[ovs-discuss] first test on an hybrid switched network

Pásztor Lénárd Zoltán lenard.pasztor at gmail.com
Fri Jul 16 14:00:22 UTC 2010


Hi,

Currently I just compiled NOX (0.8.0) and openvswitch (1.01 HEAD). They
communicate each other with openflow 1.0.0.
I attached a picture about my test network.

gw: Is a PC running NOX controller, has etherchannel configured on the Cisco
switch, and bonding on linux
vm1,vm2,vm3 nodes running xen hypervisor, have two psyhical interfaces
connected to the local lan.
br0 is a vswitchd running on vm1,2,3 nodes configured to use physical eth0
and eth1 as a bonded ports.
domU1-9 running xen pv guests with one vif per vm connected to local br0.
This physical PC-s connected together using a Cisco switch.

If I don't use NOX controller for the openvswitches the domU network also
works well.

Now try to connect NOX and vswitchd:

1. start controller on GW
nox_core -i ptcp:2525 switch packetdump
NOX 0.8.0~full~beta (nox_core), compiled Jul 15 2010 22:15:41
Compiled with OpenFlow 0x01

2. set controller on vm1
ovs-vsctl set-controller tcp:172.24.15.254:2525

Then the communication stopped, and came back after I shut down the
controller and the openvswitch goes to fail-open mode.

Logs on controller:
[(Intel Corporate):1f:e4:10>(Quanta Computer
Inc.):03:55:ce:IP]([v:4hl:5l:100t:64]TCP
cs:b826[172.24.1.1>172.24.15.252]){22>4484} seq:3994195419 ack:511570489
f:0x18
[(Quanta Computer Inc.):03:55:ce>(Intel
Corporate):1f:e4:10:IP]([v:4hl:5l:52t:64]TCP
cs:c47a[172.24.15.252>172.24.1.1]){4484>22} seq:511570489 ack:3994195467
f:0x10
[(Quanta Computer Inc.):03:55:ce>(Intel
Corporate):1f:e4:10:IP]([v:4hl:5l:64t:64]TCP
cs:c46d[172.24.15.252>172.24.1.1]){4484>22} seq:511570489 ack:3994195467
f:0x10
[(Intel Corporate):1f:e4:10>(Quanta Computer
Inc.):03:55:ce:IP]([v:4hl:5l:292t:64]TCP
cs:c608[172.24.1.1>172.24.15.252]){22>33158} seq:586669489 ack:1406889139
f:0x18
[(Quanta Computer Inc.):03:55:ce>(Intel
Corporate):1f:e4:10:IP]([v:4hl:5l:52t:64]TCP
cs:cdd[172.24.15.252>172.24.1.1]){33158>22} seq:1406889139 ack:586669729
f:0x10
[28:93:fe:d0:92:07>01:80:c2:00:00:00:llc]
00001|openflow|WARN:stream: no response to inactivity probe after 15
seconds, disconnecting
00002|nox|WARN:stream: disconnected (Protocol error)

