From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: with ECARTIS (v1.0.0; list xfs); Mon, 28 Apr 2008 17:43:15 -0700 (PDT) Received: from cuda.sgi.com (cuda1.sgi.com [192.48.168.28]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id m3T0h534027383 for ; Mon, 28 Apr 2008 17:43:07 -0700 Received: from sandeen.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 0314E10D0B7D for ; Mon, 28 Apr 2008 17:43:48 -0700 (PDT) Received: from sandeen.net (sandeen.net [209.173.210.139]) by cuda.sgi.com with ESMTP id lkbseXJseshRulof for ; Mon, 28 Apr 2008 17:43:48 -0700 (PDT) Message-ID: <48166F42.50104@sandeen.net> Date: Mon, 28 Apr 2008 19:43:46 -0500 From: Eric Sandeen MIME-Version: 1.0 Subject: Re: Problems with xfs_grow on large LVM + XFS filesystem 20TB size check 2 failed References: <6A32BC807C106440B7E23208F280DDAF01D21F36FD@bcmail1.VIDMARK.LOCAL> <481650F5.40205@sandeen.net> <6A32BC807C106440B7E23208F280DDAF01D21F3718@bcmail1.VIDMARK.LOCAL> <481656F6.5030300@sandeen.net> <48166E18.10008@sgi.com> In-Reply-To: <48166E18.10008@sgi.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: xfs-bounce@oss.sgi.com Errors-to: xfs-bounce@oss.sgi.com List-Id: xfs To: markgw@sgi.com Cc: Lance Reed , "xfs@oss.sgi.com" Mark Goodwin wrote: > > Eric Sandeen wrote: >> Lance Reed wrote: >>> Thanks for the quick response. >>> >>> Actually, I was able to run an xfs_repair and all was well.(took 45 minutes...) >>> >>> But I would love to be able to expand the XFS file system out to the max. >>> >>> I guess I could expand it in < 2TB increments then maybe? >>> >>> Thanks for the update and I will look around. >>> >>> I did find this which I think is related. >>> >>> http://oss.sgi.com/archives/xfs/2008-01/msg00097.html >> Yep, I think expanding it in, say, 1T increments should be fine, and it >> should all (I think...) end up the same as if you'd done it all at once >> (modulo the bug, of course...) > > I think we verified a while back that growing in 2T increments is an > effective workaround - the bug is due to a signed 32 bit overflow. > The fix has been available for some time now, but apparently hasn't > made it's way into Centos yet. Honestly, I think nothing makes its way back to Centos... I really don't have the time to maintain it. If anyone on the list uses Centos + xfs and wants to backport patches that appear to be bugfixes, I'm sure it'd be welcomed. I'd be happy to facilitate w/ review or whatnot, but probably won't have time to actually take on this task myself (despite being the instigator of the centos module originally, which was, in retrospect, perhaps a tad irresponsible...) :) -Eric