Re: [REVIEW] Prototype: In-place upgrade v02

From: Zdenek Kotala <Zdenek(dot)Kotala(at)Sun(dot)COM>
To: abbas(dot)butt(at)enterprisedb(dot)com
Cc: pgsql-hackers(at)postgresql(dot)org
Subject: Re: [REVIEW] Prototype: In-place upgrade v02
Date: 2008-09-30 07:08:14
Message-ID: 48E1D05E.7020807@sun.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Abbas napsal(a):
> On Mon, 2008-09-29 at 14:42 +0200, Zdenek Kotala wrote:
>>
>>> Do I have to perform performance tests too?
>> Yes, please. My colleague tested it and got 5% performance drop, but it was not
>> complete version and I tested full patch on Friday and It was surprise for me
>> ... I got little bit better throughput (about 0,5%) with patch. I'm going to
>> retest it again but it would be good to get result also from others.
>
>
> Are there defined ways of benchmarking the current performance of pg
> database on my system?
> Are there any guide lines on what type of data or table design should be
> used?

I think integer, numeric, text and some composite type should be enough.

> One option of getting the benchmarks could be to use the run time
> statistics logging with and without the patch.

Run time statistic should be same. Patch add extra wrapper around page and
heaptuple processing. By my opinion pg_stat* output shouldn't show extra
difference.

> And I have to test the performance of the 4 basic queries INSERT,
> UPDATE, DELETE and SELECT, right?
>

Yes. It would be nice compare simple SELECT and complex SELECT too.

Thanks Zdenek

Zdenek Kotala Sun Microsystems
Prague, Czech Republic http://sun.com/postgresql

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Heikki Linnakangas 2008-09-30 07:23:22 Re: FSM rewrite: doc changes
Previous Message Robins Tharakan 2008-09-30 06:19:31 Function management in PG