Github user StephanEwen commented on the issue:

    https://github.com/apache/flink/pull/5427
  
    Given the already non-trivial complexity of the `CheckpointCoordinator`, I 
am wondering if there is a way to do this without adding the 
`resetForNewExecution()` method.
    
    Adding another life cycle method (even if it is not exploited for other 
purposes currently) would need more involved tests and make future maintenance 
harder.
    
    Can we reset all hooks in the `restoreLatestCheckpointedState()` method? 
Would we miss any cases if we do only that? 


---

Reply via email to