Deleted DataFactory Trigger still executing daily for months after being disabled and deleted

Use the management APIs and monitoring views to locate and stop the hidden trigger and its pipeline, then clean up the artifacts explicitly.

1. Confirm which factory and pipeline are actually running
   • In Monitor → Pipeline runs, note the Data Factory name, pipeline name, and run IDs that are still executing twice daily.
   • Optionally, use PowerShell to query the run details:

          Get-AzDataFactoryV2PipelineRun -ResourceGroupName <rg-name> `
            -DataFactoryName <factory-name> `
            -PipelineRunId <run-id>
     

     This confirms the pipeline and factory that are still active.
2. Check for triggers via API/PowerShell (even if Studio shows none)
   • Triggers have states started or stopped, and deployment/publishing does not delete them automatically. They must be explicitly deleted via API.
   • Use PowerShell or REST to list triggers on the factory and find the one still associated with the pipeline. If any are in started state, stop them first, then delete them using the Create or Update Trigger API and the trigger delete API.
   • Important behavior: deployment won’t delete artifacts; delete APIs must be called for each artifact type (pipeline, dataset, linked service, trigger) to fully clean up a factory.
3. Stop and delete the Integration Runtime
   • In Monitor → Integration runtimes, confirm the Azure-SSIS IR that is still running.
   • If Studio won’t let it be stopped, use PowerShell/REST against the factory to:
     1. Stop the IR.
     2. Delete the IR definition so it cannot be started again by any lingering pipeline or trigger.
   • If there are capacity/queue issues or “no active worker”/unhealthy status errors, follow the IR monitoring guidance under Azure-SSIS integration runtime and the Integration Runtime Management Troubleshooting Guide.
4. Clean up orphaned artifacts in the factory
   • Because deployment does not delete artifacts, there can be “ghost” pipelines or triggers that still exist in the backend even if Studio shows them as deleted.
   • Use the REST delete APIs to explicitly remove:
     • Pipelines: pipeline delete API
     • Datasets: dataset delete API
     • Linked services: linked service delete API
   • This ensures no hidden artifact can continue to schedule or run the SSIS package.
5. Ongoing monitoring and verification
   • After stopping/deleting the trigger and IR, verify in Monitor → Pipeline runs that no new runs are created at the old schedule.
   • Optionally, set up periodic monitoring (for example, with a Logic App) using the pipelineruns/querybyfactory endpoint to detect any unexpected runs and alert a ticketing system, as described in Query By Factory.

If the factory is in a production environment with many live triggers, ensure only the specific offending trigger and pipeline are stopped/deleted so as not to impact other workloads.

References:

• Start and stop an Azure-SSIS integration runtime on a schedule
• Troubleshoot pipeline orchestration and triggers in Azure Data Factory
• Applying DataOps to Azure Data Factory
• Troubleshoot package execution in the SSIS integration runtime