Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
Hi, list When I umount a DRBD device used by a DRBD client, the 'umount' program hang, and I got the following call trace: ``` [28076.944996] XFS (drbd100): Unmounting Filesystem [28198.933553] drbd web: State change failed: Device is held open by someone [28198.935451] drbd web: Failed: role( Primary -> Secondary ) [28200.828646] INFO: task umount:21558 blocked for more than 120 seconds. [28200.830496] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [28200.831238] umount D ffff8801379fc890 0 21558 21271 0x00000084 [28200.831253] ffff8800a164fd90 0000000000000082 ffff880136345c00 ffff8800a164ffd8 [28200.831282] ffff8800a164ffd8 ffff8800a164ffd8 ffff880136345c00 ffff8801379fc880 [28200.831284] ffff8800a1337d10 ffff8801379fc8c0 ffff8801379fc8e8 ffff8801379fc890 [28200.831289] Call Trace: [28200.831321] [<ffffffff8163a909>] schedule+0x29/0x70 [28200.831516] [<ffffffffa023fc41>] xfs_ail_push_all_sync+0xc1/0x110 [xfs] [28200.831554] [<ffffffff810a6ae0>] ? wake_up_atomic_t+0x30/0x30 [28200.831572] [<ffffffffa022ab08>] xfs_unmountfs+0x68/0x190 [xfs] [28200.831584] [<ffffffffa022b5eb>] ? xfs_mru_cache_destroy+0x6b/0x90 [xfs] [28200.831597] [<ffffffffa022d9c2>] xfs_fs_put_super+0x32/0x90 [xfs] [28200.831621] [<ffffffff811e0ab6>] generic_shutdown_super+0x56/0xe0 [28200.831623] [<ffffffff811e0f07>] kill_block_super+0x27/0x70 [28200.831625] [<ffffffff811e1249>] deactivate_locked_super+0x49/0x60 [28200.831627] [<ffffffff811e1846>] deactivate_super+0x46/0x60 [28200.831632] [<ffffffff811fe785>] mntput_no_expire+0xc5/0x120 [28200.831635] [<ffffffff811ff8bf>] SyS_umount+0x9f/0x3c0 [28200.831642] [<ffffffff81645909>] system_call_fastpath+0x16/0x1b [28320.831678] INFO: task umount:21558 blocked for more than 120 seconds. [28320.833583] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [28320.834275] umount D ffff8801379fc890 0 21558 21271 0x00000084 [28320.834286] ffff8800a164fd90 0000000000000082 ffff880136345c00 ffff8800a164ffd8 [28320.834290] ffff8800a164ffd8 ffff8800a164ffd8 ffff880136345c00 ffff8801379fc880 [28320.834292] ffff8800a1337d10 ffff8801379fc8c0 ffff8801379fc8e8 ffff8801379fc890 [28320.834295] Call Trace: [28320.834349] [<ffffffff8163a909>] schedule+0x29/0x70 [28320.834653] [<ffffffffa023fc41>] xfs_ail_push_all_sync+0xc1/0x110 [xfs] [28320.834688] [<ffffffff810a6ae0>] ? wake_up_atomic_t+0x30/0x30 [28320.834708] [<ffffffffa022ab08>] xfs_unmountfs+0x68/0x190 [xfs] [28320.834724] [<ffffffffa022b5eb>] ? xfs_mru_cache_destroy+0x6b/0x90 [xfs] [28320.834752] [<ffffffffa022d9c2>] xfs_fs_put_super+0x32/0x90 [xfs] [28320.834771] [<ffffffff811e0ab6>] generic_shutdown_super+0x56/0xe0 [28320.834775] [<ffffffff811e0f07>] kill_block_super+0x27/0x70 [28320.834780] [<ffffffff811e1249>] deactivate_locked_super+0x49/0x60 [28320.834783] [<ffffffff811e1846>] deactivate_super+0x46/0x60 [28320.834791] [<ffffffff811fe785>] mntput_no_expire+0xc5/0x120 [28320.834795] [<ffffffff811ff8bf>] SyS_umount+0x9f/0x3c0 [28320.834803] [<ffffffff81645909>] system_call_fastpath+0x16/0x1 ``` Before umount hang, I did the following steps: 1. mount the drbd device at DRBD client (pure client) node; 2. poweroff 2 of the 3 nodes that contains the resource for the DRBD client above 3. write to a file under the mount point at step 1 4. umount the drbd device at step 1 <= umount hang My environment: # cat /etc/centos-release CentOS Linux release 7.2.1511 (Core) # rpm -qa | grep drbd kmod-drbd-9.0.1_3.10.0_327-1.el7.centos.x86_64 drbdmanage-0.99.4-1.noarch drbd-utils-8.9.11-1.el7.centos.x86_64 The call trace doesn't show anything obviously related to DRBD 9. But just in case. -- Thanks, Li Qun -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20170419/dac4c909/attachment.htm>