Logs on vm1:
Jul 16 12:23:17 vm1 ovs-vsctl: 00001|vsctl|INFO|Called as ovs-vsctl
set-controller tcp:172.24.15.254:2525
Jul 16 12:23:17 vm1 ovs-vswitchd: 01305|rconn|INFO|dp0<->tcp:
172.24.15.254:2525: connecting...
Jul 16 12:23:17 vm1 ovs-vswitchd: 01306|rconn|INFO|dp0<->tcp:
172.24.15.254:2525: connection timed out
Jul 16 12:23:17 vm1 ovs-vswitchd: 01307|rconn|INFO|dp0<->tcp:
172.24.15.254:2525: waiting 1 seconds before reconnect
Jul 16 12:23:18 vm1 ovs-vswitchd: 01308|rconn|INFO|dp0<->tcp:
172.24.15.254:2525: connecting...
Jul 16 12:23:18 vm1 ovs-vswitchd: 01309|rconn|INFO|dp0<->tcp:
172.24.15.254:2525: connected
Jul 16 12:23:18 vm1 ovs-vswitchd: 01310|ofproto|INFO|dp0<->tcp:
172.24.15.254:2525: normalization changed ofp_match, details:
Jul 16 12:23:18 vm1 ovs-vswitchd: 01311|ofproto|INFO| pre:
wildcards=0xffffffff  in_port=    0  dl_src=00:00:80:e4:73:00
 dl_dst=00:00:00:00:80:e4  dl_vlan=29440  dl_vlan_pcp=  0  dl_type=     0
 nw_tos=   0  nw_proto=   0  nw_src=         0  nw_dst=         0  tp_src=
 0  tp_dst=    0
Jul 16 12:23:18 vm1 ovs-vswitchd: 01312|ofproto|INFO|post: wildcards=
0x23fffff  in_port=    0  dl_src=00:00:00:00:00:00  dl_dst=00:00:00:00:00:00
 dl_vlan=29440  dl_vlan_pcp=  0  dl_type=     0  nw_tos=   0  nw_proto=   0
 nw_src=         0  nw_dst=         0  tp_src=    0  tp_dst=    0
Jul 16 12:23:28 vm1 ovs-vswitchd: 01313|rconn|ERR|dp0<->tcp:
172.24.15.254:2525: no response to inactivity probe after 5 seconds,
disconnecting
Jul 16 12:23:29 vm1 ovs-vswitchd: 01314|rconn|INFO|dp0<->tcp:
172.24.15.254:2525: connecting...
Jul 16 12:23:30 vm1 ovs-vswitchd: 01315|rconn|INFO|dp0<->tcp:
172.24.15.254:2525: connection timed out
Jul 16 12:23:30 vm1 ovs-vswitchd: 01316|rconn|INFO|dp0<->tcp:
172.24.15.254:2525: waiting 2 seconds before reconnect
Jul 16 12:23:32 vm1 ovs-vswitchd: 01317|rconn|INFO|dp0<->tcp:
172.24.15.254:2525: connecting...
Jul 16 12:23:33 vm1 ovs-vswitchd: 01318|fail_open|WARN|Could not connect to
controller (or switch failed controller's post-connection admission control
policy) for 15 seconds, failing open

hwaddresses:
gw 00:23:8b:03:55:ce (it is a bonded interface on an etherchannel)
vm1 br0 00:15:17:1f:e4:10
vm1 eth0 00:15:17:1f:e4:10
vm1 eth1 00:15:17:1f:e4:10

If I don't use openvswitch's port-bonding on vm nodes, but only one port at
br0 the network seems fine with nox controller too, but the controller logs
this:

[(Xensource, Inc.):64:80:78>(Quanta Computer
Inc.):03:55:ce:IP]([v:4hl:5l:67t:64]UDP
cs:fe9b[172.24.1.222>172.24.15.252]){60900>53} l:47 c: 16991(id:8720 fl:|RD
| op:0 nq:1 na:0 nath:0 nadd:0)(q? us.archive.ubuntu.com A IN)
00568|packet|ERR:'next_rr: data truncated'
[(Quanta Computer Inc.):03:55:ce>(Xensource,
Inc.):64:80:78:IP]([v:4hl:5l:227t:64]UDP
cs:91d4[172.24.15.252>172.24.1.222]){53>60900} l:207 c: 8427
[(Xensource, Inc.):64:80:78>(Quanta Computer
Inc.):03:55:ce:IP]([v:4hl:5l:67t:64]UDP
cs:fe9b[172.24.1.222>172.24.15.252]){49850>53} l:47 c: 52087(id:2932 fl:|RD
| op:0 nq:1 na:0 nath:0 nadd:0)(q? us.archive.ubuntu.com A IN)
00569|packet|ERR:'next_rr: data truncated'

What happends, when I got this messages?

About bonding, what I miss to configure, to let the bonding work with
openflow controller?


thx for your answers,
Lenard
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://openvswitch.org/pipermail/ovs-discuss/attachments/20100716/97f2831f/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.png
Type: image/png
Size: 16400 bytes
Desc: not available
URL: <http://openvswitch.org/pipermail/ovs-discuss/attachments/20100716/97f2831f/attachment-0004.png>


More information about the discuss mailing list