On Tue, Apr 18, 2017 at 07:32:04PM +0200, Stefan Priebe - Profihost AG wrote: > Hi, > > the FS is mounted and in production. I didn't know how to dump a running FS. Since there's no lvm here, unmounting, metadumping, and mounting is probably the only reasonable option. --D > > Greets, > Stefan > > Am 18.04.2017 um 18:59 schrieb Darrick J. Wong: > > On Tue, Apr 18, 2017 at 09:01:17AM +0200, Stefan Priebe - Profihost AG wrote: > >> System crashed again but output was the same. > >> > >> 2017-04-18 07:30:18 XFS (sda3): xfs_iunlink_remove: xfs_imap > >> returned error -22. > >> 2017-04-18 07:30:18 XFS (sda3): xfs_inactive_ifree: xfs_ifree > >> returned error -22 > >> 2017-04-18 07:30:18 XFS (sda3): xfs_do_force_shutdown(0x1) called > >> from line 1872 of file fs/xfs/xfs_inode.c. Return address = > >> 0xffffffff9b33cf93 > >> 2017-04-18 07:30:18 XFS (sda3): I/O Error Detected. Shutting down > >> filesystem > >> 2017-04-18 07:30:18 XFS (sda3): Please umount the filesystem and > >> rectify the problem(s) > >> 2017-04-18 07:30:18 XFS (sda3): xfs_log_force: error -5 returned. > >> 2017-04-18 07:30:28 XFS (sda3): xfs_log_force: error -5 returned. > >> 2017-04-18 07:30:58 XFS (sda3): xfs_log_force: error -5 returned. > >> 2017-04-18 07:31:28 XFS (sda3): xfs_log_force: error -5 returned. > > > > That looks like some sort of problem with a garbage value in > > di_next_unlinked that we hit when removing inodes from the AGI unlinked > > inode list. > > > > Could you take a metadump and get it to us for further analysis? > > Eric might be interested. ;) > > > > --D > > > >> > >> Stefan > >> Am 17.04.2017 um 17:06 schrieb Stefan Priebe - Profihost AG: > >>> > >>> Am 17.04.2017 um 17:04 schrieb Stefan Priebe - Profihost AG: > >>>> Hi Eric, > >>>> > >>>>> On 4/17/17 2:54 AM, Stefan Priebe wrote: > >>>>>> Hello list, > >>>>>> > >>>>>> one of my servers is crashing every night with the following errors: > >>>>>> > >>>>>> XFS (sda3): xfs_iunlink_remove: xfs_imap returned error -22. > >>>>>> XFS (sda3): xfs_inactive_ifree: xfs_ifree returned error -22 XFS > >>>>> > >>>>> > >>>>> -EINVAL > >>>>> > >>>>>> (sda3): xfs_do_force_shutdown(0x1) called from line 1872 of file > >>>>>> fs/xfs/xfs_inode.c. Return address = 0xffffffff9e33cf53 XFS (sda3): > >>>>>> I/O Error Detected. Shutting down filesystem > >>>>>> > >>>>>> I ran xfs_repair v4.9.0 but it seems unable to recover / fix this > >>>>>> error. I'm running an OpenSUSE 42.2 kernel based von v4.4. Is there > >>>>>> anything i can do to fix this? > >>>>> > >>>>> # echo 11 > /proc/sys/fs/xfs/error_level > >>>>> > >>>>> to get a full backtrace from the shutdown (if you didn't already) > >>>>> so we can at least see what process is encountering the errors. > >>>> > >>>> done. > >>>> > >>>>> I have a hunch, though - are you running xfs_fsr overnight? > >>>> > >>>> No only rsync from dev/a to dev/b. But the reported device is a (source). > >>> > >>> verified: but that job is done 3-4 hours before crashing. > >>> > >>>> > >>>> I also have two empty folders in /lost+found which i cannot delete. Even > >>>> after running xfs_repair. > >>>> > >>>> # rm -r /lost+found/ > >>>> rm: cannot remove ‘/lost+found/536874081’: Directory not empty > >>>> rm: cannot remove ‘/lost+found/536874085’: Directory not empty > >>>> > >>>> Greets, > >>>> Stefan > >>>> > >> -- > >> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in > >> the body of a message to majordomo@xxxxxxxxxxxxxxx > >> More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
