What causes CANCELED?

A user or automated process manually canceled the run via the Databricks UI or Jobs API. A job is configured with 'Cancel previous run' concurrency policy — a new run canceled the still-running previous run. An orchestration tool (e.g., ADF, Airflow) canceled the run as part of a timeout or error handling flow. A long-running job was canceled before a scheduled maintenance window. A CI/CD pipeline canceled the run during a redeployment

How do I fix CANCELED?

Check the Databricks audit log to see which user or service principal triggered the cancellation. Review the job concurrency policy — if 'Cancel previous run' is set, consider whether the policy matches your intent. If canceled by an orchestration tool, check the orchestrator logs for the reason. Re-trigger the job if the cancellation was unintentional. Adjust the job timeout or concurrency policy if runs are being canceled on a regular schedule

Medium severityjob

Power BI Refresh Error:
CANCELED

What does this error mean?

A Databricks job run was manually canceled, or canceled by the system due to a concurrent run limit or cancellation policy. No error occurred in the job code — the run was stopped externally.

Common causes

1A user or automated process manually canceled the run via the Databricks UI or Jobs API
2A job is configured with 'Cancel previous run' concurrency policy — a new run canceled the still-running previous run
3An orchestration tool (e.g., ADF, Airflow) canceled the run as part of a timeout or error handling flow
4A long-running job was canceled before a scheduled maintenance window
5A CI/CD pipeline canceled the run during a redeployment

How to fix it

1Check the Databricks audit log to see which user or service principal triggered the cancellation
2Review the job concurrency policy — if 'Cancel previous run' is set, consider whether the policy matches your intent
3If canceled by an orchestration tool, check the orchestrator logs for the reason
4Re-trigger the job if the cancellation was unintentional
5Adjust the job timeout or concurrency policy if runs are being canceled on a regular schedule

Frequently asked questions

Does a CANCELED run count as a failure?

In Databricks it is a distinct terminal state from FAILED. However, if a CANCELED run means important data was not processed, it has the same operational impact as a failure.

What is the 'Cancel previous run' concurrency policy?

When a new run starts and the previous run is still in progress, Databricks cancels the old run so only the latest run executes. Useful for idempotent jobs but dangerous if the old run was doing critical work.