Project

General

Profile

Actions
Copy link

Bug #200

closed

umount hangs with clustered mds

Added by Yehuda Sadeh almost 14 years ago. Updated over 13 years ago.

Status:
Resolved
Priority:
Normal
Assignee:
-
Category:
-
Target version:
-
% Done:

0%

Source:
Tags:
Backport:
Regression:
Severity:
Reviewed:
Affected Versions:
ceph-qa-suite:
Crash signature (v1):
Crash signature (v2):

Description

Happens on both current master and the unstable branches (6/14). The umount at the following scenario never exits:

  1. cd /mnt
  2. tar xvfj ~/linux-2.6.32.4.tar.bz2
    ...
  3. rm -fR linux-2.6.32.4
  4. cd; umount /mnt

after a while we get the following:

[ 1383.000000] INFO: task umount:1270 blocked for more than 120 seconds.
[ 1383.000000] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1383.000000] umount D 00000000402b3210 0 1270 1234 0x00000000
[ 1383.000000] 7021b988 7021b4b8 6032fc00 64398000 64399a20 60015cd0 64399a20 7001e5e0
[ 1383.000000] 64398000 6f8aed00 64399a70 60237cb0 6f80a140 6004576b 64398000 7001e5e0
[ 1383.000000] 64399aa0 6f9b2b70 7001e3a0 7001e3a8 64399b00 70c96935 00013bd1 00007e65
[ 1383.000000] Call Trace:
[ 1383.000000] 643999f8: [<60015cd0>] _switch_to+0x5e/0xae
[ 1383.000000] 64399a28: [<60237cb0>] schedule+0x1d9/0x21f
[ 1383.000000] 64399a38: [<6004576b>] prepare_to_wait+0x61/0x6e
[ 1383.000000] 64399a78: [<70c96935>] ceph_mdsc_sync+0x26a/0x36a [ceph]
[ 1383.000000] 64399ab0: [<600454f4>] autoremove_wake_function+0x0/0x38
[ 1383.000000] 64399b08: [<70c77ad7>] ceph_syncfs+0x85/0xde [ceph]
[ 1383.000000] 64399b28: [<600a1c7a>] __sync_filesystem+0x66/0x7e
[ 1383.000000] 64399b48: [<600a1d36>] sync_filesystem+0x34/0x4a
[ 1383.000000] 64399b68: [<6008533d>] generic_shutdown_super+0x26/0xe1
[ 1383.000000] 64399b78: [<70c966bf>] ceph_mdsc_pre_umount+0x287/0x293 [ceph]
[ 1383.000000] 64399b98: [<60085443>] kill_anon_super+0x11/0x4a
[ 1383.000000] 64399bb8: [<70c76436>] ceph_kill_sb+0x78/0x85 [ceph]
[ 1383.000000] 64399bd8: [<60084862>] deactivate_locked_super+0x20/0x40
[ 1383.000000] 64399bf8: [<60084cba>] deactivate_super+0x3f/0x43
[ 1383.000000] 64399c18: [<60098e35>] mntput_no_expire+0x6b/0x96
[ 1383.000000] 64399c38: [<600993d6>] sys_umount+0x2dd/0x30a
[ 1383.000000] 64399cd8: [<60018179>] handle_syscall+0x59/0x70
[ 1383.000000] 64399cf8: [<600273eb>] userspace+0x3c0/0x465
[ 1383.000000] 64399fc8: [<60015bcf>] fork_handler+0x62/0x69
[ 1383.000000] 64399fd0: [<70c8fa73>] ceph_msg_kfree+0x40/0x75 [

Actions
Copy link
 #1

Updated by Yehuda Sadeh almost 14 years ago

Seems to be waiting forever here:

wait_event(mdsc->cap_flushing_wq, check_cap_flush(mdsc, want_flush));
Actions
Copy link
 #2

Updated by Sage Weil almost 14 years ago

  • Subject changed from umount hangs to umount hangs with clustered mds
Actions
Copy link
 #3

Updated by Yehuda Sadeh almost 14 years ago

  • Status changed from New to Resolved

Fixed by 364521acb070ab60540f869f4ccd79b468c3d9b0.

Actions
Copy link

Also available in: Atom PDF