On Wed, Aug 07, 2013 at 12:57:55PM -0400, Maxim Khitrov wrote: > On Wed, Aug 7, 2013 at 11:44 AM, Florian Obser <flor...@narrans.de> wrote: > > On Wed, Aug 07, 2013 at 10:26:22AM -0400, Maxim Khitrov wrote: > > [...] > >> Increasing the MTU on both ix0 interfaces to 9000 gives me ~7.2 Gbps: > > > > you expect a lot of jumbo frames in front of / behind your firewall? > > (if the answer is no, why are you testing that?) > > It's a possibility. What this tells me, however, is that the the > throughput isn't the (main) problem. The per-packet processing > overhead appears to be the limiting factor, which is why I asked about
indeed, during my tests systat showed that the system is spending 99% in interrupt handlers. Having context switches because you are running iperf localy is not good[tm] in this situation. > checksum offloading. > > > anyway, I was testing an Intel 82599 system in July which will become > > a border router. All of this is forwarding rate; it took me 2 days to > > beg, borrow and steal enough hw to actually generate the traffic. (I > > had 4 systems in front of and 4 systems behind the router, all doing > > 1Gb/s) > > What tools were you using to generate the traffic and to calculate > bytes/packets per second? I assume interrupts per second came from > systat? > right, the interrupt rate came from systat, traffic was generated with iperf and measured with bwm-mg in 30 second average mode. iperf was running in dualtest mode and instructed to run for an hour so that I had a chance to start all iperfs before the first one would finish ;) no other switches (besides -c and -s of course). -- I'm not entirely sure you are real.
