Re: maintenance_work_mem + create index - Mailing list pgsql-performance

From Uwe Bartels
Subject Re: maintenance_work_mem + create index
Date
Msg-id AANLkTikQO+y2nWz4SjjUMKkGDCVFFL0Z_8bu9poUyk=9@mail.gmail.com
Whole thread Raw
In response to Re: maintenance_work_mem + create index  (Stephen Frost <sfrost@snowman.net>)
Responses Re: maintenance_work_mem + create index
Re: maintenance_work_mem + create index
List pgsql-performance
OK. I didn't now that. Thanks for sharing that information.
Can anybody tell if we have this limitation on maintenance_work_mem as well?

Does anybody know of a solution out of that on Linux?
Or is there a dynamic way to put $PGDATA/base/pgsql_tmp into RAM without blocking it completely like a ram disk?

Best Regards,
Uwe

On 24 March 2011 15:13, Stephen Frost <sfrost@snowman.net> wrote:
Uwe,

* Uwe Bartels (uwe.bartels@gmail.com) wrote:
> So I checked this again and raised afterwards maintenance_work_mem step by
> step up 64GB.
> I logged in via psql, run the following statements
> set maintenance_work_mem = '64GB';

I believe maintenance_work_mem suffers from the same problem that
work_mem has, specifically that PG still won't allocate more than
1GB of memory for any single operation.

       Thanks,

               Stephen

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iEYEARECAAYFAk2LUW8ACgkQrzgMPqB3kigZMwCfUVL/5nSdK5xiV+/SjWB6BG9B
Fm0An2V5Tald8PUYXc5VIuKL/C1WNYTp
=MSxh
-----END PGP SIGNATURE-----


pgsql-performance by date:

Previous
From: Stephen Frost
Date:
Subject: Re: maintenance_work_mem + create index
Next
From: Shaun Thomas
Date:
Subject: Re: maintenance_work_mem + create index