[GitHub] [beam] scwhittle commented on a diff in pull request #23882: Enforce splitting invariants by ensuring split state is reset in the same synchronized block as window index increments.

[GitBox <gi...@apache.org>]

scwhittle commented on code in PR #23882:
URL: https://github.com/apache/beam/pull/23882#discussion_r1011466147


##########
sdks/java/harness/src/main/java/org/apache/beam/fn/harness/FnApiDoFnRunner.java:
##########
@@ -999,14 +1022,7 @@ public void onClaimFailed(PositionT position) {}
         }
       }
     } finally {

Review Comment:
   Is that guaranteed somewhere? I'm concerned that if things change so that the exception is caught that we've left this object in an invalid state.  We had a recent bug in state where the user was able to catch the exception and the invalid state was then used. I don't think that is possible here but is it possible that we add some generic exception handling (perhaps by configuration) and would rather clean up as we go here? Or is that unlikely in your opinion?
   
   



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@beam.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org