From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: with ECARTIS (v1.0.0; list xfs); Thu, 03 Jul 2008 20:50:25 -0700 (PDT) Received: from cuda.sgi.com ([192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id m643oKkV012452 for ; Thu, 3 Jul 2008 20:50:21 -0700 Received: from web34507.mail.mud.yahoo.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with SMTP id 226321877C48 for ; Thu, 3 Jul 2008 20:51:23 -0700 (PDT) Received: from web34507.mail.mud.yahoo.com (web34507.mail.mud.yahoo.com [66.163.178.173]) by cuda.sgi.com with SMTP id wzh0PhFdVmhGNAHW for ; Thu, 03 Jul 2008 20:51:23 -0700 (PDT) Date: Thu, 3 Jul 2008 20:51:23 -0700 (PDT) From: Mark Reply-To: MusicMan529@yahoo.com Subject: Re: user defined cache In-Reply-To: <1215139205.7187.33.camel@quadski> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Message-ID: <328802.35603.qm@web34507.mail.mud.yahoo.com> Sender: xfs-bounce@oss.sgi.com Errors-to: xfs-bounce@oss.sgi.com List-Id: xfs To: xfs@oss.sgi.com, Warwick Boote --- On Thu, 7/3/08, Warwick Boote wrote: > Hi all, > > Do you know of setting or layered file system that enables > me to define > a large buffer in main memory and cache writes to this > buffer? > > Obviously i'd need the file system to also read from > this buffer if the > data hasn't been flushed to the disk yet. XFS supports the following mount options, and they do increase throughput somewhat on my system: logbufs=N (helps concurrency) logbsize=N (helps log buffering) If you are using an internal XFS log, you may also try "nobarrier" to reduce force-to-disk write frequency. NB: I have seen this option discussed on this mailing list, but I do not see it in the mount(8) man page. Bear in mind that the XFS log is for metadata, rather than actual file data. This may be what you're looking for, or at least sufficient; if not, my apologies. -- Mark "What better place to find oneself than on the streets of one's home village?" --Capt. Jean-Luc Picard, "Family"