Received: from localhost (maia-4.hub.org [200.46.204.183]) by postgresql.org (Postfix) with ESMTP id B96C39FA2FD for ; Sat, 16 Dec 2006 04:27:09 -0400 (AST) Received: from postgresql.org ([200.46.204.71]) by localhost (mx1.hub.org [200.46.204.183]) (amavisd-new, port 10024) with ESMTP id 78308-04 for ; Sat, 16 Dec 2006 04:27:05 -0400 (AST) X-Greylist: from auto-whitelisted by SQLgrey-1.7.4 Received: from smtp4.libero.it (smtp4.libero.it [193.70.192.54]) by postgresql.org (Postfix) with ESMTP id C59A29F9F47 for ; Sat, 16 Dec 2006 04:27:05 -0400 (AST) Received: from localhost (172.16.1.205) by smtp4.libero.it (7.0.027-DD01) id 453F690A03E9BCFE; Sat, 16 Dec 2006 09:27:03 +0100 Received: from smtp3.libero.it ([172.16.1.53]) by localhost (asav19.libero.it [193.70.193.6]) (amavisd-new, port 10024) with ESMTP id 31993-22; Sat, 16 Dec 2006 09:27:03 +0100 (CET) Received: from [151.25.103.14] (ppp-14-103.25-151.libero.it [151.25.103.14]) by smtp3.libero.it (Postfix) with ESMTP id 21638ABE48; Sat, 16 Dec 2006 09:27:01 +0100 (MET) Message-ID: <4583ADD5.4070703@streppone.it> Date: Sat, 16 Dec 2006 09:27:01 +0100 From: Cosimo Streppone User-Agent: Mozilla Thunderbird 1.0.2 (X11/20050317) X-Accept-Language: it, it-it, en-us, en MIME-Version: 1.0 To: Alexander Staubo CC: Postgresql Performance list Subject: Re: New to PostgreSQL, performance considerations References: <4CD99FDA-14B5-4EA9-9440-55E2548C34C1@purefiction.net> <09D6591B-4AA9-4994-BED6-0C7C583BF933@purefiction.net> <44088FE9-4AA9-4FD6-AE8F-34560A6401BA@purefiction.net> In-Reply-To: <44088FE9-4AA9-4FD6-AE8F-34560A6401BA@purefiction.net> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Scanned: with antispam and antivirus automated system at libero.it X-Virus-Scanned: Maia Mailguard 1.0.1 X-Archive-Number: 200612/434 X-Sequence-Number: 22490 Alexander Staubo wrote: > On Dec 15, 2006, at 17:53 , Ron wrote: > >> At 09:50 AM 12/15/2006, Greg Smith wrote: >> >>> On Fri, 15 Dec 2006, Merlin Moncure wrote: >>> >>>> The slower is probably due to the unroll loops switch which can >>>> actually hurt code due to the larger footprint (less cache coherency). >>> >>> The cache issues are so important with current processors that I'd >>> suggest throwing -Os (optimize for size) into the mix people test. > > So far I have been compiling PostgreSQL and running my pgbench script > manually, but this makes me want to modify my script to run pgbench > automatically using all possible permutations of a set of compiler flags. I don't know if it's practical, but this link comes to mind: http://clusty.com/search?query=acovea -- Cosimo