Clusters and High Availability
An OpenVMS cluster is a set of nodes sharing storage and communicating over a System Communications Services (SCS) interconnect. Candidates must understand the cluster components: the cluster alias, the quorum disk, and the quorum algorithm. Quorum determines whether the cluster continues to run; votes come from nodes and the quorum disk. A cluster must maintain a majority of votes to stay operational, and a node that loses contact may be expelled or may hang depending on quorum.
Know the roles of the cluster interconnect (e.g., Ethernet, Memory Channel) and the shared system disk concept: multiple nodes can boot from the same system disk or use a common system disk. Understand that SCSNODE/SCSSYSTEMID must be unique per node, and that the cluster configuration is stored in a cluster database (e.g., the cluster authorization and configuration files). High availability features include volume shadowing, MSCP serving (disk serving), and failover of batch/print queues and applications.
Traps: assuming a cluster always survives losing any single node (quorum matters); confusing the quorum disk with a data disk; and forgetting that the cluster alias is a network identity, not a node. Also be ready for questions on how a node joins a cluster (boot with the cluster parameters and interconnect) and on shadow-set membership.
Rule of thumb: before declaring a cluster healthy, verify quorum (votes vs. expected), then verify the interconnect, then verify shared storage access.