All of lore.kernel.org
 help / color / mirror / Atom feed
From: bugzilla-daemon@bugzilla.kernel.org
To: linux-xfs@kernel.org
Subject: [Bug 200835] New: XFS hangs in xfs_reclaim_inode()
Date: Thu, 16 Aug 2018 16:03:18 +0000	[thread overview]
Message-ID: <bug-200835-201763@https.bugzilla.kernel.org/> (raw)

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

            Bug ID: 200835
           Summary: XFS hangs in xfs_reclaim_inode()
           Product: File System
           Version: 2.5
    Kernel Version: 4.14.62
          Hardware: Intel
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: normal
          Priority: P1
         Component: XFS
          Assignee: filesystem_xfs@kernel-bugs.kernel.org
          Reporter: peter.klotz99@gmail.com
        Regression: No

Created attachment 277895
  --> https://bugzilla.kernel.org/attachment.cgi?id=277895&action=edit
Hang of kernel 4.14.62

The attached file shows backtraces that occur in quick succession and
ultimately lead to a complete server hang.

The backtraces seem related to inode reclamation.

For example the first backtrace:

Aug 16 02:33:30 hpmicroserver kernel: INFO: task khugepaged:28 blocked for more
than 120 seconds.
Aug 16 02:33:30 hpmicroserver kernel:       Not tainted 4.14.62-1-lts #1
Aug 16 02:33:30 hpmicroserver kernel: "echo 0 >
/proc/sys/kernel/hung_task_timeout_secs" disables this message.
Aug 16 02:33:30 hpmicroserver kernel: khugepaged      D    0    28      2
0x80000000
Aug 16 02:33:30 hpmicroserver kernel: Call Trace:
Aug 16 02:33:30 hpmicroserver kernel:  ? __schedule+0x284/0x860
Aug 16 02:33:30 hpmicroserver kernel:  schedule+0x28/0x80
Aug 16 02:33:30 hpmicroserver kernel:  schedule_timeout+0x292/0x370
Aug 16 02:33:30 hpmicroserver kernel:  ? check_preempt_curr+0x62/0x90
Aug 16 02:33:30 hpmicroserver kernel:  wait_for_completion+0xaf/0x140
Aug 16 02:33:30 hpmicroserver kernel:  ? wake_up_q+0x70/0x70
Aug 16 02:33:30 hpmicroserver kernel:  flush_work+0x116/0x1d0
Aug 16 02:33:30 hpmicroserver kernel:  ? worker_detach_from_pool+0xa0/0xa0
Aug 16 02:33:30 hpmicroserver kernel:  xlog_cil_force_lsn+0x78/0x210 [xfs]
Aug 16 02:33:30 hpmicroserver kernel:  ? enqueue_task_fair+0x5a/0x500
Aug 16 02:33:30 hpmicroserver kernel:  ? native_sched_clock+0x37/0x90
Aug 16 02:33:30 hpmicroserver kernel:  ? __switch_to_asm+0x40/0x70
Aug 16 02:33:30 hpmicroserver kernel:  _xfs_log_force_lsn+0x71/0x340 [xfs]
Aug 16 02:33:30 hpmicroserver kernel:  ? try_to_wake_up+0x54/0x4b0
Aug 16 02:33:30 hpmicroserver kernel:  ? update_group_capacity+0x27/0x1e0
Aug 16 02:33:30 hpmicroserver kernel:  ? xfs_reclaim_inode+0xe3/0x340 [xfs]
Aug 16 02:33:30 hpmicroserver kernel:  __xfs_iunpin_wait+0xa7/0x160 [xfs]
Aug 16 02:33:30 hpmicroserver kernel:  ? bit_waitqueue+0x30/0x30
Aug 16 02:33:30 hpmicroserver kernel:  xfs_reclaim_inode+0xe3/0x340 [xfs]
Aug 16 02:33:30 hpmicroserver kernel:  xfs_reclaim_inodes_ag+0x1b1/0x300 [xfs]
Aug 16 02:33:30 hpmicroserver kernel:  xfs_reclaim_inodes_nr+0x31/0x40 [xfs]
Aug 16 02:33:30 hpmicroserver kernel:  super_cache_scan+0x152/0x1a0
Aug 16 02:33:30 hpmicroserver kernel:  shrink_slab.part.45+0x1e8/0x3c0
Aug 16 02:33:30 hpmicroserver kernel:  shrink_node+0x123/0x310
Aug 16 02:33:30 hpmicroserver kernel:  do_try_to_free_pages+0xc3/0x330
Aug 16 02:33:30 hpmicroserver kernel:  try_to_free_pages+0xf4/0x1b0
Aug 16 02:33:30 hpmicroserver kernel:  __alloc_pages_slowpath+0x3e4/0xd80
Aug 16 02:33:30 hpmicroserver kernel:  ? __switch_to+0x170/0x4b0
Aug 16 02:33:30 hpmicroserver kernel:  ? __switch_to_asm+0x34/0x70
Aug 16 02:33:30 hpmicroserver kernel:  ? __switch_to_asm+0x34/0x70
Aug 16 02:33:30 hpmicroserver kernel:  ? __switch_to_asm+0x40/0x70
Aug 16 02:33:30 hpmicroserver kernel:  __alloc_pages_nodemask+0x226/0x240
Aug 16 02:33:30 hpmicroserver kernel:  khugepaged_alloc_page+0x17/0x50
Aug 16 02:33:30 hpmicroserver kernel:  khugepaged+0xbcb/0x2120
Aug 16 02:33:30 hpmicroserver kernel:  ? wait_woken+0x80/0x80
Aug 16 02:33:30 hpmicroserver kernel:  ? collapse_shmem+0xb90/0xb90
Aug 16 02:33:30 hpmicroserver kernel:  kthread+0x119/0x130
Aug 16 02:33:30 hpmicroserver kernel:  ? __kthread_parkme+0xa0/0xa0
Aug 16 02:33:30 hpmicroserver kernel:  ret_from_fork+0x22/0x40


The problem has occurred twice so far. Once in kernel 4.14.52 and once in
4.14.62. I am using the Arch Linux LTS kernel so it is more or less the vanilla
kernel.

The server has two XFS filesystems. One is a 30TB XFS/LUKS/RAID setup, the
other one is a plain 4TB XFS volume.

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

             reply	other threads:[~2018-08-16 19:02 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-08-16 16:03 bugzilla-daemon [this message]
2018-08-16 16:04 ` [Bug 200835] XFS hangs in xfs_reclaim_inode() bugzilla-daemon
2018-08-17  2:36 ` bugzilla-daemon
2018-08-17  5:32 ` bugzilla-daemon
2018-08-17  5:34 ` bugzilla-daemon
2018-08-17  8:43 ` bugzilla-daemon
2018-08-17 11:55 ` bugzilla-daemon
2018-08-17 13:06 ` bugzilla-daemon
2018-08-17 13:25 ` bugzilla-daemon
2018-08-17 14:04 ` bugzilla-daemon
2018-08-17 14:06 ` bugzilla-daemon

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=bug-200835-201763@https.bugzilla.kernel.org/ \
    --to=bugzilla-daemon@bugzilla.kernel.org \
    --cc=linux-xfs@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.