Re: memory problems and crash of db when deleting data from table with thousands of partitions

From: Tomas Vondra <tomas(dot)vondra(at)2ndquadrant(dot)com>
To: Josef Machytka <josef(dot)machytka(at)gmail(dot)com>
Cc: pgsql-bugs(at)lists(dot)postgresql(dot)org
Subject: Re: memory problems and crash of db when deleting data from table with thousands of partitions
Date: 2019-10-28 16:24:46
Message-ID: 20191028162446.wlktskknjpsrj4rt@development
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

On Mon, Oct 28, 2019 at 04:18:59PM +0100, Josef Machytka wrote:
>Thank you for your email, FYI - we now did tests with PostgreSQL 12 and
>unfortunately it is also not able to handle to our case. Only difference is
>that PG 12 is not killed by OOM killer and even does not crash - which is
>good. But it reports error "out of memory" and stops the statement. So at
>least it looks like much more stable then PG 11.
>

Hmmm, this seems a bit weird to me:

MessageContext: 52197329840 total in 8274 blocks; 75904 free (9 chunks);
52197253936 used

That context is generally meant for parse trees and other long-lived
stuff, and I wouldn't expect it to grow to 52GB of data, even if there
are many many partitions.

I wonder if this might be just another manifestation of the memory leak
from [1]. Can you provide a self-contained reproducer, i.e. a script I
could use to reproduce the issue?

[1] https://www.postgresql.org/message-id/20191024221758.vfv2enubnwmy3deu@development

regards

--
Tomas Vondra http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services

In response to

Responses

Browse pgsql-bugs by date

  From Date Subject
Next Message Juan José Santamaría Flecha 2019-10-28 17:13:58 Re: BUG #15858: could not stat file - over 4GB
Previous Message Stephen Frost 2019-10-28 15:47:54 Re: BUG #16079: Question Regarding the BUG #16064