Re: Slow Index Creation, why is it not consuming more memory.

From: Peter Geoghegan <peter(dot)geoghegan86(at)gmail(dot)com>
To: Tory M Blue <tmblue(at)gmail(dot)com>
Cc: "pgsql-performance(at)postgresql(dot)org" <pgsql-performance(at)postgresql(dot)org>
Subject: Re: Slow Index Creation, why is it not consuming more memory.
Date: 2015-12-08 06:40:21
Message-ID: CAEYLb_XBryqSJ9e+4px3sw=Od9wLSB-1hjdvXfAe-1p6ZebrHw@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

On Mon, Dec 7, 2015 at 10:36 PM, Tory M Blue <tmblue(at)gmail(dot)com> wrote:
> What am I not understanding missing?

Yes. There is a hard limit on the number of tuples than can be sorted
in memory prior to PostgreSQL 9.4. It's also the case that very large
work_mem or maintenance_work_mem settings are unlikely to help unless
they result in a fully internal sort.

There is evidence that the heap that tuple sorting uses benefits from
*lower* settings. Sometimes as low as 64MB.

We're working to make this better in 9.6.

--
Regards,
Peter Geoghegan

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message Tory M Blue 2015-12-10 09:12:19 checkpoints, proper config
Previous Message Tory M Blue 2015-12-08 06:36:56 Slow Index Creation, why is it not consuming more memory.