I have created the overlay network that appears on my first mail.
The hosts and the OVS are running on slivers on PlaneLab. Each one is a
different PLE node.
On Planetlab the OVS is running on userspace mode.
On the OVS node:
The output of "sliver-ovs show" or "ovs-vsctl show" is
###
Bridge inria_nepi
Port inria_nepi
Interface inria_nepi
type: internal
options: {local_ip="192.168.3.1", local_netmask="24"}
Port LSENDER-WEB
Interface LSENDER-WEB
type: tunnel
options: {remote_ip="192.38.109.144", remote_port="2222"}
Port LSENDER-USER
Interface LSENDER-USER
type: tunnel
options: {remote_ip="148.81.140.193", remote_port="2222"}
###
and the output of ifconfig is
###
eth0 Link encap:Ethernet HWaddr 00:18:8B:F8:E2:75
inet addr:147.83.30.167 Bcast:147.83.30.191 Mask:255.255.255.192
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:184958309 errors:0 dropped:0 overruns:0 frame:0
TX packets:241174705 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:38807681873 (36.1 GiB) TX bytes:33298506912 (31.0 GiB)
Interrupt:16
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:49006888 errors:0 dropped:0 overruns:0 frame:0
TX packets:49006888 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:8164153718 (7.6 GiB) TX bytes:8164153718 (7.6 GiB)
tap1184-1 Link encap:Ethernet HWaddr 66:52:4B:75:76:C9
inet addr:192.168.3.1 Bcast:192.168.3.255 Mask:255.255.255.0
UP BROADCAST RUNNING PROMISC MULTICAST MTU:1500 Metric:1
RX packets:72 errors:0 dropped:0 overruns:0 frame:0
TX packets:107 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:500
RX bytes:6284 (6.1 KiB) TX bytes:10730 (10.4 KiB)
###
and also the output of "ovs-ofctl show inria_nepi"
###
1(LSENDER-WEB): addr:fe:ff:ff:00:00:00
config: PORT_DOWN
state: LINK_DOWN
speed: 0 Mbps now, 0 Mbps max
2(LSENDER-USER): addr:fe:ff:ff:00:00:01
config: PORT_DOWN
state: LINK_DOWN
speed: 0 Mbps now, 0 Mbps max
LOCAL(inria_nepi): addr:66:52:4b:75:76:c9
config: 0
state: 0
speed: 0 Mbps now, 0 Mbps max
OFPT_GET_CONFIG_REPLY (xid=0x4): frags=normal miss_send_len=0
###
On each host the ifconfig output is:
###
eth0 Link encap:Ethernet HWaddr 00:1D:09:F0:95:4E
inet addr:148.81.140.193 Bcast:148.81.140.199
Mask:255.255.255.248
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:550814908 errors:0 dropped:0 overruns:0 frame:0
TX packets:588989063 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:253532467683 (236.1 GiB) TX bytes:122759328585 (114.3
GiB)
Interrupt:16
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:9527169 errors:0 dropped:0 overruns:0 frame:0
TX packets:9527169 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:3351470425 (3.1 GiB) TX bytes:3351470425 (3.1 GiB)
tap1422-0 Link encap:Ethernet HWaddr 96:F2:23:57:65:5A
inet addr:192.168.3.2 Bcast:192.168.3.2 Mask:255.255.255.255
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:52 errors:0 dropped:0 overruns:0 frame:0
TX packets:36 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:500
RX bytes:5416 (5.2 KiB) TX bytes:3016 (2.9 KiB)
###
So, when I start the network (without running a controller) I can have
pings from each host to the other that passes through the OVS.
After that I delete the flows on the OVS with "ovs-ofctl del-flows
inria_nepi" and I try to insert manually the flows.
So, on the OVS terminal I write:
$ ovs-ofctl add-flow inria_nepi in_port=1,actions=output:2
$ ovs-ofctl add-flow inria_nepi in_port=2,actions=output:1
On the port 1 is the host USER and on the port 2 is the WEB host.
When I am pinging from the USER to the WEB I got no reply.
BUT when I run on the OVS "ovs-ofctl dump-flows inria_nepi" I can see that
the switch receives packets from the host WEB but I cannot see received
packets from the USER host. In fact by running "ovs-ofctl dump-flows
inria_nepi" on the OVS I get:
"cookie=0x0, duration=20.845s, table=0, n_packets=3, n_bytes=180,
idle_age=14, in_port=2 actions=output:1"
I noticed with tcpdump on the tap device of the USER is not listening to
any packet. Also, the tap device of the OVS is not listening to packets.
What exactly am I doing wrong in the manually configuration of the flows?
In my mind, when I delete the flows from the OVS, the communication between
the eth0 interface and the tap device, stops.
How comes that?
2013/4/23 Kouvakas Alexandros <questa...@gmail.com>
> @ Felician Nemeth
> I am using sliver-ovs to create the bridge on OF switch but I am not using
> the Makefile because it creates OVSwitchs on every node, which I don't want
> it. I want to have OVS only on the node with IP 193.168.3.1 and just hosts
> on the nodes ..3.2 and ..3.3.
>
> Finally, I found the solution by creating tunnels between the OVS and the
> hosts from a tutorial of NEPI.
>
> Thanks for the help.
>
>
>
> 2013/4/19 Felician Nemeth <neme...@tmit.bme.hu>
>
>> > I am trying to create an overlay network with planetlab nodes.
>>
>> I think it is much more easier to use the freshly launched sliver-ovs
>> functionality:
>> https://www.planet-lab.eu/doc/guides/user/practices/openflow
>>
>> cheers,
>> Felicián
>> _______________________________________________
>> discuss mailing list
>> discuss@openvswitch.org
>> http://openvswitch.org/mailman/listinfo/discuss
>>
>
>
>
> --
> Kouvakas Alexandros
>
--
Kouvakas Alexandros
_______________________________________________
discuss mailing list
discuss@openvswitch.org
http://openvswitch.org/mailman/listinfo/discuss
