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 (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mBLL34B8017629 for ; Sun, 21 Dec 2008 15:03:04 -0600 Received: from sandeen.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 6938D176BF7F for ; Sun, 21 Dec 2008 13:03:02 -0800 (PST) Received: from sandeen.net (sandeen.net [209.173.210.139]) by cuda.sgi.com with ESMTP id 1EiRZU03zAKck6pR for ; Sun, 21 Dec 2008 13:03:02 -0800 (PST) Message-ID: <494EAF05.5020901@sandeen.net> Date: Sun, 21 Dec 2008 15:03:01 -0600 From: Eric Sandeen MIME-Version: 1.0 Subject: Re: Big filesystem > (4TiB) recommendations (mkfs parameter tuning?) References: In-Reply-To: List-Id: XFS Filesystem from SGI List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: xfs-bounces@oss.sgi.com Errors-To: xfs-bounces@oss.sgi.com To: Justin Piszcz Cc: xfs@oss.sgi.com Justin Piszcz wrote: > Often, on this list, there are issues where there is not enough memory, > etc, to repair a large filesystem. > > If one is going to create a 10-16TiB filesystem (for example), should one > use the mkfs.xfs defaults, or should there be special tuning for > filesystems of this size, so they will be less susceptible to failure for > one reason or another? > > Justin. Memory to repair will depend more on total inodes on the system etc, not so much the absolute size of the filesystem. The defaults are defaults for a reason; they are updated when warranted, and unless you have a specific issue with them, it's best to just stick with that. IOW, if best practices warranted different tuning for a generic 10-16TiB filesystem, then the defaults would need to be updated, rather than offering hints 'n' tips on mailing lists. :) -Eric _______________________________________________ xfs mailing list xfs@oss.sgi.com http://oss.sgi.com/mailman/listinfo/xfs