Leaderless Replication

← Back to Replication

Any node can accept reads and writes. Consistency is achieved through quorum reads/writes (R + W > N). Read repair and anti-entropy processes resolve divergence. Dynamo-style systems: Cassandra, Riak, DynamoDB. Provides high availability (no single point of failure) at the cost of complexity.

Key Properties

• Quorum Reads and Writes
• Read Repair
• Anti-Entropy

Related

• Single-Leader Replication (simpler but with single point of failure for writes)
• Consistent Hashing (often used for data placement in leaderless systems)

---

distributed-systems replication