CAP Theorem

The CAP Theorem, often referred to as Brewer's Theorem, is a foundational concept in distributed systems. It states that a distributed data store can provide at most two of the following three guarantees simultaneously:

The Three Pillars of CAP

• Consistency (C): Every read request receives the most recent write or an error. Essentially, all nodes in the system see the same data at the same time.
• Availability (A): Every request made to a non-failing node in the system receives a response (even if it is not the most recent data).
• Partition Tolerance (P): The system continues to operate despite arbitrary network failures (partitions) that prevent nodes from communicating with each other.

The Core Trade-off

In a real-world distributed system, network partitions (P) are inevitable because network failures can and will happen. Because you cannot avoid network failures, the theorem effectively forces architects to choose how the system behaves during a partition:

• CP (Consistency + Partition Tolerance): The system prioritizes data accuracy. If a partition occurs, the system will return an error or time out rather than risk serving inconsistent (stale) data.
• AP (Availability + Partition Tolerance): The system prioritizes staying online. If a partition occurs, all nodes will continue to respond to requests, but they might return older, potentially stale data because they cannot synchronize with the rest of the cluster.

Summary Table

Combination	Priority	Behavior During Partition
CP	Consistency	System shuts down non-consistent nodes/returns errors.
AP	Availability	System returns the most recent local data, even if it is stale.

Key Takeaway for System Design: The CAP theorem is not about choosing two out of three forever. It is a framework for deciding how your system should handle network failures when they occur. Most modern distributed systems are designed to be "partition-tolerant" by default, leaving the meaningful choice between consistency and availability.