Received: from localhost (postgresql.org [64.49.215.8]) by postgresql.org (Postfix) with ESMTP id C6192476048 for ; Thu, 2 Jan 2003 15:57:32 -0500 (EST) Received: from local.iboats.com (local.iboats.com [209.63.105.131]) by postgresql.org (Postfix) with SMTP id 52746476419 for ; Thu, 2 Jan 2003 15:57:08 -0500 (EST) Received: (qmail 17032 invoked by uid 82); 2 Jan 2003 13:56:43 -0700 Received: from nw@codon.com by local.iboats.com by uid 81 with qmail-scanner-1.14 ( Clear:. Processed in 0.027759 secs); 02 Jan 2003 20:56:43 -0000 Received: from unknown (HELO WEASEL) (209.63.105.136) by 0 with SMTP; 2 Jan 2003 13:56:42 -0700 Message-ID: <006401c2b2a1$d322fec0$88693fd1@WEASEL> From: "Steve Wolfe" To: References: <008701c2b286$4c9c2ae0$88693fd1@WEASEL> <3255.1041537454@sss.pgh.pa.us> Subject: Re: Question on hardware & server capacity Date: Thu, 2 Jan 2003 13:59:19 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1106 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Virus-Scanned: by AMaViS new-20020517 X-Archive-Number: 200301/9 X-Sequence-Number: 656 > So an average backend only processes ~ 5 queries before exiting? > > 7.3 has reduced some of those startup costs a little, so if you're still > on 7.2 then an update might help. But I'd strongly recommend getting > connection re-use in place before you go off and buy hardware. I've been fooling around with some connection pooling, and it hasn't make the sort of difference we're looking for. Going from 3 queries per back-end to 100 queries per backend made only about a 20% difference. While that's nothing to scoff at, we're looking for at least a 100% improvement. Either way, the connection pooling WILL be put in place, but I'm certainly not counting on it preventing the need for a hardware upgrade. steve