Github user HeartSaVioR commented on the issue:
https://github.com/apache/storm/pull/2723
@danny0405
Could you address @revans2 comment, mostly backing up your proposed patch
with numbers? You may want to get help to correctly measure numbers from
well-known library like JMH.
If you're not familiar with JMH, this blog post would help you:
https://blog.codecentric.de/en/2017/10/performance-measurement-with-jmh-java-microbenchmark-harness/
I'd say showing the numbers is more powerful for persuading than let others
dive to the code and find value, especially the patch is not addressing Storm's
performance issues what we already know about. STORM-2693 addressed our
long-lived issue, so that was less needed to persuade others but this doesn't
look like the case.
---
