I have a cluster running R75.40 (yes I know... ^^). I have upgraded member2 to R80.10.
Basically I would like to perform the following steps :
- switch the traffic to member2 (newly upgraded to R80.10) (cphacu start no_dr + cpstop as per ClusterXL admin guide)
- perform pre-upgrade actions on member1 such as Download agent upgrade, etc. These actions include reboots. The actions to be performed are scripted so that our admins only have to run one script that manages the whole upgrade (pre-upgrade actions, CPUSE upgrade, JHF installation, customization...).
My problem is that due to different version numbers (member2 already upgraded) member1 will systematically be active after the reboots.
I have tried to deregister the pnote registered I guess due to version number mismatch : $FWDIR/bin/cphaprob -d admin_down unregister, and then register a new one with the "-p" : $FWDIR/bin/cphaprob -d admin_down -t 0 -s problem -p register but the failover does not occur and the persistant info is not taken into account, most probably due to different versions.
Is there a way to make a cpstop survive to a reboot? This is in order to avoid member1 to become active after pre-upgrade actions and reboot.
Maybe I missed something but I couldn't find anything that could help.