mundaym commented on pull request #29819: URL: https://github.com/apache/spark/pull/29819#issuecomment-701985693
Thanks for the review @dongjoon-hyun. > Why don't we simply use new value? Then, this PR can be one-liner. Could do. It's seems to me though that having the values the same misleadingly implies that the socket receive buffer should be the same size as the UDP payload, when in fact the socket receive buffer needs to be a lot bigger (multiple packets + kernel overhead). The more I think about though it the more I'm inclined to say we should just leave the size of the socket receive buffer alone (i.e. delete the `socket.setReceiveBufferSize` call completely). The default is always going to be large enough for this test, overriding it seems unnecessary. ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
