Re: [nox-dev] Nox + Invalid Ethernet Addr
Thanks, Michael, for sending the pcap and the debugging info offline. I believe we may have uncovered a NOX bug. I'll let KK and team confirm that. The source address you have translates to: >>> val=array.array('B', [0, 48, 5, 204, 90, 232]) >>> val = val.tostring() >>> val '\x000\x05\xccZ\xe8' The destination MAC addr you have translates to: >>> val=array.array('B', [0, 38, 85, 218, 58, 64]) >>> val = val.tostring() >>> val '\x00&U\xda:@' Unfortunately, the destination MAC address has a ":" character in binary. This causes NOX to incorrectly interpret this val string as a text (instead of binary) at Line 78 of src/nox/lib/util.py, and then perform a text conversation (instead of binary). This leads to the invalid ethernet address. KK and team will patch this. In the meantime, please try the hacky patch that I mailed you and write back to the list if that solves your problem. Thanks Srini. ___ nox-dev mailing list nox-dev@noxrepo.org http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org
Re: [nox-dev] Nox + Invalid Ethernet Addr
Hi Srini, your quick patch worked for me. The NetFPGA is now once again forwarding packets using Nox and OF 1.0 :) For everybody else: Srini made a change in the util.py file located in src/nox/lib/. Replace "if isinstance(val, str) and not (':' in val):" with "if isinstance(val, str) and string.count(val, ':') < (ethernetaddr.LEN - 1):" and "import string" Thx guys Am 17.08.2010 16:08, schrieb Srini Seetharaman: Hi Michael We need the detailed pcap and not the verbose mode output. Coud you please run "tcpdump -s0 port XX -w file.pcap" and email us (KK and I) the file offline. This is the way we will be able to look into a packet_in to see what MAC address the NetFPGA is using. Thanks! Srini. On Tue, Aug 17, 2010 at 6:29 AM, Michael Jarschel wrote: Hi, here's the output from tshark during while pinging: 36.479743 172.16.40.35 -> 172.16.40.31 OFP Packet Out (CSM) (BufID=964) (24B) Ver Warning! 36.479938 172.16.40.31 -> 172.16.40.35 OFP Packet In (AM) (BufID=965) (78B) Ver Warning! 36.516587 172.16.40.35 -> 172.16.40.31 TCP 6633> 51576 [ACK] Seq=673 Ack=4369 Win=108 Len=0 TSV=543152383 TSER=101721801 37.479228 172.16.40.31 -> 172.16.40.35 OFP Packet In (AM) (BufID=966) (78B) Ver Warning! 37.479335 172.16.40.35 -> 172.16.40.31 TCP 6633> 51576 [ACK] Seq=673 Ack=4447 Win=108 Len=0 TSV=543152623 TSER=101722800 37.479784 172.16.40.35 -> 172.16.40.31 OFP Packet Out (CSM) (BufID=966) (24B) Ver Warning! 37.479971 172.16.40.31 -> 172.16.40.35 OFP Packet In (AM) (BufID=967) (78B) Ver Warning! 37.516625 172.16.40.35 -> 172.16.40.31 TCP 6633> 51576 [ACK] Seq=697 Ack=4525 Win=108 Len=0 TSV=543152633 TSER=101722801 35 ist the controller, 31 is the switch. Regards, Michael On Mon, 16 Aug 2010 20:11:03 +0200, Michael Jarschel wrote: Hi KK, we are running the pyswitch and the packetdump module for testing. It works fine for the reference software switch. I am not sure wether this is a problem of the netfpga or nox-related. I'll do a tcpdump in the morning. Thx. Michael Am 16.08.10 19:43, schrieb kk yap: Hi, I am assuming you are running some Python module. Can you tell us which modules you are running? A tcpdump of the controller traffic might help too. Can you also run just switch or routing that as C? Thanks. Regards KK On 16 August 2010 08:19, Michael Jarschel wrote: >> Hi all, first off, I'm sorry, if this has been answered before, but Google didn't turn up any results and I don't have the time to scan all posts. We have recently switched to OF 1.0 from 0.89 and are running the current git version of the Nox branch for OpenFlow 1.0 as a controller for our NetFPGA OpenFlow switch. We have two machines connected to the switch and as a basic