[Bug 200835] XFS hangs in xfs_reclaim_inode()

[bugzilla-daemon@bugzilla.kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=200835

--- Comment #3 from Peter Klotz (peter.klotz99@gmail.com) ---
Hello Dave

Thanks for the reply. I will upload a complete journald leading to one of the
crashes.

It seems the problem now occurs on a daily basis. The trigger is the deletion
of old backups.

I have a directory structure of this kind:

Backups
 Machine1
  Day1
  Day2
  ...
 Machine2
  Day1
  Day2
  ...

The files in the daily directories are hardlinked, so a minimum amount of space
is wasted. However subdirectories still consume inodes, so lots of them are in
use.

[root@hpmicroserver ~]# df -i
Filesystem                Inodes    IUsed      IFree IUse% Mounted on
/dev/mapper/halvar     390701504 12805954  377895550    4% /mnt/halvar
/dev/mapper/faxe      2929890048 24276605 2905613443    1% /mnt/faxe

The only unusual thing I had to configure is "libata.force=noncq" since the
SATA controller had otherwise problems when talking to the SSD.

Any help (maybe a workaround) is appreciated.

-- 
You are receiving this mail because:
You are watching the assignee of the bug.