Search before asking

• I had searched in the issues and found no similar feature requirement.

Description

@EricGao888 has released a plugin earlier to implement batch job submission on Aliyun Serverless Spark. After a careful review, i think there are still something to be optimized. as listed below:

1. retry for remote api call is missing now, which may cause inconsistence state when some temporary issue occur like network failure
2. idempotent submitting is necessary especially for the same scheduled node in the same process instance
3. termination should be ensured cuz a single call on CancelJobRun Api can only turn a remote job into cancelling state, but not really canceled

Are you willing to submit a PR?

• Yes I am willing to submit a PR!

Code of Conduct

• I agree to follow this project's Code of Conduct