Before you start
Come from Virtualisation — clustering concepts build directly on understanding what a VM or server actually is.
CLUSTER 0.8
One server is a single point of failure. Clustering links multiple servers together so that if one fails, another takes over — often without anyone noticing.
Drag to rotate · click a node to jump to that topic
Think of it as: a backup singer who already knows every lyric, ready to take the mic instantly
A cluster is a group of servers working together so that if one fails, the workload it was running moves to another member of the cluster — ideally with zero or near-zero interruption.
Think of it as: the actual moment the backup takes over
Failover is the process of detecting a failure and automatically shifting the workload to a healthy node. How fast and how seamless that handoff is — measured in seconds or milliseconds — is what separates a good HA design from a mediocre one.
This cluster is the two-server, on-premises version of an idea the Resilience & Continuity domain takes much further at cloud scale — availability zones, region pairs, and multi-region failover are the same core concept, just spread across entire datacenters instead of two machines in a rack.
REAL-WORLD SCENARIO
A company's order-processing database runs on a two-node cluster. One night, the primary server's hardware fails entirely. The cluster detects the failure and fails over to the second node automatically — orders keep processing, and most customers never know anything happened.