Re: Can the V7.3 EXPLAIN ANALYZE be trusted? - Mailing list pgsql-performance

From Leeuw van der, Tim
Subject Re: Can the V7.3 EXPLAIN ANALYZE be trusted?
Date
Msg-id BF88DF69D9E2884B9BE5160DB2B97A85010D7046@nlshl-exch1.eu.uis.unisys.com
Whole thread Raw
In response to Can the V7.3 EXPLAIN ANALYZE be trusted?  (Steven Rosenstein <srosenst@us.ibm.com>)
Responses Re: Can the V7.3 EXPLAIN ANALYZE be trusted?  (Greg Stark <gsstark@mit.edu>)
Re: Can the V7.3 EXPLAIN ANALYZE be trusted?  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-performance

-----Original Message-----
From: pgsql-performance-owner@postgresql.org [mailto:pgsql-performance-owner@postgresql.org]On Behalf Of Steven
Rosenstein
Sent: Sunday, February 06, 2005 8:51 PM
To: pgsql-performance@postgresql.org
Subject: [PERFORM] Can the V7.3 EXPLAIN ANALYZE be trusted?






While working on a previous question I posed to this group, I ran a number
of EXPLAIN ANALYZE's to provide as examples.  After sending up my last
email, I ran the same query *without* EXPLAIN ANALYZE.  The runtimes were
vastly different.  In the following example, I ran two identical queries
one right after the other.  The runtimes for both was very close (44.77
sec).  I then immediately ran the exact same query, but without EXPLAIN
ANALYZE.  The same number of rows was returned, but the runtime was only
8.7 sec.  I don't think EXPLAIN ANALYZE puts that much overhead on a query.
Does anyone have any idea what is going on here?

--- Steve


Caching by the OS?

(Did you try to *first* run the query w/o EXPLAIN ANALYZE, and then with? What's the timing if you do that?)

--Tim

pgsql-performance by date:

Previous
From: Steven Rosenstein
Date:
Subject: Re: Are JOINs allowed with DELETE FROM
Next
From: Greg Stark
Date:
Subject: Re: Can the V7.3 EXPLAIN ANALYZE be trusted?