Re: Problem analyzing explain analyze output - Mailing list pgsql-performance

From Guillaume Smet
Subject Re: Problem analyzing explain analyze output
Date
Msg-id 435C801F.1030709@openwide.fr
Whole thread Raw
In response to Re: Problem analyzing explain analyze output  ("Steinar H. Gunderson" <sgunderson@bigfoot.com>)
List pgsql-performance
Steinar,

> which seems to make sense; you have one run of about 257ms, plus 514 runs
> taking about 0.035ms each (ie. about 18ms), which should add up to become
> about 275ms (which is close enough to the reality of 281ms).

Yep. The line that disturbed me was the bitmap index scan with a cost of
"actual time=254.143..254.143". I was more looking for something like
"actual time=0..254.143" which is what I usually have for an index scan.
So I suppose that the bitmap index scan returns rows only when it's
totally computed.

Thanks for your help.

Regards.

--
Guillaume

pgsql-performance by date:

Previous
From: Christopher Kings-Lynne
Date:
Subject: Re: prepared transactions that persist across sessions?
Next
From: Alex Turner
Date:
Subject: Re: What gets cached?