In 2000, Eric Brewer presented the CAP theorem to the world in his keynote speech at the ACM Symposium on the Principles of Distributed Computing. The theorem states that in the presence of a network partition, the system design can either provide availability or consistency. We shall dig deeper into the CAP theorem in Chapter 5, Going Distributed, but the immediate implications for storage systems are that for a distributed system providing ACID compliance is tough to engineer and does not scale well.

To overcome this, some modern storage systems offer an alternative model of consistency, called BASE—short for Basically Available, Soft state, Eventual consistency. Let's look at what this means:

• Basically Available: This system guarantees the availability that there will be a response to a request, but the response might not be consistent at all times.
• Soft state: The state of the system could change over time, even when there is no external input.
• Eventual consistency: The system will eventually become consistent. Generally, this happens when various concurrent operations are reconciled and a steady state is reached.