Glossary · System coordination, integration and orchestration
Eventual consistency
Also known as: eventually consistent
German: Eventuelle Konsistenz
In distributed systems, eventual consistency is a consistency model in which replicas of data may temporarily differ after an update, but converge to the same value if no new updates are made for long enough.
- System integration
In one sentence
Under eventual consistency, replicas may differ briefly after an update but converge to the same value once updates stop.
Example
A new tool offset appears on the central dashboard a few seconds after it was changed on the machine, because the dashboard's copy is eventually consistent.
How it applies
- Architecture: Eventual consistency lets systems stay available and fast across unreliable networks, which suits dashboards, analytics and many enterprise integrations.
- Engineering: It is unsuitable where decisions depend on the latest value, such as recipe release or setpoint changes. Consumers of eventually consistent data should check timestamps or versions and treat Stale data explicitly.
- Documentation: State the expected convergence time and the conditions under which data may be outdated. User documentation should warn when displayed values may lag behind the machine.
Eventual consistency vs. strong consistency
With strong consistency, every read returns the latest confirmed write. With eventual consistency, a read can return an older value for a while. The German list term “eventuelle Konsistenz” is a loan translation; “eventual” means “in the end,” not “possibly.”