Definition: Distributed consistency patterns are techniques for keeping data correct and coordinated when it is spread across multiple services, databases, or servers.
Because you usually cannot use one normal database transaction across all of them, patterns such as Saga, Outbox, replication, and locking are used instead.
Example
Imagine an online shop:
Order Service
↓
Inventory Service
↓
Payment Service
↓
Shipping Service
A single business action — “buy a product” — affects several services.
The question is:
What happens if Payment succeeds but Shipping fails?
Distributed consistency patterns are the different strategies for dealing with situations like this.
Important: There isn’t one single “distributed consistency pattern.” It’s a category of solutions.