Unable to extend the file system in 4 node cluster

[rajasekar_kh]

root@prodbc07:/dev/rdsk]\# vgextend vgplat /dev/dsk/c104t8d7 /dev/dsk/c104t8d6
Current path "/dev/dsk/c107t0d3" is an alternate link, skip.
Current path "/dev/dsk/c107t0d4" is an alternate link, skip.
Current path "/dev/dsk/c107t0d5" is an alternate link, skip.
Current path "/dev/dsk/c107t0d6" is an alternate link, skip.
Current path "/dev/dsk/c107t0d7" is an alternate link, skip.
Current path "/dev/dsk/c107t1d0" is an alternate link, skip.
Current path "/dev/dsk/c107t1d1" is an alternate link, skip.
Current path "/dev/dsk/c107t1d2" is an alternate link, skip.
Current path "/dev/dsk/c107t1d3" is an alternate link, skip.
Current path "/dev/dsk/c107t1d4" is an alternate link, skip.
Current path "/dev/dsk/c107t1d5" is an alternate link, skip.
Current path "/dev/dsk/c107t1d6" is an alternate link, skip.
Current path "/dev/dsk/c107t1d7" is an alternate link, skip.
vgextend: Couldn't install the physical volume "/dev/dsk/c104t8d7".
Device busy
vgextend: Couldn't install the physical volume "/dev/dsk/c104t8d6".
Device busy

==

i tried the disk in another cluster node.

[root@prodbc05:/dev/rdsk]\# ls -l |grep -i c104t8d7
crw-r-----   1 bin        sys        188 0x688700 May 25  2010 c104t8d7

===

Newly allocated luns

crw-r-----   1 bin        sys        188 0x688700 Jul  8 00:04 c104t8d7
crw-r-----   1 bin        sys        188 0x689000 Jul  8 00:04 c104t9d0
crw-r-----   1 bin        sys        188 0x688600 Jul  8 00:04 c104t8d6
crw-r-----   1 bin        sys        188 0x6b8700 Jul  8 00:04 c107t8d7
crw-r-----   1 bin        sys        188 0x6b8600 Jul  8 00:04 c107t8d6
crw-r-----   1 bin        sys        188 0x6b9000 Jul  8 00:04 c107t9d0

how to extend the file systems

[Patrick Wallek]

The general steps to take to extend a SG volume group are:

1) Extend the VG on one node to include the appropriate disks

2) run a 'vgexport -m /tmp/vgplat.map -p -v -s vgplat' on that node

3) Copy the /tmp/vgplat.map to the other nodes in the cluster

On all other cluster nodes:

4) 'vgexport vgplat' to get rid of the VG (assuming it is only used on one node at a time)

5) Recreate the /dev/vgplat directory and /dev/vgplat/group file (via mknod)

6) Import the VG via 'vgimport -m /tmp/vgplat.map -v -s vgplat'