Installation manual

Preparing to Upgrade in an OpenVMS Cluster Environment
6.5 Rolling Upgrade
6.5 Rolling Upgrade
This section describes the following:
How a rolling upgrade works
Notes and restrictions
Tasks you need to perform to prepare your system for a rolling upgrade
6.5.1 How a Rolling Upgrade Works
During a rolling upgrade, you upgrade each system disk individually, allowing
old and new versions of the operating system to run together in the same cluster,
creating a mixed-version cluster. Because rolling upgrades allow mixed-version
clusters, the systems that you are not upgrading remain available.
6.5.2 Notes and Restrictions
The following restrictions apply to rolling upgrades:
Compaq recommends that all VAX computers in a cluster run the same
version of the operating system. Mixed-version clusters are supported only
for the purpose of upgrading the entire cluster incrementally. Compaq further
recommends that all VAX systems in an OpenVMS Cluster run the latest
version of the OpenVMS VAX operating system.
A rolling upgrade does not apply when all systems boot from a single system
disk. When all systems boot from a single system disk, perform a concurrent
upgrade using the procedure described in Section 6.4.
During the rolling upgrade, the queuing system is disabled on the nodes being
upgraded. The queuing system on the other nodes should be unaffected by
the upgrade as long as a START/QUEUE/MANAGER/ON=(upgrade
node . . .) command is not executed. A START/QUEUE/MANAGER
command could attempt to move the queue manager process to the node
being upgraded if the /ON qualifier indicates that it is one of the preferred
nodes. If the queue manager attempts to run on the node being upgraded, the
queuing system will hang on the entire cluster.
If your startup file contains a START/QUEUE/MANAGER/ON=(upgrade
node1, node2 . . . noden) command, do not boot in any nonupgrading
nodes until you have removed that command from the startup file.
Booting in a node causes the queuing system to hang on the cluster if the
START/QUEUE/MANAGER/ON=(upgrade node, . . . ) command is executed.
When the upgrade has completed, you can add the command back to the
startup file.
If the cluster queuing system hangs during the upgrade, it is most likely
caused by the queue manager attempting to run on the upgrading node. To
clear the condition, move the queue manager to a node you are not upgrading
by entering the following command:
START/QUEUE/MANAGER/ON=(nonupgrading node1, -
node2 . . . noden)
6–4 Preparing to Upgrade in an OpenVMS Cluster Environment