Error Message

Jun 7 16:27:17 node1 kernel: GAB INFO V-15-1-20032 Port d closed
Jun 7 16:27:28 node1 kernel: GAB ERROR V-15-1-20015 unconfigure failed: clients still registered <---- GAB is unable to offline the other ports, due to the dependent port is still running.

>>>> Because of this, the network service gets stopped before other VCS services are supposed to be stopped, due to which VCS detects N/W partition at LLT level :
Jun 7 16:27:09 node1 shutdown[14279]: shutting down for system reboot
Jun 7 16:27:10 node1 init: Switching to runlevel: 6
...
...
Jun 7 16:27:17 node1 kernel: GAB INFO V-15-1-20032 Port d closed
Jun 7 16:27:28 node1 kernel: GAB ERROR V-15-1-20015 unconfigure failed: clients still registered
Jun 7 16:27:28 node1 ntpd[9491]: ntpd exiting on signal 15
...
...
Jun 7 16:27:50 node1 kernel: LLT INFO V-14-1-10205 link 1 (eth3) node 1 in trouble
Jun 7 16:27:50 node1 kernel: LLT INFO V-14-1-10205 link 0 (eth2) node 1 in trouble

 

Here is a typical example of the reboot sequence, as found in rc6.d. By default, the kill (or rc-stop) sequence for VCS is K10.

In this example, it has been changed to K99. Because of this, the ports are being unloaded in the wrong order.

# ls -l etc/rc.d/rc6.d/*v*

-rwxr-xr-x 1 support fcf 2619 Feb 1 2011 etc/rc.d/rc6.d/K64vxodm
-rwxr-xr-x 1 support fcf 2077 Oct 4 2010 etc/rc.d/rc6.d/K65vxgms
-r-xr--r-- 1 support fcf 6248 Jan 20 2011 etc/rc.d/rc6.d/K68vxfen
-rwxr-xr-x 1 support fcf 2315 Feb 1 2011 etc/rc.d/rc6.d/K69vxglm
-r-xr-xr-x 1 support fcf 6301 Mar 6 02:33 etc/rc.d/rc6.d/K80vxdbdctrl
-r-xr-x--- 1 support fcf 8749 Apr 8 23:47 etc/rc.d/rc6.d/K80vxpbx_exchanged
-rwxr-xr-x 1 support fcf 3615 Mar 6 02:32 etc/rc.d/rc6.d/K84vras-vradmind.sh
-rwxr-xr-x 1 support fcf 3334 Mar 6 02:32 etc/rc.d/rc6.d/K84vxrsyncd.sh
-r-xr-xr-x 1 support fcf 2578 Mar 6 02:32 etc/rc.d/rc6.d/K85vxnm-vxnetd
-rwxr-xr-x 1 support fcf 4289 Jun 4 23:46 etc/rc.d/rc6.d/K99vcs <<<<<<<<<<<
-rwxr-xr-x 1 support fcf 1763 Mar 6 02:32 etc/rc.d/rc6.d/K99vxfs
-rwxr--r-- 1 support fcf 3313 Mar 6 02:32 etc/rc.d/rc6.d/K99vxvm-boot

>>>> By default the kill (or rc-stop) sequence for VCS is K10, however, in customer's setup, it was K99 (in all the run levels):>>>> Checking the reboot sequence in rc6.d:
# ls -l etc/rc.d/rc6.d/*v*

-rwxr-xr-x 1 support fcf 2619 Feb 1 2011 etc/rc.d/rc6.d/K64vxodm
-rwxr-xr-x 1 support fcf 2077 Oct 4 2010 etc/rc.d/rc6.d/K65vxgms
-r-xr--r-- 1 support fcf 6248 Jan 20 2011 etc/rc.d/rc6.d/K68vxfen
-rwxr-xr-x 1 support fcf 2315 Feb 1 2011 etc/rc.d/rc6.d/K69vxglm
-r-xr-xr-x 1 support fcf 6301 Mar 6 02:33 etc/rc.d/rc6.d/K80vxdbdctrl
-r-xr-x--- 1 support fcf 8749 Apr 8 23:47 etc/rc.d/rc6.d/K80vxpbx_exchanged
-rwxr-xr-x 1 support fcf 3615 Mar 6 02:32 etc/rc.d/rc6.d/K84vras-vradmind.sh
-rwxr-xr-x 1 support fcf 3334 Mar 6 02:32 etc/rc.d/rc6.d/K84vxrsyncd.sh
-r-xr-xr-x 1 support fcf 2578 Mar 6 02:32 etc/rc.d/rc6.d/K85vxnm-vxnetd
-rwxr-xr-x 1 support fcf 4289 Jun 4 23:46 etc/rc.d/rc6.d/K99vcs <<<<<<<<<<<
-rwxr-xr-x 1 support fcf 1763 Mar 6 02:32 etc/rc.d/rc6.d/K99vxfs
-rwxr--r-- 1 support fcf 3313 Mar 6 02:32 etc/rc.d/rc6.d/K99vxvm-boot

To correct the VCS stop sequence, run the following command:

# chkconfig vcs on

This will reset the correct stop/start sequence number for VCS script.