<div>Hi,Murphy,</div><div> </div><div>I use your pox /murphypox . Becasue thi version solves a lot more bugs which are necessary for me.</div><div> </div><div>I run a component based on l2_Multi with inserting my own functionality( no influence on flood), spanning_tree and a component to manage flow entries ( manageflow.py).th</div>
<div> </div><div>Here I run " sudo ./pox.py log.level --WARNING l2_multiUDP openflow.spanning_tree manageflow</div><div> </div><div>The thing is that sometimes I can see that " Not flooding, holding down active" appear, when this appears, it seems that switches refuse to flood other than the one which connects to source. From Sam's message, it seems that this is a self protection to avoid overload, but it shouldn't happen with spanning_tree and 100pkts/sec packets.</div>
<div> </div><div>Thanks</div><div> </div><div>Weiyang<br><br></div><div class="gmail_quote">2012/10/9 Murphy McCauley <span dir="ltr"><<a href="mailto:murphy.mccauley@gmail.com" target="_blank">murphy.mccauley@gmail.com</a>></span><br>
<blockquote style="margin:0px 0px 0px 0.8ex;padding-left:1ex;border-left-color:rgb(204,204,204);border-left-width:1px;border-left-style:solid" class="gmail_quote">I think you'll get better answers if you provide more information.<br>
<br>
What branch of POX are you running? What revision?<br>
What components are you running? What was your commandline?<br>
What are the exact messages you see? Can you paste the log?<br>
<span class="HOEnZb"><font color="#888888"><br>
-- Murphy<br>
</font></span><div class="HOEnZb"><div class="h5"><br>
On Oct 9, 2012, at 11:38 AM, Weiyang Mo wrote:<br>
<br>
> Hi,<br>
><br>
> I saw some strange behaviors when I play with a few OpenFlow switches and IXIA traffic generator.<br>
><br>
> The traffic generator server/host don't act as normal PCs thus learning switch will not work. That means that even the packets from source are flooded to the destination, the destination will not reply.<br>
><br>
> I try to solve this issue by also sending some packets from destination to let controller know where is the destination, then the source packets will be forwarded to destination according to my OSPF algorithm.<br>
><br>
> However I found something really strange that the packets are not flooded to all ports although using OFPP.ALL. the topology is as follows:<br>
><br>
> PC1 ----- SW1 --------SW2--------PC2<br>
> | |<br>
> IXIA1 IXIA2<br>
> I send traffic from IXIA1 to IXIA2. Since IXIA2 never replys as I mention above, the controller should flood packets out of all ports. However I find that not all ports are flooded such as PC2 and IXIA2 which connects to SW2. From the controller side, I see some information "Not flooding, holding down active"<br>
><br>
> I'm not sure whether the "hold down" stops packets flooding all ports. And what does hold_down mean?<br>
><br>
> BTW, another strange behavior is that this "not all flooded" issue not always happen. If I run the controller a few times, sometimes flooding all ports, sometimes not... And when it's not, it seems packets are dropped on some switches?<br>
><br>
> Thanks very much.<br>
<br>
</div></div></blockquote></div><br>