Instance Status and Alerts¶
Observability provides the progress and status of each instance. The system shows an instance as:
- Active when in progress and has no alerts.
- Completed when ended with no alerts.
- Warning when in progress or ended with warning-level alerts.
- Error when in progress or ended with error-level alerts.
- Ended when the instance is no longer active. The instance may or may not have alerts.
- Upcoming when the instance will occur in the future, based on the journey's instance conditions.
Instance alerts¶
Alerts give additional context about what happens to components during an instance. Instance alerts consider:
- Run statuses.
- Any component expectations you set.
- The journey sequence.
From the Instances page or above the instance details tabs, click any or to open a dialog with information about the alerts.
| Alert level | Example situations |
|---|---|
| Error | A batch pipeline run failed. A batch pipeline run was missed (did not occur per its expected schedule). A batch pipeline was still running when a subsequent run in the journey started. The instance has ended, but a batch pipeline run is pending. The instance has ended, but a batch pipeline run is still running. The instance has ended, but not all batch pipelines in the journey had a run. The order of batch pipeline runs did not align with the journey's sequence. Tests in a component failed. |
| Warning | A batch pipeline run started later than its expected schedule. A batch pipeline run ended later than its expected schedule. A batch pipeline run completed, but with warnings. A batch pipeline run status changed from its end state. A dataset component asset did not arrive when expected. Tests in a component had warnings. |
Tip
Set rules based on instance alerts, so you can get notified about key situations in your data estate.
Troubleshooting: Pending or Missing Runs