Message-ID: <4BACA611.8050607@archonet.com>
Date: Fri, 26 Mar 2010 12:18:25 +0000
From: Richard Huxton <dev@archonet.com>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.9.1.9pre) Gecko/20100217 Lightning/1.0b1 Shredder/3.0.3pre
MIME-Version: 1.0
To: Eliot Gable <egable+pgsql-performance@gmail.com>
CC: Merlin Moncure <mmoncure@gmail.com>,
 pgsql-performance@postgresql.org
Subject: Re: Performance Tuning Large PL/PGSQL Stored Procedure
References: <bf6923ed1003231400x6c9a6166me835385568851ea@mail.gmail.com>	
	<b42b73151003251900k6cfe4bc8j7a6b796039482a22@mail.gmail.com>
	<bf6923ed1003252056m4c40660em6c7d92fc3a917e0c@mail.gmail.com>
In-Reply-To: <bf6923ed1003252056m4c40660em6c7d92fc3a917e0c@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

On 26/03/10 03:56, Eliot Gable wrote:
>
> I really am chasing milliseconds here, and I appreciate all your feedback.
> You've given me a relatively large number of possible optimizations I can
> try out. I will definitely try out the libpqtypes. That sounds like a
> promising way to further cut down on execution time. I think most of my
> performance penalty is in transfering the results back to the C++
> application.

In addition to all of Merlin's good advice, if the client is on a 
different machine to the server then try sticking wireshark or similar 
onto the connection. That should make it pretty clear where the main 
costs are in getting your data back.

-- 
   Richard Huxton
   Archonet Ltd