Re: Query Plan choice with timestamps - Mailing list pgsql-performance

From Richard Huxton
Subject Re: Query Plan choice with timestamps
Date
Msg-id 489AEC42.5080007@archonet.com
Whole thread Raw
In response to Re: Query Plan choice with timestamps  (Giorgio Valoti <giorgio_v@mac.com>)
Responses Re: Query Plan choice with timestamps  (Giorgio Valoti <giorgio_v@mac.com>)
List pgsql-performance
Giorgio Valoti wrote:
>
> On 07/ago/08, at 10:35, Richard Huxton wrote:
>
>> Giorgio Valoti wrote:
>>> Hi, I have a timestamptz field that I want to use with a query, but I
>>> don’t need the full timestamp resolution, so I’ve created a
>>> day_trunc(timestamptz) immutable function which I’ll use with the
>>> query and with a new index:
>>> logs=> create index test_idx on blackbox (day_trunc(ts));
>>> However, the query plan doesn’t use the index:
>>
>> Does it use it ever? e.g. with
>>  SELECT * FROM blackbox WHERE day_trunk(ts) = '...'
>
> It’s used:
[snip]

OK - so the index is working.

If you disable seq-scans before running the query, does it use it then?

SET enable_seqscan = off;

 > logs=>explain select count(*) from blackbox group by day_trunc(ts) order
 > by day_trunc(ts);
 >                                         QUERY PLAN
 >
------------------------------------------------------------------------------------------

 >
 >  GroupAggregate  (cost=98431.58..119773.92 rows=74226 width=8)

In particular:
1. Is the estimated cost more or less than 119773.92?
2. How does that match the actual time taken?

--
   Richard Huxton
   Archonet Ltd

pgsql-performance by date:

Previous
From: Giorgio Valoti
Date:
Subject: Re: Query Plan choice with timestamps
Next
From: dforums
Date:
Subject: Re: Plz Heeeelp! performance settings