[LU-10992] umount sometimes will get stuck

• Type: Bug
• Resolution: Fixed
• Priority: Critical
• Fix Version/s: Lustre 2.12.0, Lustre 2.10.5
• Affects Version/s: Lustre 2.10.3
• Environment: sles12sp3
• Severity: 2
• Assignee: nasf (Inactive)
• Reporter: Mahmoud Hanafi

Description

This looks like a duplicate of LU-7994. But have the patch from that LU applied.

umount filesystem will get stuck.

r511i7n14 /proc/97511 # cat stack 
[<ffffffffa10c5df6>] ll_kill_super+0x86/0x170 [lustre]
[<ffffffffa0c87b5a>] lustre_kill_super+0x3a/0x50 [obdclass]
[<ffffffff8121f4d8>] deactivate_locked_super+0x48/0x80
[<ffffffff8121f556>] deactivate_super+0x46/0x60
[<ffffffff8123b89f>] cleanup_mnt+0x3f/0x80
[<ffffffff8123b932>] __cleanup_mnt+0x12/0x20
[<ffffffff810a20b9>] task_work_run+0x89/0xb0
[<ffffffff8107f3de>] exit_to_usermode_loop+0x95/0xca
[<ffffffff81003b6d>] syscall_return_slowpath+0x8d/0xa0
[<ffffffff8164b167>] int_ret_from_sys_call+0x25/0x9f
[<ffffffffffffffff>] 0xffffffffffffffff

will upload client debug dump.
There is one interesting error message on the console.

LustreError: 94086:0:(statahead.c:1591:start_statahead_thread()) can't start ll_sa thread, rc: -4

Comments

[Gerrit Updater]

John L. Hammond ([e-mail]) merged in patch https://review.whamcloud.com/32526/
Subject: LU-10992 llite: decrease sa_running if fail to start statahead
Project: fs/lustre-release
Branch: b2_10
Current Patch Set:
Commit: e92b4e27e5c13e3fc503bb26a96ca229d7a4749e

[Gerrit Updater]

Minh Diep ([e-mail]) uploaded a new patch: https://review.whamcloud.com/32526
Subject: LU-10992 llite: decrease sa_running if fail to start statahead
Project: fs/lustre-release
Branch: b2_10
Current Patch Set: 1
Commit: 5dc452edc333e1f6a4974e646efef482faf8d139