test wanted them to ping each other. Using dpctl to configure flow rules manually this works flawlessly. However, when the Nox Controller is connected to the datapath, the intial arp-reply packet is dropped by the controller because of an invalid ethernet address: 00084|openflow-event|DBG:received packet-in event from 0023207f7017 (len:60) invalid ethernet addr [(Fujitsu Siemens Computers):cc:5a:e8>00:26:55:da:3a:40:ARP](REPLY[hw:1p:2048[00:30:05:cc:5a:e8>00 :26:55:da:3a:40]:[192.168.0.2>192.168.0.1]) My guess would be, that the vendor code resolution displayed here causes the problem. However, I am not sure wether this is the case and in which library this happens. Any ideas ? Regards, Michael -- Dipl.-Inform. Michael Jarschel Chair of Communication Networks (Lehrstuhl für Informatik III) University of Wuerzburg, Am Hubland, 97074, Wuerzburg, Germany http://www3.informatik.uni-wuerzburg.de/staff/michael.jarschel/ mailto:michael.jarsc...@informatik.uni-wuerzburg.de Tel: +49 931 31-88768 Fax: +49 931 31-86632 ___ nox-dev mailing list nox-dev@noxrepo.org http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org -- Dipl.-Inform. Michael Jarschel Chair of Communication Networks (Lehrstuhl für Informatik III) University of Wuerzburg, Am Hubland, 97074, Wuerzburg, Germany http://www3.informatik.uni-wuerzburg.de/staff/michael.jarschel/ mailto:michael.jarsc...@informatik.uni-wuerzburg.de Tel: +49 931 31-88768 Fax: +49 931 31-86632 ___ nox-dev mailing list nox-dev@noxrepo.org http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org -- Dipl.-Inform. Michael Jarschel Chair of Communication Networks (Lehrstuhl für Informatik III) University of Wuerzburg, Am Hubland, 97074, Wuerzburg, Germany http://www3.informatik.uni-wuerzburg.de/staff/michael.jarschel/ mailto:michael.jarsc...@informatik.uni-wuerzburg.de Tel: +49 931 31-88768 Fax: +49 931 31-86632 ___ nox-dev mailing list nox-dev@noxrepo.org http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org
Re: [nox-dev] Nox + Invalid Ethernet Addr
Hi Michael We need the detailed pcap and not the verbose mode output. Coud you please run "tcpdump -s0 port XX -w file.pcap" and email us (KK and I) the file offline. This is the way we will be able to look into a packet_in to see what MAC address the NetFPGA is using. Thanks! Srini. On Tue, Aug 17, 2010 at 6:29 AM, Michael Jarschel wrote: > Hi, > > here's the output from tshark during while pinging: > > 36.479743 172.16.40.35 -> 172.16.40.31 OFP Packet Out (CSM) > (BufID=964) (24B) Ver Warning! > 36.479938 172.16.40.31 -> 172.16.40.35 OFP Packet In (AM) (BufID=965) > (78B) Ver Warning! > 36.516587 172.16.40.35 -> 172.16.40.31 TCP 6633 > 51576 [ACK] Seq=673 > Ack=4369 Win=108 Len=0 TSV=543152383 TSER=101721801 > 37.479228 172.16.40.31 -> 172.16.40.35 OFP Packet In (AM) (BufID=966) > (78B) Ver Warning! > 37.479335 172.16.40.35 -> 172.16.40.31 TCP 6633 > 51576 [ACK] Seq=673 > Ack=4447 Win=108 Len=0 TSV=543152623 TSER=101722800 > 37.479784 172.16.40.35 -> 172.16.40.31 OFP Packet Out (CSM) > (BufID=966) (24B) Ver Warning! > 37.479971 172.16.40.31 -> 172.16.40.35 OFP Packet In (AM) (BufID=967) > (78B) Ver Warning! > 37.516625 172.16.40.35 -> 172.16.40.31 TCP 6633 > 51576 [ACK] Seq=697 > Ack=4525 Win=108 Len=0 TSV=543152633 TSER=101722801 > > 35 ist the controller, 31 is the switch. > > Regards, > Michael > > > On Mon, 16 Aug 2010 20:11:03 +0200, Michael Jarschel > wrote: >> Hi KK, >> >> we are running the pyswitch and the packetdump module for testing. It >> works fine for the reference software switch. >> I am not sure wether this is a problem of the netfpga or nox-related. >> I'll do a tcpdump in the morning. >> >> Thx. >> Michael >> >> Am 16.08.10 19:43, schrieb kk yap: >>> Hi, >>> >>> I am assuming you are running