X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (av.hub.org [200.46.204.144]) by svr1.postgresql.org (Postfix) with ESMTP id B2F43D7842 for ; Wed, 24 Aug 2005 11:55:16 -0300 (ADT) Received: from svr1.postgresql.org ([200.46.204.71]) by localhost (av.hub.org [200.46.204.144]) (amavisd-new, port 10024) with ESMTP id 91406-09 for ; Wed, 24 Aug 2005 14:55:10 +0000 (GMT) Received: from ns.snowman.net (ns.snowman.net [66.92.160.21]) by svr1.postgresql.org (Postfix) with ESMTP id 97D37D7805 for ; Wed, 24 Aug 2005 11:55:08 -0300 (ADT) Received: by ns.snowman.net (Postfix, from userid 1000) id C2A2817B11; Wed, 24 Aug 2005 10:55:33 -0400 (EDT) Date: Wed, 24 Aug 2005 10:55:33 -0400 From: Stephen Frost To: Donald Courtney Cc: William Yu , pgsql-performance@postgresql.org Subject: Re: Caching by Postgres Message-ID: <20050824145533.GN6026@ns.snowman.net> Mail-Followup-To: Donald Courtney , William Yu , pgsql-performance@postgresql.org References: <20050823171045.12112.qmail@web51309.mail.yahoo.com> <20050823124323.710a9dd3.frank@wiles.org> <430B6DE3.6040109@sun.com> <430C7448.7030103@sun.com> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="BU7+kJFeeDlNltZg" Content-Disposition: inline In-Reply-To: <430C7448.7030103@sun.com> X-Editor: Vim http://www.vim.org/ X-Info: http://www.snowman.net X-Operating-System: Linux/2.4.24ns.3.0 (i686) X-Uptime: 10:48:33 up 74 days, 7:06, 6 users, load average: 0.05, 0.13, 0.09 User-Agent: Mutt/1.5.9i X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, hits=0.004 required=5 tests=[AWL=0.004] X-Spam-Level: X-Archive-Number: 200508/382 X-Sequence-Number: 14129 --BU7+kJFeeDlNltZg Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable * Donald Courtney (Donald.Courtney@sun.com) wrote: > To be clear - > I built postgreSQL 8.1 64K bit on solaris 10 a few months ago > and side by side with the 32 bit postgreSQL build saw no improvement.=20 > In fact the 64 bit result was slightly lower. That makes some sense actually. It really depends on what you're doing alot of the time. On a Sparc system you're not likely to get much of a speed improvment by going to 64bit (unless, maybe, you're doing lots of intensive 64bit math ops). You'll have larger pointers and whatnot though. > I used the *same 64 bit S10 OS* for both versions. I think your > experience makes sense since your change was from 32 to 64 bit Linux. 32bit to 64bit Linux on a Sparc platform really shouldn't affect performance all that much (I'd expect it to be similar to 32bit to 64bit under Solaris actually, at least in terms of the performance difference). 32bit to 64bit Linux on an amd64 platform is another matter entirely though, but not because of the number of bits involved. Under amd64, 32bit is limited to 32bit on i386 which has a limited number of registers and whatnot. Under amd64/64bit you get more registers (and I think some other niceities) which will improve performance. That's not a 32bit vs. 64bit thing, that's i386 vs. native amd64. It's really mainly an oddity of the platform. On a mips system I'd expect the same kind of performance difference between 32bit and 64bit as you'd see on a sparc platform. Enjoy, Stephen --BU7+kJFeeDlNltZg Content-Type: application/pgp-signature; name="signature.asc" Content-Description: Digital signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFDDIplrzgMPqB3kigRAgdZAJ9qbWNbxrPc6bSy7foIPir/sPv0RwCfWhVT 9IKq1E6qwm6VW6vw3arTAhc= =pqT1 -----END PGP SIGNATURE----- --BU7+kJFeeDlNltZg--