Hi Ben, Yes, you are right. I find the bottleneck is tcprelay which only generates packets at 750Kpps. But even when I add more hosts to send packets in parallel, I can't get a higher generation rate. E.g., if I use only one host, tcpreplay sends packets at 300Kpps, while if I use 6 hosts, each tcpreplay sends packets at around 130Kpps and the total rate is still 750Kpps. How do you get a higher generation rate in your experiment? Thanks.
+++++++++++++++++ Best, Hao On Thu, Jul 16, 2015 at 12:28 AM, Ben Pfaff <[email protected]> wrote: > On Wed, Jul 15, 2015 at 03:10:33PM -0700, Hao Wu wrote: > > I'm using OVS 2.3.0 and testing the packet processing throughput. > > Currently, I build a simple network via Mninet which consists of 3 hosts > > connecting to 1 ovs instance. Then, I use the 3 hosts send out 64-byte > > packets via Tcpreplay, and I measure the packet processing throughput in > > OVS and get the result at 750Kpps. > > > > I want to know is there any method to enhance OVS performance? like > > binding multi cores or some other settings? Since I notice the > performance > > in NSDI 15 paper is several Mpps. > > Have you determined for sure that OVS is the bottleneck in the setup? >
_______________________________________________ discuss mailing list [email protected] http://openvswitch.org/mailman/listinfo/discuss