some Python module. Can you tell us >>> which modules you are running? A tcpdump of the controller traffic >>> might help too. Can you also run just switch or routing that as C? >>> >>> Thanks. >>> >>> Regards >>> KK >>> >>> On 16 August 2010 08:19, Michael Jarschel >>> wrote: >>> >> Hi all, first off, I'm sorry, if this has been answered before, but Google didn't turn up any results and I don't have the time to scan all posts. We have recently switched to OF 1.0 from 0.89 and are running the current git version of the Nox branch for OpenFlow 1.0 as a controller for our NetFPGA OpenFlow switch. We have two machines connected to the switch and as a basic test wanted them to ping each other. Using dpctl to configure flow rules manually this works flawlessly. However, when the Nox Controller is connected to the datapath, the intial arp-reply packet is dropped by the controller because of an invalid ethernet address: 00084|openflow-event|DBG:received packet-in event from 0023207f7017 (len:60) invalid ethernet addr [(Fujitsu Siemens Computers):cc:5a:e8>00:26:55:da:3a:40:ARP](REPLY[hw:1p:2048[00:30:05:cc:5a:e8>00 :26:55:da:3a:40]:[192.168.0.2>192.168.0.1]) My guess would be, that the vendor code resolution displayed here causes the problem. However, I am not sure wether this is the case and in which library this happens. Any ideas ? Regards, Michael -- Dipl.-Inform. Michael Jarschel Chair of Communication Networks (Lehrstuhl für Informatik III) University of Wuerzburg, Am Hubland, 97074, Wuerzburg, Germany http://www3.informatik.uni-wuerzburg.de/staff/michael.jarschel/ mailto:michael.jarsc...@informatik.uni-wuerzburg.de Tel: +49 931 31-88768 Fax: +49 931 31-86632 ___ nox-dev mailing list nox-dev@noxrepo.org http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org >> >> -- >> Dipl.-Inform. Michael Jarschel >> Chair of Communication Networks (Lehrstuhl für Informatik III) >> University of Wuerzburg, Am Hubland, 97074, Wuerzburg, Germany >> http://www3.informatik.uni-wuerzburg.de/staff/michael.jarschel/ >> mailto:michael.jarsc...@informatik.uni-wuerzburg.de >> Tel: +49 931 31-88768 >> Fax: +49 931 31-86632 > > > ___ > nox-dev mailing list > nox-dev@noxrepo.org > http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org > ___ nox-dev mailing list nox-dev@noxrepo.org http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org
Re: [nox-dev] Nox + Invalid Ethernet Addr
Hi, here's the output from tshark during while pinging: 36.479743 172.16.40.35 -> 172.16.40.31 OFP Packet Out (CSM) (BufID=964) (24B) Ver Warning! 36.479938 172.16.40.31 -> 172.16.40.35 OFP Packet In (AM) (BufID=965) (78B) Ver Warning! 36.516587 172.16.40.35 -> 172.16.40.31 TCP 6633 > 51576 [ACK] Seq=673 Ack=4369 Win=108 Len=0 TSV=543152383 TSER=101721801 37.479228 172.16.40.31 -> 172.16.40.35 OFP Packet In (AM) (BufID=966) (78B) Ver Warning! 37.479335 172.16.40.35 -> 172.16.40.31 TCP 6633 > 51576 [ACK] Seq=673 Ack=4447 Win=108 Len=0 TSV=543152623 TSER=101722800 37.479784 172.16.40.35 -> 172.16.40.31 OFP Packet Out (CSM) (BufID=966) (24B) Ver Warning! 37.479971 172.16.40.31 -> 172.16.40.35 OFP Packet In (AM) (BufID=967) (78B) Ver Warning! 37.516625 172.16.40.35 -> 172.16.40.31 TCP 6633 > 51576 [ACK] Seq=697 Ack=4525 Win=108 Len=0 TSV=543152633 TSER=101722801 35 ist the controller, 31 is the switch. Regards, Michael On Mon, 16 Aug 2010 20:11:03 +0200, Michael Jarschel wrote: > Hi KK, > > we are running the pyswitch and the packetdump module for testing. It > works fine for the reference software switch. > I am not sure wether this is a problem of the netfpga or nox-related. > I'll do a tcpdump in the morning. > > Thx. > Michael > > Am 16.08.10 19:43, schrieb kk yap: >> Hi, >> >> I am assuming you are running some Python module. Can you