This post has been republished via RSS; it originally appeared at: New blog articles in Microsoft Tech Community.
[Following material applies to Tumbling Window Triggers in Azure Data Factory. Learn more about Tumbling Window Trigger]
With the newly released feature, you now have the capabilities to cancel Tumbling Window Trigger runs if the window is in Running or Waiting on Dependency state.
- If the window is in Running state, cancel the associated Pipeline Run, and the trigger run will be marked as Cancelled afterwards
- If the window is in Waiting or Waiting on Dependency state, you can cancel the window from Monitoring
You can also rerun a cancelled window. The rerun will take the latest published definition of the trigger; and dependencies for the specified window will be re-evaluated upon rerun.
Combining these two capabilities give you a powerful way to resolve the hanging/blocking dependencies view. Imagine the following scenario: TWT-B depends on TWT-A, and for some reasons TWT-A has been consistently failing lately. To unblock downstream work flows, you decided to remove TWT-A from TWT-B's dependency list.
Instead of stopping and restarting TWT-B, you can now take following steps:
- Cancel all Waiting on Dependency runs of TWT-B (these are the windows waiting for TWT-A to finish)
- Edit and publish new trigger definitions of TWT-B to remove dependencies
- Rerun the cancelled runs
Since the rerun will take the new definition, it will by-pass the blocking TWT-A, and enable downstream workflow to continue