The CAP Theorem

One-Liner

A distributed system can only provide two of three guarantees: Consistency, Availability, and Partition Tolerance.

What It Is

A fundamental theorem in distributed systems that states it is impossible for a distributed data store to simultaneously provide more than two of the following three guarantees: Consistency, Availability, and Partition Tolerance.

Why It Exists

To force system designers to consciously choose which guarantees to prioritize when facing network partitions, which are inevitable in distributed systems.

How It Works

Consistency (C): Every read receives the most recent write or an error.
Availability (A): Every request receives a (non-error) response, without the guarantee that it contains the most recent write.
Partition Tolerance (P): The system continues to operate despite an arbitrary number of messages being dropped (or delayed) by the network between nodes.

In a distributed system, you must choose P, which means you are forced to trade off between C and A.

Tradeoffs

Pros

CP (Consistency/Partition Tolerance): The system remains consistent even during a partition.
AP (Availability/Partition Tolerance): The system remains available during a partition.

Cons

CP (Consistency/Partition Tolerance): May become unavailable to some clients.
AP (Availability/Partition Tolerance): Some clients may see stale data.

Failure Modes

Split-brain in AP systems: Nodes in different partitions can accept conflicting writes.
Cascading unavailability in CP systems: A partition can cause large parts of the system to become read-only or completely unavailable.

Interview Traps

Forgetting that in practice, the choice is always between C and A, because P (partitions) are a fact of life in distributed systems.
Treating the CAP theorem as a binary choice, rather than a spectrum of tradeoffs.

Anti-Patterns

Claiming a system is “CAP” (all three).
Ignoring the theorem and not having a clear strategy for handling partitions.

Consistency Models
Distributed Consensus