There are some issues around after you have installed the
Synopsis: VRTSvxvm 5.0_MP1_RP5: Rolling Patch 5 for Volume Manager 5.0 MP1
This patch changes the handling of VxVM devices which is leading to disputes with Sun Cluster 3.2.
host0 # scswitch -z -D testdg -h host1
Sep 26 10:20:17 host0 Cluster.CCR: build_devlink_list: readlink failed for /dev/vx/dsk//global/.devices/node@1/dev/vx/dsk/testdg: No such file or directory
Sep 26 10:23:41 host0 SC[SUNW.HAStoragePlus:6,test-rg,test-hastp-rs,hastorageplus_prenet_start]: Failed to analyze the device special file associated with file system mount point /test/data/AB: No such file or directory
host0 # clrg create test-rg
host0 # clresource create -g test-rg -t SUNW.HAStoragePlus -p FileSystemMountPoints="/testdata" test-rs
clresource: host1 - Failed to analyze the device special file associated with file system mount point /testdata: No such file or directory.
clresource: (C189917) VALIDATE on resource test-rs, resource group test-rg, exited with non-zero exit status.
clresource: (C720144) Validation of resource test-rs in resource group test-rg on node node1 failed.
clresource: (C891200) Failed to create resource "test-rs".
On the other node:
host1# Sep 26 14:27:38 host1 SC[SUNW.HAStoragePlus:4,test-rg,test-rs,hastorageplus_validate]: Failed to analyze the device special file associated with file system mount point /testdata: No such file or directory.
Sep 26 14:27:38 host1 Cluster.RGM.rgmd: VALIDATE failed on resource , resource group , time used: 0% of timeout <1800, seconds>
Do not install patch 124361-06 use patch 124361-05.
Important if patch is already installed: Before backing out 124361-06 ensure that Solaris 10 patch 125731-02 is installed to avoid bug 6622037.
Update 10.Oct.2008: New patch 122058-11 is released which fix the problem and obsoletes 124361-06
Synopsis: VRTSvxvm 5.0MP3: Maintenance Patch for Volume Manager 5.0
Basically the problems all arise when 124361-06 is installed and a VxVM volume is created on a Sun Cluster configuration. With patch 124361-06, when a vxvm volume is created it creates the special device under /devices and then we have symbolic links under /dev/vx/[r]dsk// that point to the /devices entries. This behaviour does not happen when 122058-11 is installed. The special files are created under /dev/vx/[r]dsk/< dg >/ and NOT /devices.
Check if the devices are correct. It's quite important that NO links in the mentioned directory of a device group. Two workarounds available if the wrong links exist. Besure that 122058-11 is already installed and volumes are inactive.
node1# cd /global/.devices/node@1/dev/vx/dsk/testdg
node1# ls -l
lrwxrwxrwx 1 root root 46 Oct 15 16:27 vol01 -> /devices/pseudo/vxio@0:testdg,vol01,59000,blk
lrwxrwxrwx 1 root root 46 Oct 15 16:27 vol02 -> /devices/pseudo/vxio@0:testdg,vol02,59001,blk
node1# rm vol01 vol02
node1# cd /global/.devices/node@1/dev/vx/rdsk/testdg
node1# ls -l
lrwxrwxrwx 1 root root 46 Oct 15 16:27 vol01 -> /devices/pseudo/vxio@0:testdg,vol01,59000,raw
lrwxrwxrwx 1 root root 46 Oct 15 16:27 vol02 -> /devices/pseudo/vxio@0:testdg,vol02,59001,raw
node1# rm vol01 vol02
node1# cldg sync testdg
node1# ls -l /global/.devices/node@1/dev/vx/dsk/testdg
brw------- 1 root root 282, 59000 Oct 15 16:32 vol01
brw------- 1 root root 282, 59001 Oct 15 16:32 vol02
node1# ls -l /global/.devices/node@1/dev/vx/rdsk/testdg
crw------- 1 root root 282, 59000 Oct 15 16:32 vol01
crw------- 1 root root 282, 59001 Oct 15 16:32 vol02
If symlink exist remove the symbolic links
node1# rm /dev/vx/[r]dsk/testdg/symlink
and then recreate the special files using the cluster command
Afterwards for security purposes a reconfiguation boot of all nodes is recommended.
Alert 1019694.1 Sun Cluster Resource "HAstoragePlus" May Fail if Veritas Volume Manager Patch 124361-06 is Installed