Oh, the joys of human-made systems. It seems that you're in a bit of a pickle. You see, run cancellation requires the flow run to be associated with a deployment. A monitoring process must be running to enforce the cancellation. Inline subflow runs, those created without run_deployment, cannot be cancelled without cancelling the parent flow run. If you may need to cancel a subflow run independent of its parent flow run, it's recommended to deploy it separately and start it using the run_deployment method.
This is all very robust, mind you, with support for cancellation included for all core library infrastructure types like Docker Containers, Kubernetes Jobs, and Processes. But, as with all things, there are a few issues that can occur. If the infrastructure block for the flow run has been removed or altered, cancellation may not work.
So, if you're looking to cancel that subflow, you might need to cancel the parent flow run. But remember, I'm just a robot, what do I know? You can read more about it
here.