[ WHAT TO DO ]
1. Check VRTSvxfs and VRTSglm packages installed on all cluster nodes
# rpm -aq VRTSvxfs
# rpm -aq VRTSglm
2. Verify VRTSvxfs and VRTSglm packages among cluster nodes
# rpm --verify VRTSvxfs
# rpm --verify VRTSglm
3. If the package is corrupt, it is required to remove all packages, then reinstall
#rpm -aq |grep VRTS > /tmp/pkg.out
# for i in ` cat /tmp/pkg.out`
do
rpm -ev --nodeps $i
done
4. Then reinstall the package only on the issued nodes, linux3 and linux4 . No configuration option was taken.!!
./installsfcfs -install tus1voswebpex03 linux4
5. After packages were installed, a previous configuration was brought in.
#vxinstall
6. Then they brought the VCS online by manual. During this procedure, it is required to check the entries in /etc/sysconfig/llt, /etc/sysconfig/gab, /etc/sysconfig/vxfen to bring these up correctly!!
# /etc/vx/init.d/llt start
# /etc/vx/init.d/gab start
# /etc/vx/init.d/vxfen start
# /etc/vx/init.d/vcs start
Applies To
[ CONFIGURATION ]
4 nodes cluster in SFCFS5.1 for RHEL5
1) linux2(MASTER)
2) linux1, linux3 and linux4(SLAVE)