The reason I use the reference switch is cause I understood that it is user-space and it will be easier to add new features. I haven´t tried l2_pairs, i would do it. I know its strange that l3_learning does make the flow modification and i think the reason is that theres a problem with the switch implementation. I would give a look to the openvswitch. And I just update pox to the beta branch and when I run the "./pox.py openflow.of_01 --address=192.168.0.100 --port=6633 forwarding.l2_learning" it doesn´t print the debug, could you tell me you can I print it? Thanks for the quick reply
Subject: Re: [pox-dev] Problem with stanford's software reference design openflow switch From: [email protected] Date: Mon, 8 Apr 2013 15:07:33 -0700 CC: [email protected] To: [email protected] On Apr 8, 2013, at 2:57 PM, Mayumi Park Campos wrote:I have implemented the pox controller with a openflow stanford's software reference design openflow switch emulated on a pc and my problem is that when a try to run pox as forwarding l2_learning behavior my “switch” do not work as switch and do not install any flow but when I run pox as forwarding l3_learning it works fine. Could anyone tell me what should I do to make my emulated switch works as a switch and not as a router? I have not make any modification to the original code. Are you using the reference switch for any particular reason? Open vSwitch is considerably more maintained. It seems strange that it's working for l3_learning (which, incidentally, is more like an L3 switch than anything that anyone would call a router) and not l2_learning. Have you tried l2_pairs? Can you post the entire POX log at debug level for a test run? See the manual for info on configuring the log (there's a FAQ entry for troubleshooting).https://openflow.stanford.edu/display/ONL/POX+Wiki -- Murphy
