> -----Original Message----- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Alexandre Frigon > Sent: Tuesday, January 27, 2015 8:31 PM > To: dev at dpdk.org > Subject: [dpdk-dev] Pktgen-DPDK rate and traffic inconsistency problem > > Hi all, > > I'm using dpdk 1.8 and pktgen-dpdk 2.8 to generate traffic on a back-to-back > setup both equipped with 82599EB 10-Gigabit NIC. > The problem is when I start it, pktgen indicates 10000Mbits/s Tx with 64B > packet > size, but I'm receiving about 15% of it on the other end. > This percentage seems to be proportional with the packet size. > > e.g. > Using nload to read Rx traffic > Pktgen: Tx: 10000Mbits/s ==> Other end: Rx 1660 Mbits/s > Rate: 100% > Pkt size: 64B > > > e.g 2 > Pktgen: Tx: 10000Mbits/s ==> Other end: Rx 9385 Mbits/s > Rate: 100% > Pkt size: 1518B > > > Pktgen is started with this command on a Xeon(R) CPU E31270 @ 3.40GHz > ./app/pktgen -c 1f -n 3 --proc-type auto --socket-mem 1024 --file-prefix pg > -- -p > 0x3 -P -N -m "[1:3].0, [2:4].1"
>From past experience I don't assign more than 1 core per port. It had some race conditions issues and one core I capable to RX or TX full 10G. Also check if you assign proper cores/memory for your NICs (the same NUMA node). > > Is there something I'm not configuring correctly or something I have miss? > > Also, the % rate is acting strangely since anything above 50% doesn't change > the Tx rate and anything below is modifying it > e.g Tx: 10000Mbits/s 5000Mbits/s > %Rate: >=50% 25% > > Actually I am getting exactly opposite results :) If I set rate to 50% I get MBits/s Rx/Tx : 0/9942 9942/0 9942/9942 For 10%: MBits/s Rx/Tx : 0/1997 1997/0 1997/1997 Which is about 2x set :D Additionaly I am getting message when "start 0" -> "stop 0" -> "start 0" is issued PMD: ixgbe_dev_rx_init(): forcing scatter mode So there is definitely something wrong there but don't know where. Another issue I encountered is build system that fail when building out-of-tree. Till this is fixed you can try version 2.7.1 that is working for me. Pawel