Hi Alex, Thanks for the answer. Yes, you are right, for the software dispather, when scaling from one thread to two threads, it shall have packet disorder issue.
B.T.W. how did you test the short-lived connections? Regards, CY. On Tue, Nov 26, 2013 at 1:34 AM, Alex Wang <al...@nicira.com> wrote: > Hey, > > Want to try answering your question, > > > 1. If your hardware uses hyperthreading, I recommend do not specify > 'n-handler-threads' more than # of real cores. > > > >> For the the upcall dispatcher, currently it's using hash function to find >> a miss-handler, it's some kind of load balance selection for all the >> threads. Will it be possible to select one miss-handler always, and only >> the miss-handler's load/cpu usage is above a threshold, then select the >> second one, and then the third one, ...? >> > > 2. One thing we are concerned about is maintaining packet processing > order. Your suggestion looks hard to achieve that to me. >
_______________________________________________ discuss mailing list discuss@openvswitch.org http://openvswitch.org/mailman/listinfo/discuss
