Unlike Solaris, RedHat Linux assigns a minor number foreach volume instead for each diskgroup:

[guest@posterboy /]# ls -l/dev/vx/dsk/datadg/vol01
brw-------    1root     root     199,   7 Aug 22 16:40/dev/vx/dsk/datadg/vol01

The minor number assigned to avolume will depend on in what order the disk groups are created. So when ClusterServer fails over the service groups and import disk groups, if Volume Managerfinds out there are conflicting minor numbers, it will assign a temporary minornumber to the conflicting volume, this will cause an NFS stale handle on NFSclients if the volume is used by an NFSshare.

[guest@posterboy /]# vxdg importdata2dg
vxvm:vxdg: Warning: Volumevol01: Temporarily renumbered due toconflict
vxvm:vxdg: Warning: Volumevol02: Temporarily renumbered due to conflict

Best practiceto configure the NFS service group's disk groups:

1.  Create all NFSdisk groups and volumes in one node ( system )
2.  Verify the all diskgroup volumes minor number; make sure these minor numbers are not used in othernodes by other volumes
3.  Change the conflicting minor numbers ifneeded.  It will be a good idea to make a range of minor numbers availableon both system to use for NFS disk groups only

To change the minornumber, use the vxdg reminor command asfollowing. Please note that the first volume in the disk group will use thegiven minor number, the second volume will use the next serialized minor number,and so on.

[guest@posterboy /]# vxdg -gdata2dg -f reminor data2dg11
[guest@posterboy /]# ls -l/dev/vx/dsk/data2dg/vol01
brw-------    1root     root     199,  11 Aug 29 17:46/dev/vx/dsk/data2dg/vol01
[guest@posterboy/]# ls -l/dev/vx/dsk/data2dg/vol02
brw-------    1root     root     199,  12 Aug 29 17:47/dev/vx/dsk/data2dg/vol02