Re: index speed and failed expectations? - Mailing list pgsql-general

From Adam Rich
Subject Re: index speed and failed expectations?
Date
Msg-id 005201c8f636$f5e5e3d0$e1b1ab70$@r@sbcglobal.net
Whole thread Raw
In response to index speed and failed expectations?  (rihad <rihad@mail.ru>)
Responses Re: index speed and failed expectations?
Re: index speed and failed expectations?
Re: index speed and failed expectations?
List pgsql-general
> This query from the console:
>
> select * from stats order by start_time;
>
> takes 8 seconds before starting its output. Am I wrong in assuming that
> the index on start_time should make ORDER BY orders of magnitude
> faster?
> Or is this already fast enough? Or should I max up some memory (buffer)
> setting to achieve greater speeds? Not that the speed is crucial, just
> curious.
>

Postgresql won't use the index for queries like this.  Due to the
MVCC implementation, the index does not contain all necessary information
and would therefore be slower than using the table data alone.

(What postgresql lacks is a first_row/all_rows hint like oracle)

However, if you limit the number of rows enough, you might force it
to use an index:

select * from stats order by start_time limit 1000;




pgsql-general by date:

Previous
From: rihad
Date:
Subject: index speed and failed expectations?
Next
From: rihad
Date:
Subject: Re: index speed and failed expectations?