Wow... thank you very much, Scott.
SELECT pg_stat_reset(); ANALYZE;
And everything working like a charm....
What is the reason for this to happen? Is it common?
-----Original Message-----
From: Scott Marlowe <
scott.marlowe@gmail.com>
To: Iñigo Martinez Lasala <
imartinez@vectorsf.com>
Cc: pgsql-admin <
pgsql-admin@postgresql.org>, Jaume Sabater <
jsabater@gmail.com>, Sergio Chavarria <
sergio.chavarria@gmail.com>
Subject: Re: [ADMIN] Strange behaviour with a query
Date: Fri, 17 Apr 2009 04:30:19 -0600
On Fri, Apr 17, 2009 at 4:29 AM, Scott Marlowe <scott.marlowe@gmail.com> wrote:
> On Fri, Apr 17, 2009 at 4:17 AM, Scott Marlowe <scott.marlowe@gmail.com> wrote:
>> On Fri, Apr 17, 2009 at 3:00 AM, Iñigo Martinez Lasala
>> <imartinez@vectorsf.com> wrote:
>>> Hi everybody again.
>>>
>>> Deleting rel_dis_can_fk index has solved the problem! But.... why??
>>
>> Hard to say without explain analyze output.
>>
>
> I'd hazard a guess that you're getting a lot more rows back from the
> bitmap scan of rel_dis_can_fk than the query planner expects. But
> that's just a guess.
Try craniking up your default stats setting and running analyze again
and seeing if it runs fast even with the index.