On a Solaris 10 host, if the file system fails to mount after reboot, OS and cluster does not start.
After rebooting the Solaris 10 host, if the service
svc:/system/filesystem/local
fails to mount any of the following file system, the dependent OS or cluster services does not start.
- Local file system
- Shared file system
Note: The preceding file system must be created by
SnapDrive for UNIX.
Workaround
- Find the service that failed to start after the reboot by using the following command:
svcs -vx /network/inetd
Example:
# svcs -vx /network/inetd
svc:/network/inetd:default (inetd)
State: offline since Fri 25 May 2008 04:06:59 PM IST
Reason: Service svc:/system/filesystem/local:default
is not running because a method failed.
- If the
service/system/filesystem/local:default fails, find the persistent file system that is failed to mount by using the following command:
tail /var/svc/log/system-filesystem-local:default.log
Example:
# tail /var/svc/log/system-filesystem-local:default.log
[ May 25 15:58:35 Method "start" exited with status 95 ]
[ May 25 16:07:17 Executing start method
("/lib/svc/method/fs-local") ]
UX:vxfs fsck: ERROR: V-3-20693: cannot access device
/dev/vx/rdsk/dg2/fs1_SdHv:
UX:vxfs mount: ERROR: V-3-20003: Cannot open
/dev/vx/dsk/dg2/fs1_SdHv: No such device or address
UX:vxfs mount: ERROR: V-3-24996: Unable to get disk
layout version
WARNING: /sbin/mountall -l failed: exit status 1
[ May 25 16:07:18 Method "start" exited with status 95 ]
- Comment the entries in
/etc/vfstab for the file system that failed to mount after rebooting the host.
- Restart the OS
service: svc:/system/filesystem/local:default by executing the following commands:
- svcadm disable
svc:/system/filesystem/local:default
- svcadm enable
svc:/system/filesystem/local:default
The OS and cluster service starts.
- For local file system that is failed to mount, complete the following steps:
- Manually mount the file system.
- Uncomment the
/etc/vfstab entry.
- For cluster services, complete the following steps:
- Use the
cfsmount
command to check if the shared file systems are properly mounted.
- Uncomment the
/etc/vfstab entry on all cluster nodes for this file system.
svcadm enable
svc:/system/filesystem/local:default
- If local or cluster file system fails to mount even after completing the preceding steps, contact NetApp technical support.