Failed to mount cluster mount point

[Muhammad Ahmad]

it's a SG-CFS 3-nodes cluster, one of SG-CFS-pkg is not down, not coming up.

ALL SG-CFS-DG-x packages are up, but one of SG-CFS-MP Package was down.

package logs showing:

08/03/10 23:14:05: ########### Node "node-x": Starting package ###########
UX:vxfs mount: ERROR: V-3-21268: /dev/vx/dsk/dgxx/volxxx is corrupted. needs checking
List volume device file /dev/vx/dsk/dgxx/volxxx:
brw------- 1 root root 2 0x006d60 Aug 3 23:13 /dev/vx/dsk/dgxx/volxxx
List if /dev/vx/dsk/dgxx/volxxx is mounted:
List if /cfs/xxxxx is mounted:
List processes using /dev/vx/dsk/dgxx/volxxx:
/dev/vx/dsk/dgxx/volxxx:
List processes using /cfs/xxxx:
/cfs/xxxx:
08/03/10 23:14:05: ERROR: Failed to mount cluster mount point "/cfs/xxxx" on "dgxx/volxxx" with "cluster,rw" option
awk: Cannot find or open file -F=.

Version details:
---------------
SG Version: A.11.19.00
HPUX 11.31
A.02.01 HP Serviceguard Cluster File System for RAC with DCOE
B.05.00.01 Veritas File System Bundle 5.0 for HP-UX
5.0.31.0 Veritas Oracle Disk Manager



An earliest would be highly oblidged.

Regards,
Muhammad Ahmad

[Matti_Kurkela · Accepted solution]

Did you notice it says: "/dev/vx/dsk/dgxx/volxxx is corrupted. needs checking"?

That means you must run "fsck -F vxfs /dev/vx/rdsk/dgxx/volxxx". (NOTE: "rdsk" not "dsk")

You may have to run this command on all three nodes, one node at a time. If the fsck operation is successful on all three nodes, the CFS filesystem should be mountable again.

MK

[Stephen Doud]

See also http://forums13.itrc.hp.com/service/forums/questionanswer.do?admit=109447627+1280926010339+28353475&threadId=1163552

Per your email to me, my suggestion to run
# fsck –F vxfs –o full /dev/vx/rdsk/dginfa/volinfa
corrected the problem.