Github user Ethanlm commented on the issue: https://github.com/apache/storm/pull/2270 @HeartSaVioR Thanks for the review. I think the easiest could be using ping times between supervisors. That could be done in a fair time frame. A more complicated scoring will need to take many statistics into consideration, for example, the transferring latency between tasks, serialization/deserialization cost, current load in downstream executors, and the execution latency of executors and etc. It's also interesting to see if we could apply any machine learning algorithms. But it should take a lot of time to do it. We don't have any plan on this currently.
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---