On Wed, Nov 13, 2013 at 5:12 PM, Li, Chen <chen...@intel.com> wrote: > Hi list, > > > > I’m working under CentOS 6.4 + Openstack Havana + Openvswitch 1.10.0-1.el6 > > > > While ports have been created like: > > ovs-vsctl show > > d4c41787-8313-49a1-9fa8-1e34f7fa2086 > > Bridge "br-eth4" > > Port "br-eth4" > > Interface "br-eth4" > > type: internal > > Port "phy-br-eth4" > > Interface "phy-br-eth4" > > Port "eth4" > > Interface "eth4" > > Bridge br-int > > Port br-int > > Interface br-int > > type: internal > > Port "qvobc57c061-99" > > tag: 1 > > Interface "qvobc57c061-99" > > Port "int-br-eth4" > > Interface "int-br-eth4" > > ovs_version: "1.10.0" > > > > ip link |grep eth4 > > 6: eth4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP qlen > 1000 > > 54: br-eth4: <BROADCAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN > > 328: phy-br-eth4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1504 qdisc > pfifo_fast state UP qlen 1000 > > 329: int-br-eth4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1504 qdisc > pfifo_fast state UP qlen 1000 > > > > ovs-ofctl dump-flows br-int > > NXST_FLOW reply (xid=0x4): > > cookie=0x0, duration=787.291s, table=0, n_packets=17099182, > n_bytes=409367360073, idle_age=0, priority=3,in_port=1,vlan_tci=0x0000 > actions=mod_vlan_vid:1,NORMAL > > cookie=0x0, duration=908.918s, table=0, n_packets=170, n_bytes=10844, > idle_age=10, priority=2,in_port=1 actions=drop > > cookie=0x0, duration=909.944s, table=0, n_packets=15442515, > n_bytes=1019231483, idle_age=0, priority=1 actions=NORMAL > > > > ovs-ofctl dump-flows br-eth4 > > NXST_FLOW reply (xid=0x4): > > cookie=0x0, duration=800.552s, table=0, n_packets=15975472, > n_bytes=1054407237, idle_age=0, priority=4,in_port=1,dl_vlan=1 > actions=strip_vlan,NORMAL > > cookie=0x0, duration=922s, table=0, n_packets=9, n_bytes=694, idle_age=800, > priority=2,in_port=1 actions=drop > > cookie=0x0, duration=922.937s, table=0, n_packets=17883113, > n_bytes=426539014687, idle_age=0, priority=1 actions=NORMAL > > > > ovs-dpctl show > > system@ovs-system: > > lookups: hit:618023724 missed:5279837 lost:121027 > > flows: 6 > > port 0: ovs-system (internal) > > port 1: br-int (internal) > > port 2: int-br-eth4 > > port 3: phy-br-eth4 > > port 4: qvobc57c061-99 > > port 5: eth4 > > port 8: br-eth4 (internal) > > > > Then I run iperf test, and I get: > > > > ovs-dpctl dump-flows > > in_port(3),eth(src=fa:16:3e:76:88:c0,dst=00:1b:21:a1:19:dc),eth_type(0x8100),vlan(vid=1,pcp=0),encap(eth_type(0x0800),ipv4(src=191.101.0.28,dst=191.101.100.254,proto=6,tos=0,ttl=64,frag=no),tcp(src=5001,dst=33032)), > packets:4728341, bytes:312070578, used:0.000s, flags:., actions:pop_vlan,5 > > package size = 312070578 / 4728341 = 66 > > > > in_port(5),eth(src=00:1b:21:a1:19:dc,dst=fa:16:3e:76:88:c0),eth_type(0x0800),ipv4(src=191.101.100.254,dst=191.101.0.28,proto=6,tos=0,ttl=64,frag=no),tcp(src=33032,dst=5001), > packets:5009744, bytes:120178504800, used:0.000s, flags:P., actions:3 > > package size = 120178504800/ 5009744 = 23989 > > > > in_port(2),eth(src=00:1b:21:a1:19:dc,dst=fa:16:3e:76:88:c0),eth_type(0x0800),ipv4(src=191.101.100.254,dst=191.101.0.28,proto=6,tos=0,ttl=64,frag=no),tcp(src=33032,dst=5001), > packets:5009744, bytes:120178504800, used:0.000s, flags:P., actions:4 > > package size = 120178504800/ 5009744 = 23989 > > > > in_port(4),eth(src=fa:16:3e:76:88:c0,dst=00:1b:21:a1:19:dc),eth_type(0x0800),ipv4(src=191.101.0.28,dst=191.101.100.254,proto=6,tos=0,ttl=64,frag=no),tcp(src=5001,dst=33032), > packets:4728343, bytes:312070710, used:0.000s, flags:., > actions:push_vlan(vid=1,pcp=0),2 > > package size = 312070710/ 4728343 = 66 > > > > > > The question I have is : > > Isn’t the device MTU restricted the max packages size ??? > > Why we still can observe such big packages ??
It's due to the use of TSO. The NIC will segment the packets to the appropriate MTU before they are transmitted on the wire. _______________________________________________ discuss mailing list discuss@openvswitch.org http://openvswitch.org/mailman/listinfo/discuss
