Github user StephanEwen commented on the issue:
https://github.com/apache/flink/pull/5427
I think we may get around adding a new method. I checked with @tillrohrmann
, here are the thoughts:
- Submitting a job initially as a new reader group, no need to reset here
- Recovering at any point calls the `restoreLatestCheckpointedState()`
method
- Also recovering from a JobManager failover basically "resubmits" the
job with a special flag causing the JM to call the
`restoreLatestCheckpointedState()` for the job.
---
