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
