How Server for NFS runs on a server cluster

Because Server for NFS is specifically designed to make use of the server cluster, you can use Cluster Administrator to configure network file system (NFS) shared directories as cluster resources. This provides for high availability of those shared directories. This ensures that if the cluster node sharing the directory should be shut down or fail, another node in the cluster will assume the sharing the NFS shared directory until the first node restarts. In addition to enhancing availability, distributing NFS shares among nodes in the cluster can provide for static load balancing among several servers. For more information about sharing NFS directories on a server cluster, see To share a directory as a cluster resource.

Because the NFS shared directory is the cluster resource (instead of the Server for NFS service itself), ownership of individual NFS shares can move easily from one node to the other, as circumstances require. However, configuration changes to Server for NFS (such as the name of the User Name Mapping server) on one node are automatically replicated to instances of Server for NFS running on other cluster nodes. This simplifies administration of multiple servers because you only have to configure Server for NFS on one computer.

For more information, see Installing Server for NFS on a server cluster and Adding a computer running Server for NFS to a server cluster.