pg.ddx.io  pgsql-performance@postgresql.org mailing list archive  
help / color / mirror / Atom feed
From: mark <markkicks@gmail.com>
To: Justin <justin@emproshunts.com>
Cc: pgsql-general@postgresql.org
Cc: pgsql-performance@postgresql.org
Subject: Re: select query takes 13 seconds to run with index
Date: Mon, 26 May 2008 16:57:06 -0700
Message-ID: <82fa9e310805261657u4a84b5fftd3abb50314f18815@mail.gmail.com> (raw)
In-Reply-To: <483B4C74.7000703@emproshunts.com>
References: <82fa9e310805261549m2362fbe8sbf18789938fee833@mail.gmail.com>
	<483B4729.7050402@emproshunts.com>
	<82fa9e310805261632y46f306fdh84e2da62ffbd7873@mail.gmail.com>
	<483B4C74.7000703@emproshunts.com>

On Mon, May 26, 2008 at 4:49 PM, Justin <justin@emproshunts.com> wrote:
> mark wrote:
>
> On Mon, May 26, 2008 at 4:26 PM, Justin <justin@emproshunts.com> wrote:
> mark wrote:
> Hi, is there anyway this can be made faster?  id is the primary key,
> and there is an index on uid..
> thanks
> EXPLAIN ANALYZE select * from pokes where uid = 578439028 order by id
> DESC limit 6;
>                                                                     QUERY
> PLAN
>
> ----------------------------------------------------------------------------------------------------------------------------------------------------
>  Limit  (cost=0.00..9329.02 rows=6 width=135) (actual
> time=13612.247..13612.247 rows=0 loops=1)
>   ->  Index Scan Backward using pokes_pkey on pokes
> (cost=0.00..5182270.69 rows=3333 width=135) (actual
> time=13612.245..13612.245 rows=0 loops=1)
>         Filter: (uid = 578439028)
>  Total runtime: 13612.369 ms
> (4 rows)
>
>
> First this should be posted on performance list.
>
>
> sorry about this.
>
>
>
> how many records are in this table?
>
>
> 22334262, 22 million records.
>
>
>
> The estimate is way off, when was the last time Vaccum was on the table?
>
>
> about a week ago i ran this VACUUM VERBOSE ANALYZE;
> this table is never updated or deleted, rows are just inserted...
>
>
>
>
> What verison of Postgresql are you running
>
>
> 8.3.1
>
>
>
> Size of the Table
>
>
> 22 million rows approximately
>
>
> I have no experience  on large datasets so people with more experience in
> this area are going to have to chime in.
> My gut feel is 13 seconds for Postgresql to sort through an index of that
> size and table is not bad.
>
> you may need to take a look at hardware and postgresql.config settings to
> improve the performance for this query
>
> This query is very simple where changing it around or adding index results
> massive improvements is not going to help in this case.
the hardware is e5405 dual quad core on a 16GB RAM machine, with 8.3.1
default settings except maximum connections increased...



view thread (16+ messages)  latest in thread

Message-ID: <82fa9e310805261657u4a84b5fftd3abb50314f18815@mail.gmail.com>
Permalink:  ../82fa9e310805261657u4a84b5fftd3abb50314f18815@mail.gmail.com/
Also on:    postgresql.org/message-id/82fa9e310805261657u4a84b5fftd3abb50314f18815@mail.gmail.com

 · 

reply

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Reply to all the recipients using the --to and --cc options:
  reply via email

  To: pgsql-performance@postgresql.org
  Cc: markkicks@gmail.com, justin@emproshunts.com, pgsql-general@postgresql.org
  Subject: Re: select query takes 13 seconds to run with index
  In-Reply-To: <82fa9e310805261657u4a84b5fftd3abb50314f18815@mail.gmail.com>

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox