Github user arunmahadevan commented on the issue: https://github.com/apache/spark/pull/21673 @tdas , thanks for your comments. Yes theres problem with the current abstraction, and I didn't consider refactoring it since there have been multiple changes to this class without changing the underlying structure and the fields of the ExecutionStats are accessed from multiple places within StreamExecution already. I did not think adding an extra field would increase the code complexity, however if you plan to do major refactoring to simplify the logic and address the issues, I am happy to discard this PR and help review your changes.
--- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org