From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from cuda.sgi.com (cuda1.sgi.com [192.48.157.11]) by oss.sgi.com (8.14.3/8.14.3/SuSE Linux 0.8) with ESMTP id o8FEkRSF193399 for ; Wed, 15 Sep 2010 09:46:27 -0500 Received: from pavilion.ashurst.eu.org (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 2300014F2E83 for ; Wed, 15 Sep 2010 07:59:01 -0700 (PDT) Received: from pavilion.ashurst.eu.org (pavilion.ashurst.eu.org [212.13.194.45]) by cuda.sgi.com with ESMTP id aWq3uGZrDvzwHSRs for ; Wed, 15 Sep 2010 07:59:01 -0700 (PDT) Message-ID: <4C90DC1E.6090002@ashurst.eu.org> Date: Wed, 15 Sep 2010 15:45:50 +0100 From: Andy Bennett MIME-Version: 1.0 References: <4C90D080.5090004@mrc-lmb.cam.ac.uk> In-Reply-To: <4C90D080.5090004@mrc-lmb.cam.ac.uk> Subject: Re: no space left on device with 662G free List-Id: XFS Filesystem from SGI List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Sender: xfs-bounces@oss.sgi.com Errors-To: xfs-bounces@oss.sgi.com To: Dave Howorth Cc: xfs@oss.sgi.com Hi, > I recently started using XFS when I set up a new server and I'm having a > problem when copying data on to it from an old server. cp reports errors > like: > > cp: cannot create symbolic link > `/nfs/scop4/data/trembl/xml/releases/2006-01-10/Q6RXU6.xml': No space > left on device > > and later > > cp: cannot create directory > `/nfs/scop4/data/trembl/xml/releases/2006-01-24': No space left on device > > The filesystem is in a 2 TB LVM on an md RAID and according to df > there's 662 GB space free: > > /dev/mapper/vg--storage-lv--data > 2.0T 1.4T 662G 68% /nfs/scop4/data > > There are a quite a few directories in the filesystem. Some of the > directories contain many millions of files and some directories consist > entirely of symlinks, if any of that's relevant. > > The filesystem is on a newish machine running openSUSE 11.2 (Linux scop4 > 2.6.31.12-0.2-desktop #1 SMP PREEMPT 2010-03-16 21:25:39 +0100 x86_64 > x86_64 x86_64 GNU/Linux) and the data is being mirrored from an older > machine that uses a reiser filesystem and which stores the data without > problem. > > Some folks on the suse mailing list said it was probably an inode > problem and suggested I run df -i: > > # df -i /nfs/scop4/data > Filesystem Inodes IUsed IFree IUse% Mounted on > /dev/mapper/vg--storage-lv--data > 429496704 -18446744073280007488 > 18446744073709504192 - /nfs/scop4/data > > I don't understand that output, especially the negative number! I then > read about the inode64 mount option in the XFS FAQ and I believe I've > now enabled that and remounted the filesystem. mtab shows > > /dev/mapper/vg--storage-lv--data /nfs/scop4/data xfs rw,noatime,inode64 0 0 > > But I'm still seeing the same errors as before. Is there something else > I need to do to enable inode64, or am I looking in the wrong direction? Seems like you have a cross between http://xfs.org/index.php/XFS_FAQ#Q:_What_is_the_inode64_mount_option_for.3F and http://xfs.org/index.php/XFS_FAQ#Q:_Why_do_I_receive_No_space_left_on_device_after_xfs_growfs.3F Perhaps not using inode64 from the beginning looks similar to a xfs_growfs? Regards, @ndy -- andyjpb@ashurst.eu.org http://www.ashurst.eu.org/ 0x7EBA75FF _______________________________________________ xfs mailing list xfs@oss.sgi.com http://oss.sgi.com/mailman/listinfo/xfs