pnowojski commented on pull request #18475:
URL: https://github.com/apache/flink/pull/18475#issuecomment-1024637122


   > I would say without volatile notifyCompletion see a obsolete future could 
happen. But, before entry the notifyCompletion, the input's future is already 
completed. In the anyOf function, it will recheck whether the future of the 
input is completed, and thus triggered the following 
availableFuture.complete(). Stream will not halt.
   
   No. In the scenario I described above:
   
   >  If Complete action happens at the very very end (around to here arrow), 
   
   when `anyOf` is being executed the input's future will be still not yet 
completed.
   
   Please do not add test coverage for that, it would be unreliable anyway. 
   
   I don't know why, I'm not able to reproduce this issue locally. I believe it 
can happen, but I simply can not verify it. Anyway, as I wrote in the ticket, I 
had a chat with other committers and we decided to merge the fix for this issue 
without test coverage for the time being. We will add later some test coverage 
as part of https://issues.apache.org/jira/browse/FLINK-25869. I would propose 
to go with my fix, as it's a bit simpler and later we need to probably re-use 
the same code to fix https://issues.apache.org/jira/browse/FLINK-25827.


-- 
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: issues-unsubscr...@flink.apache.org

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


Reply via email to