Well, I gave that "unintuitive inbound traffic suspension mechanism" in Netty 
3.x a shot in #6313. It doesn't seem so bad, so I hope I implemented it 
correctly! It seems to work in early testing on a real cluster: it exerts 
backpressure on queries to historicals without blocking I/O threads on the 
broker.

[ Full content available at: 
https://github.com/apache/incubator-druid/pull/6014 ]
This message was relayed via gitbox.apache.org for [email protected]

Reply via email to