4.1.0 HP PolyServe Software for Microsoft SQL Server administration guide (T5392-96074, October 2010)

These upgrades may require a reboot; however, you can minimize machine downtime
by perform a rolling upgrade. In this scenario, you will need to fail over the Virtual
SQL Server from the primary node, and then upgrade that node while the Virtual
SQL Server resides on the backup node. Clients will see a small downtime from two
failovers (a failover to the backup node before the upgrade, and a failback after the
upgrade) instead of a full upgrade downtime as in a non-clustered environment.
The procedure first upgrades the backup node and then upgrades the primary.
Upgrade the Backup Node
Complete these steps:
1. Ensure that all virtual SQL servers are active on their primary nodes.
2. Disable hosting on the backup node.
3. Apply the upgrade (for example, a service pack).
4. Enable hosting on the backup node.
Upgrade the Primary Node
Complete these steps:
1. Disable hosting on the primary node and verify that all virtual hosts have moved
to a backup node.
2. Apply the upgrade.
3. Enable hosting on the primary node. By default, Virtual SQL Servers are set to
NOFAILBACK and will remain on the backup node. If you want a Virtual SQL
Server to fail back to the primary node, you will need to disable an instance
associated with the Virtual SQL Server on the backup node currently hosting the
Virtual SQL Server. The Virtual SQL Server will then fall back to the primary
node.
Upgrade production SQL Servers156