High availability requires:

• Two homogenous, network systems with similar configurations in terms of resources such as CPU, memory, and so on

• The high availability package and the associated hardware

• Devices that are accessible to both nodes

• A logical volume manager (LVM) to maintain unique device path names across the cluster nodes

• Volumes or disk suite objects on the multihost disks

• Third-party vendor mirroring for media failure protection

• Logical host name that can be bound to the primary or secondary node. In a symmetric configuration, you need two logical host names, each corresponding to a primary companion.

See the documentation for Sun Cluster for more information about requirements, and for installing platform-specific high availability software.