I would do the following.
1. Case 1 -- If thats ok to reboot the systems before your maintanance.
move following files to a temporary location before you reboot the box.
Stop had on all nodes
/opt/VRTSvcs/bin/hastop -all ( if you want your resource up an running at this stage - Use hastop -all -force on any node)
1. /etc/rc2.d/S70llt
2. /etc/rc2.d/S92gab
3. /etc/rc2.d/S92vxfen
So ,
cd /etc/rc2.d
mv S70llt s70llt < destination file has a lower case "s">
mv S92gab s92gab
mv S97vxfen s97vxfen
All you need to do is stop llt from starting as gab wont start without llt., But I do move all the files and move them back to their original file name once the maintenance task is over.
Reverse the above step and reboot or manually start these drivers/modules once you are ready to start to cluster .
2. Case 2 . Disable cluster without reboot.
Stop the cluster as I mentioned earlier.
2. stop gab
/sbin/gabconfig -u
3.Stop LLT
/sbin/lltconfig -U
4. Unload llt and gab modules
First list module ids
modinfo |grep llt
modinfo |grep gab
Unload gab first and then lln
modunload -i <llt id you got from modinfo>
modunload -i <gab's id from modinfo>
Unload vxfen too if you have that one configured. I assume that yours not a CFS cluster .
Now your system have no VCS stuff in it. However if you reboot all these modules will be reloaded if you dont do the case 1.
