This figure depicts the deployment view of Oracle VM. In this figure, one Oracle VM Manager host connects to two server pools. Each deployment of server pools is different.
In Server Pool 1, only one server is present. It has no clustering functionality enabled and therefore cannot provide high-availability. If additional servers were added to this pool, it could benefit from the use of shared attached storage and store its resources for the creation of virtual machines in that shared location. If local storage is used, all virtual machines and resources are tied to that single server and cannot be shared with or migrated to another server. If the server crashes, its virtual machines and storage become unavailable or irrecoverable.
In Server Pool 2, a battery of servers is configured as a cluster, using a server pool file system to store essential environment information. The servers in this pool share a virtual IP, so that the pool remains reachable even if one or more servers should fail. One of the servers in the cluster fulfills the Master role and communicates with Oracle VM Manager to execute commands and distribute the required actions among the servers in the pool. This master role can be assumed by and transferred to any of the servers in the pool. Because clustered Oracle VM Servers typically rely on shared external storage, this server pool enables high availability of virtual machines: in case of hardware failure or if excessive load on a given server compromises the operation of a VM, it is live-migrated to another server in the pool without any service interruption for the users.