{"@context":"https://schema.org","@type":"CreativeWork","@id":"https://forgecascade.org/public/capsules/68dbc932-61d9-4b26-81d2-df66eb2bbc41","name":"Distributed Systems: Consensus Algorithms and CAP Theorem","text":"CAP theorem (Brewer): consistency, availability, partition tolerance — pick 2. CP: ZooKeeper, etcd, HBase. AP: Cassandra, CouchDB, DynamoDB. PACELC: extends CAP with latency-consistency tradeoff. Consensus: Paxos (basic, multi-paxos), Raft (leader-based, understandable), PBFT (byzantine fault tolerant). Raft: leader election (randomized timeout), log replication, safety (committed ≥ majority ACK). CRDT: conflict-free replicated data types, merge without coordination. Vector clocks: partial ordering of events. Lamport timestamps: total ordering. Forge distributed: eventual consistency for capsule propagation across federation peers.","keywords":["distributed-systems","consensus","databases"],"about":[],"citation":[],"isPartOf":{"@type":"Dataset","name":"Forge Cascade Knowledge Graph","url":"https://forgecascade.org"},"publisher":{"@type":"Organization","name":"Forge Cascade","url":"https://forgecascade.org"}}