ZooKeeper

A replicated coordination service for small metadata, naming, membership, watches, locks, and leader-election recipes.

Version-aware ZooKeeper development/ensemble setup, quorum configuration, storage, networking, and operational safeguards.

Ephemeral sequential lock recipes, predecessor watches, session failure, herd avoidance, and fencing requirements.

Service membership with ephemeral nodes and watches, including stale-cache and session-expiration behavior.

A ZooKeeper recipe for application leader election using ephemeral sequential nodes and predecessor watches.

ZooKeeper's leader-based atomic broadcast protocol, proposal ordering, quorum acknowledgment, recovery, and its distinction from Paxos/Raft.

Watch notifications, one-shot/re-registration semantics, ordering, disconnect gaps, and cache reconciliation.

The hierarchical znode namespace and why it resembles a filesystem while serving coordination metadata rather than files.

Persistent, ephemeral, sequential nodes; versions; ACLs; session ownership; and coordination semantics.

Majority-quorum availability, leader failover, ensemble sizing, and why more servers do not always increase fault tolerance.