DaanHoogland commented on PR #8694: URL: https://github.com/apache/cloudstack/pull/8694#issuecomment-1962610393
> A couple questions, if you don't mind: > > 1. Would it make sense from a latency-perspective to only wait 1ms if there's actually no data coming in? If there's a big chunk of data coming in, you might not want to wait 1ms to read ~200 bytes (which appears to be the maximum the NIO readNextBytes method would return) > > 2. Why 1ms? Could we raise this to reduce wasted CPU cycles even more? My testing showed the CPU went from 100% to 25%. This might still seem a bit excessive. valid questions @kohrar , sounds like you are arguing for configurability. This is the simplest solution to the issue in your initial ticket. What you are asking now is a bigger (functional) change however. I would prefer to discuss and implement that separately. -- 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]
