Monitor runs
Where: sidebar → Job running. The page heading is Run executions.
Each execution records an attempt to run a saved pipeline or script version. Opening the job editor does not tell you whether a previous execution succeeded.
Find the execution
Locate the run associated with your pipeline or use the execution identifier returned when you started it. Open its detail panel.
Inspect status and context
Check the workload, saved version, timestamps, and any Spark application identifier or resource profile. Confirm you are looking at the correct attempt.
Read the result and logs
Open Logs and inspect the error summary if the run failed. After a successful run, also inspect the target table; a status alone does not validate business correctness.
Read execution status
| Status | Meaning for the user |
|---|---|
| Queued / Preparing | Work has been accepted or is being prepared; no successful output is established yet. |
| Submitted | The workload has been submitted to the execution runtime. |
| Running | Execution is in progress. |
| Cancelling | Cancellation has been requested but is not complete. |
| Succeeded | The execution completed successfully; validate its output. |
| Failed / Timed out | Inspect the failure details and logs before retrying. |
| Cancelled | The execution ended as cancelled; check whether any output was already written. |
Cancel or recover
Select Cancel when offered and wait for a final status. If live updates disconnect, reopen the execution or refresh before deciding that the job has stopped.
Before a retry, check the source connection, graph validation, target permissions, write mode, and any runtime error. Do not interpret a frontend timeout as proof that the backend execution did not start.
For support, provide the execution ID, pipeline name/version, timestamp, and a sanitized error excerpt. Never share logs containing credentials or sensitive rows without review.