What happened (please include outputs or screenshots):
After onstarted_leading function throw exception and finish, the process still leading.

What you expected to happen:
The process stop leading immediatly.

How to reproduce it (as minimally and precisely as possible):
Just follow the leaderelection/example.py.

Anything else we need to know?:
Go client use ctx.Done() to find out the OnStartedLeading finish, and then release lock:
https://github.com/kubernetes/client-go/blob/1309f64d6648411b4a36a2f7fa84dd8df31884b6/tools/leaderelection/leaderelection.go#L265-L291

Environment:

• Kubernetes version (kubectl version): v1.20.15
• OS (e.g., MacOS 10.13.6): 10.15.7
• Python version (python --version): 3.8.10
• Python client version (pip list | grep kubernetes): 21.7.0