[ovs-dev] [PATCH]: Trio of small fixes
Ben Pfaff
blp at nicira.com
Thu Oct 28 17:33:07 UTC 2010
On Thu, Oct 28, 2010 at 10:20:36AM -0700, Jean Tourrilhes wrote:
> On Thu, Oct 28, 2010 at 10:16:12AM -0700, Ben Pfaff wrote:
> >
> > Can you describe the problem that this fixes in a bit more detail?
> >
> > Without your patch, on "master", when I run:
> >
> > ovs-dpctl add-dp br0
> > ovs-dpctl add-if br0 eth0
> > ovs-dpctl add-if br0 eth1
> > ovs-dpctl add-if br0 eth2
> > ovs-openflowd --out-of-band br0 unix:/tmp/socket
>
> You will also need fail-secure : --fail=closed
>
> > ovs-ofctl show br0
> >
> > I see the LOCAL port listed in the output, and when I further run:
> >
> > ifconfig br0 192.168.0.20
> >
> > I can "ping" br0 from elsewhere. So the problem isn't obvious.
>
> Can you do :
> ovs-ofctl show
> ovs-ofctl dump-ports
> And verify that LOCAL is listed there (it is for in-band).
> My symptoms is that OFPP_FLOOD does not flood on OFPP_LOCAL.
OK, here's my session:
hda:/mnt/_build# ovs-dpctl del-dp br0
hda:/mnt/_build# ovs-dpctl add-dp br0
hda:/mnt/_build# ovs-dpctl add-if br0 eth0
hda:/mnt/_build# ovs-dpctl add-if br0 eth1
hda:/mnt/_build# ovs-dpctl add-if br0 eth2
hda:/mnt/_build# ovs-openflowd --out-of-band --fail=closed br0 unix:/tmp/socket
Oct 28 10:10:26|00001|openflowd|INFO|Open vSwitch version 1.1.0pre2
Oct 28 10:10:26|00002|openflowd|INFO|OpenFlow protocol version 0x01
Oct 28 10:10:26|00003|ofproto|INFO|using datapath ID 0000002320800b82
Oct 28 10:10:26|00004|rconn|INFO|dp0<->unix:/tmp/socket: connecting...
Oct 28 10:10:26|00005|rconn|INFO|dp0<->unix:/tmp/socket: connected
In a separate terminal I'm running "ovs-controller punix:/tmp/socket
--hub" (which causes all flows to be flooded)".
In a third terminal:
hda:~# ovs-ofctl show br0
Oct 28 10:10:44|00001|ofctl|INFO|connecting to unix:/var/run/openvswitch/br0.mgmt
features_reply (xid=0x6d5b5): ver:0x1, dpid:0000002320800b82
n_tables:2, n_buffers:256
features: capabilities:0x87, actions:0xfff
1(eth0): addr:50:54:00:00:00:01, config: 0, state:0
current: 100MB-FD AUTO_NEG
advertised: 10MB-HD 10MB-FD 100MB-HD 100MB-FD COPPER AUTO_NEG
supported: 10MB-HD 10MB-FD 100MB-HD 100MB-FD COPPER AUTO_NEG
2(eth1): addr:50:54:00:00:00:02, config: 0, state:0
current: 100MB-FD AUTO_NEG
advertised: 10MB-HD 10MB-FD 100MB-HD 100MB-FD COPPER AUTO_NEG
supported: 10MB-HD 10MB-FD 100MB-HD 100MB-FD COPPER AUTO_NEG
3(eth2): addr:50:54:00:00:00:03, config: 0, state:0
current: 100MB-FD AUTO_NEG
advertised: 10MB-HD 10MB-FD 100MB-HD 100MB-FD COPPER AUTO_NEG
supported: 10MB-HD 10MB-FD 100MB-HD 100MB-FD COPPER AUTO_NEG
LOCAL(br0): addr:00:23:20:f2:4b:0e, config: 0, state:0
Oct 28 10:10:44|00002|ofctl|INFO|connecting to unix:/var/run/openvswitch/br0.mgmt
get_config_reply (xid=0xbd003): miss_send_len=0
hda:~# ovs-ofctl dump-ports br0
Oct 28 10:10:48|00001|ofctl|INFO|connecting to unix:/var/run/openvswitch/br0.mgmt
stats_reply (xid=0xf4dc3): flags=none type=4(port)
4 ports
port 65534: rx pkts=6, bytes=468, drop=0, errs=0, frame=0, over=0, crc=0
tx pkts=42, bytes=2520, drop=0, errs=0, coll=0
port 3: rx pkts=2065, bytes=170515, drop=0, errs=0, frame=0, over=0, crc=0
tx pkts=1736, bytes=189230, drop=0, errs=0, coll=0
port 2: rx pkts=1235, bytes=57104, drop=0, errs=0, frame=0, over=0, crc=0
tx pkts=1694, bytes=115806, drop=0, errs=0, coll=0
port 1: rx pkts=55018, bytes=2531136, drop=0, errs=0, frame=0, over=0, crc=0
tx pkts=1681, bytes=115026, drop=0, errs=0, coll=0
hda:~#
Running "tcpdump -i br0" first fails because br0 is not up. If I bring
it up with "ifconfig br0 up" and then rerun "tcpdump -i br0" then I do
see traffic from a "ping" that I'm running through eth0 and eth1.
More information about the dev
mailing list