Re: Parallel query execution introduces performance regressions - Mailing list pgsql-bugs

From Andres Freund
Subject Re: Parallel query execution introduces performance regressions
Date
Msg-id 20190401191304.x6ieib4bj73hywa6@alap3.anarazel.de
Whole thread Raw
In response to Re: Parallel query execution introduces performance regressions  (Jinho Jung <jinhojun@usc.edu>)
List pgsql-bugs
Hi,

Please don't top-quote on postgres mailing lists.

On 2019-04-01 15:08:59 -0400, Jinho Jung wrote:
> Sorry for the duplicated thread here and thanks for the reply.
> 
> I have another question regarding parallel execution. We noticed that
> PostgreSQL sometimes launches only one worker. Is that also an intended
> behavior?

Note that even a single worker is a form of parallelism - the "main"
backend also participates in query execution. So yes, that can be a
reasonable behaviour. Also, it might be that query was just finished
faster than a worker was able to launch.

- Andres



pgsql-bugs by date:

Previous
From: Jinho Jung
Date:
Subject: Re: Parallel query execution introduces performance regressions
Next
From: Thomas Munro
Date:
Subject: Re: BUG #15726: parallel queries failed ERROR: invalid name syntaxCONTEXT: parallel worker