From: | Rémi Cura <remi(dot)cura(at)gmail(dot)com> |
---|---|
To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Cc: | PostgreSQL General <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: GIST index : order Hack : getting the order used by CLUSTER .. USING my_index |
Date: | 2013-10-24 16:49:48 |
Message-ID: | CAJvUf_vTgcAxpzsCaMsQjf+fOxFwv0tt0LC7s0gn6DFirp0aig@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Great,
thanks.
Now you say that I never saw any improvement when clustering table with
gist.
You just saved me a lot of unnecessary queries :-)
Cheers,
Rémi-C
2013/10/24 Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
> =?UTF-8?Q?R=C3=A9mi_Cura?= <remi(dot)cura(at)gmail(dot)com> writes:
> > I don't understand how the CLUSTER .. USING index command work then.
> > It is supposed to rewrite on disk following index order. Does it do
> nothing
> > for GIST index?
>
> Nobody has ever demonstrated that CLUSTER has any value for anything
> except btree indexes. It seems likely to me that it'd actually be
> counterproductive for indexes like GiST, which depend on data arriving in
> random order for the highest index levels to end up well-distributed.
>
> regards, tom lane
>
From | Date | Subject | |
---|---|---|---|
Next Message | John R Pierce | 2013-10-24 18:31:38 | Re: PostgreSQL Point In Time Recovery |
Previous Message | Jeff Janes | 2013-10-24 16:47:13 | Re: PostgreSQL Point In Time Recovery |