Re: BUG #15577: Query returns different results when executed multiple times

Thomas Munro <thomas.munro@enterprisedb.com>

From: Thomas Munro <thomas.munro@enterprisedb.com>
To: Bartosz Polnik <bartoszpolnik@gmail.com>
Cc: David Rowley <david.rowley@2ndquadrant.com>, PostgreSQL mailing lists <pgsql-bugs@lists.postgresql.org>
Date: 2019-01-07T19:59:18Z
Lists: pgsql-bugs
On Tue, Jan 8, 2019 at 1:56 AM Bartosz Polnik <bartoszpolnik@gmail.com> wrote:
> I've modified the script:
> 1. I Added SET force_parallel_mode = on;:

What if you try that with your original query in the database that
shows the problem?  Or alternatively SET parallel_leader_participation
= on.  What I'm wondering is whether the parallel worker is
consistently getting the same result, but it's a different result from
the leader, and the variation comes from the race between leader and
worker, or whether the instability comes from somewhere else.

-- 
Thomas Munro
http://www.enterprisedb.com


Commits

  1. Avoid sharing PARAM_EXEC slots between different levels of NestLoop.