Re: PG 8.2beta reordering working for this case? - Mailing list pgsql-sql

From Tom Lane
Subject Re: PG 8.2beta reordering working for this case?
Date
Msg-id 1600.1160403323@sss.pgh.pa.us
Whole thread Raw
In response to Re: PG 8.2beta reordering working for this case?  (Kyle Bateman <kyle@actarg.com>)
List pgsql-sql
Kyle Bateman <kyle@actarg.com> writes:
> BTW, explain analyze still says 10 seconds of run time (and takes 10 
> seconds to run), but when I remove the explain analyze, the query runs 
> in about a second.  What's that all about?

Instrumentation overhead ... you're probably running this on a PC with a
very slow clock-reading capability.  Each node output row counted by
explain analyze takes two gettimeofday() calls, and apparently it's not
unusual for those to take several microseconds on cheap motherboards,
even when the CPU is nominally very fast.
        regards, tom lane


pgsql-sql by date:

Previous
From: Kyle Bateman
Date:
Subject: Re: PG 8.2beta reordering working for this case?
Next
From: "Aaron Bono"
Date:
Subject: Re: i have table