This chapter describes the functions used to manage durability and consistency policies. Durability policies are used with write operations to manage how likely your data writes are to persist in the event of a catastrophic failure, be it in your hardware or software layers. By default, your writes are highly durable. So managing durability policies is mostly about relaxing your durability guarantees in an effort to improve your write throughput.

Consistency policies are used with read operations to describe how likely it is that the data on your replicas will be identical to, or consistent with, the data on your master server. The most stringent consistency policy requires that the read operation be performed on the master server. In general, the stricter your consistency policy, the slower your store's read throughput.