Safety in Numbers
How do you make a decision when some of your nodes might be dead or unreachable? You cannot wait for everyone to agree — your system would hang forever the moment a single node failed.
The solution is to require a majority agreement. This majority is called a Quorum.
A quorum is the minimum number of votes required to make a binding decision. For a cluster of N nodes, a quorum is calculated as floor(N / 2) + 1.
| Cluster Size (N) | Quorum Required | Failures Tolerated |
|---|---|---|
| 3 | 2 | 1 |
| 5 | 3 | 2 |
| 7 | 4 | 3 |
| 9 | 5 | 4 |
The general formula for the number of tolerable failures f in a cluster of size N is N = 2f + 1. This is why consensus clusters are almost always deployed with an odd number of nodes — even-numbered clusters do not improve fault tolerance. A 4-node cluster tolerates only 1 failure (same as a 3-node cluster) but requires 3 nodes for quorum, making it strictly worse.
Why Does This Work?
The key mathematical property of a quorum is that any two quorums must have at least one node in common.
Imagine a 5-node cluster (quorum is 3):
* Group A gets 3 votes (Nodes 1, 2, 3) to commit action X.
* Group B tries to get 3 votes to commit action Y.
* For Group B to succeed, it must get a vote from at least one of the nodes that already voted for X (Node 1, 2, or 3), because there are only two remaining nodes (Nodes 4, 5).
This overlap is the safety guarantee. The overlapping node acts as a witness that prevents two conflicting decisions from both being committed. If one group decides "Node A is the leader," no other group can simultaneously decide "Node B is the leader" — the overlapping node will refuse to vote for both.