ron Johnson <[email protected]>
2026-02-04 17:42   ` Scott Ribe <[email protected]>

This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox