Cassandra在CAP定理中的定位：AP还是可配置为CP？

Great question—this is a super common point of confusion because Cassandra's behavior shifts based on how you configure it! Let me break it down clearly:

Default/Native Classification: AP #

Datastax is totally right to categorize Cassandra as AP (Availability + Partition Tolerance) out of the box. Here's why:

• Cassandra's core design prioritizes keeping the system available even when network partitions happen (a given in distributed systems). It’ll keep accepting writes and reads as long as a quorum of nodes is reachable.
• By default, consistency levels are set to enable eventual consistency, so you trade strict immediate consistency for higher availability and partition resilience.

Configurable Shift to CP Behavior #

You’re also correct that Cassandra can be adjusted to behave like a CP (Consistency + Partition Tolerance) system using the W + R > RF rule:

• First, let’s define the terms simply:
  • W: Write consistency level (number of replicas that must confirm a write before it’s marked successful)
  • R: Read consistency level (number of replicas that must respond to a read request)
  • RF: Replication factor (total number of replicas storing your data)
• When W + R > RF, it mathematically guarantees that any read will pull in the latest successful write. For example, if RF=3, setting W=2 and R=2 gives 2+2>3—this ensures the set of nodes that confirmed the write overlaps with the set of nodes you’re reading from, so you’ll always get the most up-to-date data.
• In this setup, if a network partition hits and you can’t reach enough replicas to meet the W or R requirement, the system will prioritize consistency over availability—it’ll reject writes/reads rather than returning inconsistent data.

So to wrap it up: Cassandra is natively AP, but its flexible consistency model lets you tune it to operate in a CP-like way when strict consistency matters more than 100% uptime for every request.

内容的提问来源于stack exchange，提问作者quangh