Skip to main content
Anomalies are detected data issues that surface when something prevents Twine from processing or replicating an entity correctly. They are scoped to an entity and an anomaly code, and may optionally be tied to a System Integration and/or Domain Mapping.

​
Sources

Anomalies come from two places:
  • Built-in checks that run when sending data to a target. If a required value is missing or otherwise invalid for the destination, the pipeline halts for the affected entity and the failure is also recorded against the Job that triggered the check.
  • Data Engine anomaly nodes that customers can add to their graphs to flag custom conditions. These emit an anomaly but do not alter the flow or halt processing. A separate halt node would have to be added for that.

​
Lifecycle

Some anomalies clear automatically when the underlying issue resolves - for example, when the missing value is finally populated in the source system and a subsequent sync brings it into Twine. Others have to be cleared manually in the backoffice. Anomalies can be configured to be delivered as digests via the Notifications feature, allowing the right people to be informed without monitoring the backoffice continuously. For instructions on viewing and clearing anomalies in the UI, see Working with Anomalies.
This page is a stub. More detail on anomaly codes and common causes will be added.