Data Consistency Paradigms

The two primary consistency paradigms to be aware of are referred to as ACID and BASE. We’ll define these acronyms in the following sections and understand what they are and how they differ. To help remember these acronyms, we can use an analogy from chemistry, where a BASE is the opposite of an ACID.

ACID paradigm/immediate consistency

As previously noted, ACID stands for atomic, consistent, isolated, and durable.

With immediate consistency, we get a guarantee that each transaction will be completed as a single unit of operation. Therefore, it will either entirely succeed or entirely fail. Take a hypothetical scenario of a consumer ordering a product to a new delivery address. In an immediate consistency model, the typical process for this transaction would be to first issue the address change and wait for positive confirmation that it was validated and written to the customer’s profile. Next, the order would be created, at which point the new address is immediately available for read confirmation on the customer’s profile. This is great for ensuring only accurate data exists at any given point in time. An order containing a customer address that is not on their profile is an error scenario that would fail validation and, therefore, fail to place the order.