OS crashes at nsm_mon_unmon+0x89 when using docker and NFS in Red Hat Enterprise Linux 7

Issue

• The host OS crashes while using docker. I was able to obtain a core dump of the OS and get the following traceback:

[525488.023062] Call Trace:
[525488.023078]  [<ffffffffa009d5d2>] nsm_unmonitor+0xc2/0x1b0 [lockd]
[525488.023100]  [<ffffffffa0097a1e>] nlm_destroy_host_locked+0x7e/0xe0 [lockd]
[525488.023123]  [<ffffffffa0097caa>] nlmclnt_release_host+0x8a/0x160 [lockd]
[525488.023146]  [<ffffffffa00950d6>] nlmclnt_done+0x16/0x30 [lockd]
[525488.023168]  [<ffffffffa0a4d667>] nfs_destroy_server+0x17/0x20 [nfs]
[525488.023190]  [<ffffffffa0a4e363>] nfs_free_server+0x33/0xd0 [nfs]
[525488.023212]  [<ffffffffa0a5a6fb>] nfs_kill_super+0x2b/0x30 [nfs]
[525488.023992]  [<ffffffff811e1249>] deactivate_locked_super+0x49/0x60
[525488.024759]  [<ffffffff811e1846>] deactivate_super+0x46/0x60
[525488.025523]  [<ffffffff811fe785>] mntput_no_expire+0xc5/0x120
[525488.026285]  [<ffffffff811ff8bf>] SyS_umount+0x9f/0x3c0
[525488.027067]  [<ffffffff81645909>] system_call_fastpath+0x16/0x1b

• Another pattern of call trace.

[9071522.374951] Call Trace:
[9071522.374971]  [<ffffffffa00a65d2>] nsm_unmonitor+0xc2/0x1b0 [lockd]
[9071522.374993]  [<ffffffffa00a0a1e>] nlm_destroy_host_locked+0x7e/0xe0 [lockd]
[9071522.375017]  [<ffffffffa00a0caa>] nlmclnt_release_host+0x8a/0x160 [lockd]
[9071522.375040]  [<ffffffffa009e0d6>] nlmclnt_done+0x16/0x30 [lockd]
[9071522.375067]  [<ffffffffa07f4667>] nfs_destroy_server+0x17/0x20 [nfs]
[9071522.375089]  [<ffffffffa07f5363>] nfs_free_server+0x33/0xd0 [nfs]
[9071522.375112]  [<ffffffffa08016fb>] nfs_kill_super+0x2b/0x30 [nfs]
[9071522.375134]  [<ffffffff811e1249>] deactivate_locked_super+0x49/0x60
[9071522.375154]  [<ffffffff811e1846>] deactivate_super+0x46/0x60
[9071522.375174]  [<ffffffff811fe785>] mntput_no_expire+0xc5/0x120
[9071522.375192]  [<ffffffff811fe804>] mntput+0x24/0x40
[9071522.375208]  [<ffffffff811fe938>] namespace_unlock+0x118/0x130
[9071522.375912]  [<ffffffff812011bb>] put_mnt_ns+0x4b/0x60
[9071522.376591]  [<ffffffff810aac1b>] free_nsproxy+0x1b/0x90
[9071522.377264]  [<ffffffff810aae65>] switch_task_namespaces+0x45/0x50
[9071522.377947]  [<ffffffff810aae80>] exit_task_namespaces+0x10/0x20
[9071522.378614]  [<ffffffff810815e6>] do_exit+0x2c6/0xa60
[9071522.379262]  [<ffffffff810b42ba>] ? finish_task_switch+0x14a/0x170
[9071522.379897]  [<ffffffff8163a2b8>] ? __schedule+0x2d8/0x900
[9071522.380515]  [<ffffffff81081dff>] do_group_exit+0x3f/0xa0
[9071522.381112]  [<ffffffff81092c10>] get_signal_to_deliver+0x1d0/0x6d0
[9071522.381698]  [<ffffffff8108fddb>] ? recalc_sigpending+0x1b/0x50
[9071522.382267]  [<ffffffff81014417>] do_signal+0x57/0x6c0
[9071522.382815]  [<ffffffff81093c4f>] ? do_sigtimedwait+0x21f/0x2a0

It often coincides to when the docker processes exits.