[LU-4734] umounts of OST stuck

Duplicate — Tue, 11 Mar 2014 17:31:19 +0000

Hi,

We seem to hit same issue as ~~LU-3230~~.

The umount was in :

crash> bt 3632
PID: 3632 TASK: ffff880455aec080 CPU: 7 COMMAND: "umount"
 #0 [ffff880455c53958] schedule at ffffffff81485765
 #1 [ffff880455c53a20] schedule_timeout at ffffffff81486590
 #2 [ffff880455c53ad0] cfs_schedule_timeout_and_set_state at ffffffffa051469d
[libcfs]
 #3 [ffff880455c53ae0] obd_exports_barrier at ffffffffa05eb89d [obdclass]
 #4 [ffff880455c53b30] filter_precleanup at ffffffffa0bcc0a2 [obdfilter]
 #5 [ffff880455c53b90] class_cleanup at ffffffffa0609f97 [obdclass]
 #6 [ffff880455c53c10] class_process_config at ffffffffa060c323 [obdclass]
 #7 [ffff880455c53cb0] class_manual_cleanup at ffffffffa060d069 [obdclass]
 #8 [ffff880455c53d70] server_put_super at ffffffffa0618f4c [obdclass]
 #9 [ffff880455c53e40] generic_shutdown_super at ffffffff81165f3b
#10 [ffff880455c53e60] kill_anon_super at ffffffff81166056
#11 [ffff880455c53e80] lustre_kill_super at ffffffffa060eca6 [obdclass]
#12 [ffff880455c53ea0] deactivate_super at ffffffff81166ff0
#13 [ffff880455c53ec0] mntput_no_expire at ffffffff811831cf
#14 [ffff880455c53ef0] sys_umount at ffffffff81183c98
#15 [ffff880455c53f80] system_call_fastpath at ffffffff810030f2

and we can see in the console the following messages :

Lustre: ptmp2-OST008d is waiting for obd_unlinked_exports more than 8 seconds. The obd refcount = 6. Is it stuck?
Lustre: ptmp2-OST008d is waiting for obd_unlinked_exports more than 16 seconds. The obd refcount = 6. Is it stuck?
Lustre: DEBUG MARKER: Tue Jan 14 11:45:01 2014

Lustre: ptmp2-OST008d is waiting for obd_unlinked_exports more than 32 seconds. The obd refcount = 6. Is it stuck?
Lustre: ptmp2-OST008d is waiting for obd_unlinked_exports more than 64 seconds. The obd refcount = 6. Is it stuck?

In the server crash we can find the struct obd_device *obd pointer:
0xffff880864076038
and find the connection that the server is waiting for:

crash> struct ptlrpc_connection 0xffff88083b878bc0
struct ptlrpc_connection {
  c_hash = {
    next = 0x0,
    pprev = 0xffff881035375440
  },
  c_self = 1407418007560379,
  c_peer = {
    nid = 1407418007565894,
    pid = 12345
  },
  c_remote_uuid = {
    uuid =
"NET_0x5000a0a643646_UUID\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000"
                       JO.BOO.PI.FO -> lascaux4091
                       JO.BOO.WL.BZF -> lascaux226

  },
  c_refcount = {
    counter = 3
  }
}

Thanks,
Sebastien.

Whamcloud Community JIRA

[LU-4734] umounts of OST stuck