Date: Fri, 13 Jul 2007 21:32:09 +0200
From: Adriaan van Os <postgres@microbizz.nl>
Subject: Re: TRUNCATE TABLE
In-reply-to: <23092.1184344246@sss.pgh.pa.us>
To: pgsql-performance@postgresql.org
Message-id: <4697D339.2090204@microbizz.nl>
MIME-version: 1.0
Content-type: text/plain; format=flowed; charset=ISO-8859-1
Content-transfer-encoding: 7BIT
References: <46953949.3030201@microbizz.nl>
 <87ps2ypox3.fsf@oxford.xeocode.com>
	<46972DFA.9000900@microbizz.nl> <20070713124435.GL31358@mathom.us>
	<21495.1184337630@sss.pgh.pa.us> <4697A58E.8060904@microbizz.nl>
	<23092.1184344246@sss.pgh.pa.us>
User-Agent: Thunderbird 1.5.0.4 (Macintosh/20060516)

Tom Lane wrote:
> Adriaan van Os <postgres@microbizz.nl> writes:
>> I started another test. I copied an existing database (not very large,
>> 35 tables, typically a few hundred up to a few thousand records) with
>> CREATE DATABASE testdb TEMPLATE mydb and started to remove random
>> tables from testdb with DROP TABLE and TRUNCATE TABLE. I did this with
>> the query tool of pgAdmin III, to exclude any doubts about my own
>> software (that uses pqlib).
> 
> Can you try it with plain psql?  pgAdmin is a variable that wasn't
> accounted for in my tests.

Will do that and report the results.

>> The hardware is an Intel dual-core 17-inch
>> MacBook Pro running Mac OS X 10.4.
> 
> Hmm.  I thought you said Fedora before.

Yes, the test that I mentioned yesterday was on Fedora, but as you were "of the opinion that 
there's something broken about
Adriaan's infrastructure" I tried the new test on a completely different system today.

   However, I'd done a few tests
> yesterday on my own Mac laptop (Al G4) and not gotten results that were
> out of line with HPUX or Fedora.
> 
> Does anyone else want to try replicating these tests?

Thanks,

Adriaan van Os