Re: Add the ability to limit the amount of memory that can be allocated to backends.

From: Reid Thompson <reid(dot)thompson(at)crunchydata(dot)com>
To: pgsql-hackers(at)lists(dot)postgresql(dot)org
Cc: reid(dot)thompson(at)crunchydata(dot)com
Subject: Re: Add the ability to limit the amount of memory that can be allocated to backends.
Date: 2022-09-07 00:17:25
Message-ID: 559ed491851b2e1c07d1e7e7c12883702430f57d.camel@crunchydata.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Thu, 2022-09-01 at 11:48 +0900, Kyotaro Horiguchi wrote:
> >
> > The patch seems to limit both of memory-context allocations and DSM
> > allocations happen on a specific process by the same budget. In the
> > fist place I don't think it's sensible to cap the amount of DSM
> > allocations by per-process budget.
> >
> > DSM is used by pgstats subsystem. There can be cases where pgstat
> > complains for denial of DSM allocation after the budget has been
> > exhausted by memory-context allocations, or every command complains
> > for denial of memory-context allocation after once the per-process
> > budget is exhausted by DSM allocations. That doesn't seem
> > reasonable.
> > regards.

It's intended as a mechanism for administrators to limit total
postgresql memory consumption to avoid the OOM killer causing a crash
and restart, or to ensure that resources are available for other
processes on shared hosts, etc. It limits all types of allocations in
order to accomplish this.  Our documentation will note this, so that
administrators that have the need to set it are aware that it can
affect all non-auxiliary processes and what the effect is.

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Reid Thompson 2022-09-07 00:25:06 Re: Add the ability to limit the amount of memory that can be allocated to backends.
Previous Message Jaime Casanova 2022-09-06 23:40:49 START_REPLICATION SLOT causing a crash in an assert build