How to kill working job that failed to start

[Toshimitsu Takahashi]

At first, thank you for providing useful OSS!

We struggle with jobs that remain working status.
I guess there are jobs failed not to assign a worker of the executor.
They can not be operated at all on Web UI.
We want the way to force them to stop.

[Eisuke Oishi]

1. Access your remain working "Job Instance Details" page ( /definitions/:definition_id/instances/:instance_id )
2. Press Backslash (\) key.
3. The "Cancel" button turn into the "Force Cancel" button.
   
4. Press "Force Cancel" button.

This "Force Cancel" button updates the job instance status to cancel without checking the executor state.
I'm sorry It is undocumented feature.

[Toshimitsu Takahashi]

なんと！ありがとうございます。