follow-up: compute CP and MD nodes during upgradation
As per the discussion for implementing per unit timeout, there needs to be the fine tune computation of the nodes during the upgrade process where CP nodes are upgraded first followed by parallel MD nodes upgradation.