From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.14.3/8.14.3/SuSE Linux 0.8) with ESMTP id o9F7DnDE110918 for ; Fri, 15 Oct 2010 02:13:49 -0500 Received: from mailsrv14.zmi.at (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 4EDAE191A10F for ; Fri, 15 Oct 2010 00:14:57 -0700 (PDT) Received: from mailsrv14.zmi.at (mailsrv1.zmi.at [212.69.164.54]) by cuda.sgi.com with ESMTP id cP1iwtJu9wNR7Nqk for ; Fri, 15 Oct 2010 00:14:57 -0700 (PDT) From: Michael Monnerie Subject: Re: [RFC, PATCH 0/2] xfs: dynamic speculative preallocation for delalloc Date: Fri, 15 Oct 2010 09:14:54 +0200 References: In-Reply-To: MIME-Version: 1.0 Message-Id: <201010150914.55604@zmi.at> List-Id: XFS Filesystem from SGI List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============5406079036293102373==" Sender: xfs-bounces@oss.sgi.com Errors-To: xfs-bounces@oss.sgi.com To: xfs@oss.sgi.com Cc: Ivan.Novick@emc.com, aelder@sgi.com --===============5406079036293102373== Content-Type: multipart/signed; boundary="nextPart1510069.6G2cAd0Cij"; protocol="application/pgp-signature"; micalg=pgp-sha1 Content-Transfer-Encoding: 7bit --nextPart1510069.6G2cAd0Cij Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable On Donnerstag, 14. Oktober 2010 Ivan.Novick@emc.com wrote: > I have seen ~50% performance improvement in read rate when changing > from small extents to large extents with XFS. Essentially going from > not using allocsize to setting 1gb allocsize. Also GB/s class > filesystem. In case of a database, I'd say 1GB allocsize seems fine. But when you're=20 at /var/log/, it could be a penalty, as when you have 50 open log files=20 which get permanently appended, you have 50GB preallocated which almost=20 surely won't be needed before the log is rotated anyway. The question is: what is a good value for preallocation? I'd guess for a database 1GB seems reasonable, for /var/log one full=20 stripe (nr-data-disks * stripe size), for webserver/ftp data too, as=20 well as a fileserver (storing office documents) and mailserver (storing=20 each mail as a separate file or into one flat file). Or, when you have an app that writes files of specific size, use the=20 "normal maximum" file size. We have a special server storing files of=20 normally 50M-1G size, there I set allocsize=3D1024k, so that files will be= =20 in one junk on disk. When sometimes a file is bigger, then it will be=20 fragmented but at reasonable sizes so it won't matter too much. Is that reasonable? Are there guides to help with that? I'd like to write a FAQ entry if=20 there is (can ever be?) a consensus of what would be a good value. =2D-=20 mit freundlichen Gr=FCssen, Michael Monnerie, Ing. BSc it-management Internet Services http://proteger.at [gesprochen: Prot-e-schee] Tel: 0660 / 415 65 31 ****** Radiointerview zum Thema Spam ****** http://www.it-podcast.at/archiv.html#podcast-100716 // Wir haben im Moment zwei H=E4user zu verkaufen: // http://zmi.at/langegg/ // http://zmi.at/haus2009/ --nextPart1510069.6G2cAd0Cij Content-Type: application/pgp-signature; name=signature.asc Content-Description: This is a digitally signed message part. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.15 (GNU/Linux) iEYEABECAAYFAky3/28ACgkQzhSR9xwSCbRbGACfVGwo7MAUQ0BRY4K9hIkAyPsX NVEAoMck/JZhCWVKa4KQZmy2LAoxOXv3 =MVwQ -----END PGP SIGNATURE----- --nextPart1510069.6G2cAd0Cij-- --===============5406079036293102373== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ xfs mailing list xfs@oss.sgi.com http://oss.sgi.com/mailman/listinfo/xfs --===============5406079036293102373==--