reiabreu commented on issue #8583: URL: https://github.com/apache/storm/issues/8583#issuecomment-4459058994
I mostly agree with @rzo1 analysi(thanks for the very detailed write up to both). Specially around the reordering tuples. Fields grouping uses key mod numTasks, so we wouldn't be breaking this, but we would be changing the implicit FIFO. What what would happen if we decide to hold back a tuple in an executor due to the state of the downstream task only for the next tuple that would go to the same task find a more favorable downstream jitter and be emitted. We would be changing the order of processing and breaking the sequence guarantees, despite assuring co-location. The proposal is very interesting, jitter metrics are worth exporting, but the self regulating loop will be tricky as proposed -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
