Horton

Re: strange parallel query behavior after OOM crashes

Robert Haas <robertmhaas@gmail.com>

View thread View on postgr.es
From: Robert Haas <robertmhaas@gmail.com>
To: Tomas Vondra <tomas.vondra@2ndquadrant.com>
Cc: Kuntal Ghosh <kuntalghosh.2007@gmail.com>, Thomas Munro <thomas.munro@enterprisedb.com>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2017-04-05T13:39:07Z
Lists: pgsql-hackers 
On Tue, Apr 4, 2017 at 1:52 PM, Tomas Vondra
<tomas.vondra@2ndquadrant.com> wrote:
> In any case, the comment right before BackgroundWorkerArray says this:
>
>  * These counters can of course overflow, but it's not important here
>  * since the subtraction will still give the right number.
>
> which means that this assert
>
> +       Assert(BackgroundWorkerData->parallel_register_count >=
> +               BackgroundWorkerData->parallel_terminate_count);
>
> is outright broken, just like any other attempts to rely on simple
> comparisons of these two counters, no?

Yeah, that's busted.  Good catch.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

Commits

  1. Add an Assert() to max_parallel_workers enforcement.

  2. Fix confusion of max_parallel_workers mechanism following crash.

  3. Add max_parallel_workers GUC.