Quorum
A quorum is the minimum number or proportion of distributed participants whose responses or agreement are sufficient for an operation or decision.
Kafka for Architects
A quorum architecture commits replicated state only after a majority of participating voters acknowledge the update.
- How it is used
- Coordinates fault-tolerant Kafka metadata updates and controller leadership.
- How deep it goes
- Explains majority commitment, leader election, failure sizing, and placement without a low-level Raft implementation walkthrough.
- Examples
- Illustrates metadata-update replication and compares controller placement across two and three data centers.
- What you can do after
- Evaluate/understand controller counts and placement needed to retain metadata-management availability after failures.
ScyllaDB in Action
The QUORUM consistency level requires a majority of replicas for a key to acknowledge an operation.
- How it is used
- Balances consistency and availability for ScyllaDB reads and writes by requiring majority replica acknowledgments.
- How deep it goes
- Explains majority acknowledgments, coordinator-to-replica flow, and a node-failure query experiment without detailing later read and write internals.
- Examples
- Demonstrates a three-node restaurant-review query succeeding at QUORUM after one node stops, while ALL returns NoHostAvailable.
- What you can do after
- Operate/configure it by setting QUORUM for a query and predicting its behavior when one of three replicas is unavailable.
Elasticsearch in Action, Second Edition
A quorum is the minimum number of eligible participants whose acknowledgments are required to authorize an operation in a distributed system.
- How it is used
- Uses master-eligible nodes to authorize cluster-state commits and master elections.
- How deep it goes
- Explains quorum membership, acknowledgment-based commits, the majority formula, and split-brain mitigation.
- Examples
- Demonstrates quorum with a 20-node cluster containing eight master-eligible nodes and a five-node quorum.
- What you can do after
- Evaluate cluster layouts by calculating quorum size and recognizing why three master-eligible nodes are recommended.
Read how Elasticsearch in Action, Second Edition explains quorum →
Latency
A quorum is a majority subset of nodes whose agreement is sufficient for a distributed operation to complete.
- How it is used
- Models distributed-operation latency when parallel components require only majority completion.
- How deep it goes
- Defines the majority formula and latency consequence without detailing consensus protocol internals.
- Examples
- Uses three-node and five-node clusters plus Raft, Paxos, and Viewstamped Replication.
- What you can do after
- Apply it via Raft, Paxos, or Viewstamped Replication to reason about majority completion and latency.
Think Distributed Systems
A quorum is a sufficient subset of distributed participants whose agreement or participation allows a decision to be made safely.
- How it is used
- Coordinates consensus, replicated-log commitment, and replication acknowledgments across distributed processes.
- How deep it goes
- Defines majority size, intersection, failure capacity, and leader integration without detailing protocol implementation.
- Examples
- Uses three-process quorum sets, majority replica acknowledgments, and Raft log commitment as concrete examples.
- What you can do after
- Evaluate/understand quorum requirements for consensus, replication, failure tolerance, and split-brain prevention.
Related concepts
- consensusused by · Consensus is sufficient agreement among participants or processes on a shared decision, value, state, assessment, or outcome.
- master-eligible nodehas part · A master-eligible node is an Elasticsearch node assigned the master role and therefore able to vote in master elections and become the elected master.
- consistency levelkind of · A selectable database setting that determines how many replicas or nodes must participate in an operation and what consistency it provides.
- data lossaddresses · Data loss is the destruction, corruption, deletion, or unavailability of information that is expected to be retained or remain usable.
- leader electionused by · Leader election is a distributed coordination process that selects one eligible participant to act as leader or coordinator.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.