Re: Why is explain horribly optimistic for sorts? - Mailing list pgsql-general

From Ben
Subject Re: Why is explain horribly optimistic for sorts?
Date
Msg-id Pine.LNX.4.10.10103031007180.19743-100000@gilgamesh.eos.SilentMedia.com
Whole thread Raw
In response to Re: Why is explain horribly optimistic for sorts?  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: Why is explain horribly optimistic for sorts?  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-general
Yes, it is horribly wrong.

select count(*) FROM jennyann where target like '/music/%'

gives me 93686 rows.

On Sat, 3 Mar 2001, Tom Lane wrote:

> Ben <bench@silentmedia.com> writes:
> > Here's what explain tells me:
>
> > explain SELECT * FROM jennyann where target like '/music/%' order by "LogTime" limit 1000;
> > NOTICE:  QUERY PLAN:
>
> > Sort  (cost=119.88..119.88 rows=2085 width=136)
> >   ->  Index Scan using jennyann_target_key on jennyann  (cost=0.00..4.94 rows=2085 width=136)
>
> How many records are actually getting sorted here --- ie, how many would
> you get back if you didn't have the LIMIT?  It seems to be estimating
> 2085, but I suspect that must be wildly wrong ...
>
>             regards, tom lane
>


pgsql-general by date:

Previous
From: Ben
Date:
Subject: Re: Why is explain horribly optimistic for sorts?
Next
From: Ben
Date:
Subject: Re: Why is explain horribly optimistic for sorts?