From: | Filip Rembiałkowski <plk(dot)zuber(at)gmail(dot)com> |
---|---|
To: | Tyler Hains <thains(at)profitpointinc(dot)com> |
Cc: | pgsql-general <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: Query Optimizer makes a poor choice |
Date: | 2011-11-29 22:06:34 |
Message-ID: | CAP_rwwnET3MKOpELPaTHq71GDwwwsNnQBNbbmPG=SSgRijtKJg@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
2011/11/29 Tyler Hains <thains(at)profitpointinc(dot)com>:
> I haven't had a chance to experiment with the SET STATISTICS, but that
> got me going on something interesting...
>
> Do these statistics look right?
>
> # SELECT attname, n_distinct, most_common_vals, histogram_bounds FROM
> pg_stats WHERE tablename = 'cards';
>
...
> "card_set_id" 905
> "{5201,3203,3169,5679,5143,5204,5655,4322,5236,4513}"
> "{4,3080,3896,4349,4701,5179,5445,5706,6003,6361,6784}"
This looks promising, because n_distinct is low enough that you can
cover almost all values with statistics.
raise the statistics and ANALYZE. should help.
(NOTE NOTE NOTE: assuming that the distribution is even)
...
but one thing we see for sure is that you have not tuned your
PostgreSQL instance :-)
I would recommend pgtune, -> pgfoundry.org/projects/pgtune/
it covers most important stuff, *including* default_statistics_target.
Filip
From | Date | Subject | |
---|---|---|---|
Next Message | Heiko Wundram | 2011-11-29 22:18:43 | Re: Limiting number of connections to PostgreSQL per IP (not per DB/user)? |
Previous Message | Tyler Hains | 2011-11-29 21:43:52 | Re: Query Optimizer makes a poor choice |