From: | hubert depesz lubaczewski <depesz(at)depesz(dot)com> |
---|---|
To: | Pavel Stehule <pavel(dot)stehule(at)gmail(dot)com> |
Cc: | Adrian Klaver <adrian(dot)klaver(at)aklaver(dot)com>, PostgreSQL General <pgsql-general(at)lists(dot)postgresql(dot)org>, Chris Wilson <chris+google(at)qwirx(dot)com> |
Subject: | Re: Streaming replica hangs periodically for ~ 1 second - how to diagnose/debug |
Date: | 2025-08-22 15:19:33 |
Message-ID: | aKiKhUQHTNnLDqqb@depesz.com |
Views: | Whole Thread | Raw Message | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On Fri, Aug 22, 2025 at 05:07:09PM +0200, Pavel Stehule wrote:
> > Any idea on what could it be?
> Few years ago I had a similar experience - some sec mystic lags - the
> problem was in virtualization.
While I can't discount this idea, after all it is aws ec2, so virtual
boxes, what makes me wonder is:
in case of this slow situation there is futex call, which takes
a second. But there are no futex calls in any other query processing
logs.
So, while it is obvious that futex "wasted" this time, the question is:
why would pg run futex() for this one query at all?
Best regards,
depesz
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2025-08-22 15:21:22 | Re: Streaming replica hangs periodically for ~ 1 second - how to diagnose/debug |
Previous Message | Pavel Stehule | 2025-08-22 15:07:09 | Re: Streaming replica hangs periodically for ~ 1 second - how to diagnose/debug |