From: | Robert Haas <robertmhaas(at)gmail(dot)com> |
---|---|
To: | Reydan Cankur <reydan(dot)cankur(at)gmail(dot)com> |
Cc: | pgsql-performance(at)postgresql(dot)org |
Subject: | Re: Best Profiler for PostgreSQL |
Date: | 2009-09-10 15:30:19 |
Message-ID: | 603c8f070909100830u3aa6a9ebw974468dac3a2dcd8@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-performance |
On Wed, Sep 9, 2009 at 6:15 AM, Reydan Cankur <reydan(dot)cankur(at)gmail(dot)com> wrote:
> Hi,
>
> I am running PostgreSQL-8.4.0 on a SMP Server which has 32 processors
> (32X2=64 cores). I am working on database parallelism and I need to do
> profiling in order to find the relevant parts to parallelize. I wrote 15
> queries which are performing select, sort, join, and aggregate functions and
> I want to profile these queries and try to understand where postgresql
> spends more time and then I will try to parallelize source code(time
> spending parts) by using openMP. But I need guidance about profiling tool,
> which is the best tool to profile queries. I used gprof but I want to
> profile with a more advanced tool. Options are oprofile and valgrind.
>
> 1) I can not decide which best suits for query profiling. oprofile or
> valgrind?
>
> 2) Also for both I can not find documentation about profiling steps.
You might want to start with EXPLAIN ANALYZE. That will tell you
where the time for each query is being spent.
...Robert
From | Date | Subject | |
---|---|---|---|
Next Message | tv | 2009-09-10 15:40:24 | Re: Sub-optimal plan chosen |
Previous Message | bricklen | 2009-09-10 15:25:34 | Re: Sub-optimal plan chosen |