V-16-10031-5562: Failed to lock mountpoint
Article ID: 100048850
Updated On:
Description
Error Message
VCS ERROR V-16-10031-5562 Mount:mnt_res:monitor:Failed tolock mountpoint /testmnt
Cause
This issue is normally caused by the Mount Point being mounted outside of VCS and not mount locked.
20:40:43 VCS ERROR V-16-10031-5592 Mount:mnt_res:monitor:The file system mounted on is not locked by Veritas Cluster Server. Lockingit now.20:40:43 VCS ERROR V-16-10031-5562 Mount:mnt_res:monitor:Failed tolock mountpoint /testmnt
Resolution
The Filesystem will have to be manually umounted first before VCS can take control.
# umount /testmnt
OR
# umount -f /testmnt
If the force umount fails you will need to find which process is still using the Filesystem.
Run lsof and grep for /dev/vx to find all the processes that may still be using the VxFS file systems. (Here we assume that all the VxVM volumes are with VxFS file system created. Please confirm that before killing the processes.) For example:
# lsof | grep /dev/vx | more
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
sleep 23265490 root 1w VREG 37,38001 0 4043311618 /testmnt (/dev/vx/dsk/datadg/testmnt)
Based on the above demonstration of the problem, the following process is killed.
# ps -eaf | grep 232654
root 232654 17891542 0 17:16:33 pts/2 0:00 sleep 10000000
# kill -15 232654
Once you are able to umount /testmnt you will need to bring the Mount Point back online through VCS.
If after killing the process and no other processes show up in lsof, you will need to reboot the node as the Mount Point is stuck in the kernel.
Issue/Introduction
