Upgrading Failover Zones

A failover zone is a non-global zone that is configured with the HA for Oracle Solaris Zones data service so that it can be managed by the Oracle Solaris Cluster software. If you have failover zones of brand type solaris configured on the cluster, perform this procedure in conjunction with the procedures for the cluster upgrade method that you use.

How to Upgrade Failover Zones

1. Determine the nodes where the resource group that manages the failover zone of brand type solaris is offline.

   # clresourcegroup status resourcegroup

2. Suspend the resource group that manages the failover zone of brand type solaris.

   # clresourcegroup suspend resourcegroup

3. Perform the standard or rolling upgrade on the nodes where the resource group that manages the failover zones is offline.

   Perform all steps, including booting into the new boot environment, in How to Perform a Standard Upgrade or How to Perform a Rolling Upgrade. Then return to this procedure.

4. Perform the standard or rolling upgrade on the nodes where the resource group that manages the failover zones of brand type solaris is online.

   Perform Steps 1–3 in How to Perform a Standard Upgrade or How to Perform a Rolling Upgrade. Do not perform Step 4 to boot the machine into the new boot environment. Return to this procedure after you complete this task.

5. Determine the universally unique ID (UUID) of the updated boot-environment root dataset of the node where the resource group that manages the failover zone of brand type solaris is online.

   You will see output similar to the following:

   phys-schost-1# beadm list -H
   ...
   b175b-SC;8fe53702-16c3-eb21-ed85-d19af92c6bbd;R;/;756
   ...

   In this example, the UUID is 8fe53702-16c3-eb21-ed85-d19af92c6bbd and the updated boot environment is b175b-SC.

6. Set the same UUID on the updated boot environment for the other nodes where the resource group that manages the failover zone of brand type solaris is offline.

   phys-schost-2# zfs set org.opensolaris.libbe:uuid=uuid rpool/ROOT/BE

   For example:

   phys-schost-2# zfs set org.opensolaris.libbe:uuid=8fe53702-16c3-eb21-ed85-d19af92c6bbd rpool/ROOT/b175b-SC

7. Resume the resource group that manages the failover zone of brand type solaris.

   # clresourcegroup resume resourcegroup

8. Switch the resource group that manages the failover zone of brand type solaris to a node that was already booted into the updated boot environment and verify that the zone started correctly.

   # clresourcegroup switch -n phys-schost-2 resourcegroup

9. Boot the node from Step 4, where the resource group that manages the failover zone of brand type solaris had initially been online, into its updated boot environment.

Next Steps

Go to Chapter 5, Completing the Upgrade.