What causes RPCKilledException?

A team member cancelled the job manually in the dbt Cloud UI to stop a long-running or incorrectly triggered run. A CI/CD pipeline cancelled the run via the dbt Cloud API when a newer commit superseded the current run. A dbt Cloud job orchestrator (e.g., Airflow, Prefect) cancelled the run because a dependency failed. The dbt Cloud account hit a concurrent run limit and the scheduler killed the oldest run to start a queued one. An automated cleanup script cancelled stale or stuck runs

How do I fix RPCKilledException?

Check the dbt Cloud run history to see who or what triggered the cancellation — the audit log shows the user or API token that called the cancel endpoint.. If the cancellation was unintentional, re-trigger the job manually from the dbt Cloud UI.. If a CI/CD pipeline is cancelling runs, review the pipeline logic to ensure it only cancels runs when truly superseded.. For orchestrator-triggered cancellations, check the orchestrator logs to identify which upstream dependency failed and why.. If concurrent run limits are the cause, review the dbt Cloud account plan and job scheduling to reduce overlap between jobs.

Low severityoperational

Power BI Refresh Error:
RPCKilledException

What does this error mean?

A dbt Cloud run was manually cancelled by a user or programmatically via the dbt Cloud API. All models that had not yet completed are marked as cancelled with RPCKilledException. This is not a failure caused by bad code — it is an intentional termination of the run.

Common causes

1A team member cancelled the job manually in the dbt Cloud UI to stop a long-running or incorrectly triggered run
2A CI/CD pipeline cancelled the run via the dbt Cloud API when a newer commit superseded the current run
3A dbt Cloud job orchestrator (e.g., Airflow, Prefect) cancelled the run because a dependency failed
4The dbt Cloud account hit a concurrent run limit and the scheduler killed the oldest run to start a queued one
5An automated cleanup script cancelled stale or stuck runs

How to fix it

1Check the dbt Cloud run history to see who or what triggered the cancellation — the audit log shows the user or API token that called the cancel endpoint.
2If the cancellation was unintentional, re-trigger the job manually from the dbt Cloud UI.
3If a CI/CD pipeline is cancelling runs, review the pipeline logic to ensure it only cancels runs when truly superseded.
4For orchestrator-triggered cancellations, check the orchestrator logs to identify which upstream dependency failed and why.
5If concurrent run limits are the cause, review the dbt Cloud account plan and job scheduling to reduce overlap between jobs.

Frequently asked questions

Does RPCKilledException mean my dbt code has a bug?

No — the code was never the problem. The run was cancelled externally. Check the dbt Cloud audit log to identify the cancel trigger.

Will a cancelled run leave partial data in the warehouse?

Models that completed before the cancellation wrote their results. Models in progress at cancellation time depend on the adapter's transaction behaviour — some warehouses roll back incomplete transactions, others leave partial writes.