tell us >> which modules you are running? A tcpdump of the controller traffic >> might help too. Can you also run just switch or routing that as C? >> >> Thanks. >> >> Regards >> KK >> >> On 16 August 2010 08:19, Michael Jarschel >> wrote: >>>> Hi all, >>> >>> first off, I'm sorry, if this has been answered before, but Google didn't >>> turn up any results and I don't have the time to scan all posts. >>> We have recently switched to OF 1.0 from 0.89 and are running the current >>> git version of the Nox branch for OpenFlow 1.0 as a controller for our >>> NetFPGA OpenFlow switch. We have two machines connected to the switch and as >>> a basic test wanted them to ping each other. Using dpctl to configure flow >>> rules manually this works flawlessly. However, when the Nox Controller is >>> connected to the datapath, the intial arp-reply packet is dropped by the >>> controller because of an invalid ethernet address: >>> >>> 00084|openflow-event|DBG:received packet-in event from 0023207f7017 (len:60) >>> invalid ethernet addr >>> [(Fujitsu Siemens >>> Computers):cc:5a:e8>00:26:55:da:3a:40:ARP](REPLY[hw:1p:2048[00:30:05:cc:5a:e8>00 >>> :26:55:da:3a:40]:[192.168.0.2>192.168.0.1]) >>> >>> My guess would be, that the vendor code resolution displayed here causes the >>> problem. However, I am not sure wether this is the case and in which library >>> this happens. Any ideas ? >>> >>> Regards, >>> Michael >>> >>> -- >>> Dipl.-Inform. Michael Jarschel >>> Chair of Communication Networks (Lehrstuhl für Informatik III) >>> University of Wuerzburg, Am Hubland, 97074, Wuerzburg, Germany >>> http://www3.informatik.uni-wuerzburg.de/staff/michael.jarschel/ >>> mailto:michael.jarsc...@informatik.uni-wuerzburg.de >>> Tel: +49 931 31-88768 >>> Fax: +49 931 31-86632 >>> >>> >>> ___ >>> nox-dev mailing list >>> nox-dev@noxrepo.org >>> http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org >>> >>> > > -- > Dipl.-Inform. Michael Jarschel > Chair of Communication Networks (Lehrstuhl für Informatik III) > University of Wuerzburg, Am Hubland, 97074, Wuerzburg, Germany > http://www3.informatik.uni-wuerzburg.de/staff/michael.jarschel/ > mailto:michael.jarsc...@informatik.uni-wuerzburg.de > Tel: +49 931 31-88768 > Fax: +49 931 31-86632 ___ nox-dev mailing list nox-dev@noxrepo.org http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org
Re: [nox-dev] Nox + Invalid Ethernet Addr
Hi KK, we are running the pyswitch and the packetdump module for testing. It works fine for the reference software switch. I am not sure wether this is a problem of the netfpga or nox-related. I'll do a tcpdump in the morning. Thx. Michael Am 16.08.10 19:43, schrieb kk yap: Hi, I am assuming you are running some Python module. Can you tell us which modules you are running? A tcpdump of the controller traffic might help too. Can you also run just switch or routing that as C? Thanks. Regards KK On 16 August 2010 08:19, Michael Jarschel wrote: Hi all, first off, I'm sorry, if this has been answered before, but Google didn't turn up any results and I don't have the time to scan all posts. We have recently switched to OF 1.0 from 0.89 and are running the current git version of the Nox branch for OpenFlow 1.0 as a controller for our NetFPGA OpenFlow switch. We have two machines connected to the switch and as a basic test wanted them to ping each other. Using dpctl to configure flow rules manually this works flawlessly. However, when the Nox Controller is connected to the datapath, the intial arp-reply packet is dropped by the controller because of an invalid ethernet address: 00084|openflow-event|DBG:received packet-in event from 0023207f7017 (len:60) invalid ethernet addr [(Fujitsu Siemens Computers):cc:5a:e8>00:26:55:da:3a:40:ARP](REPLY[hw:1p:2048[00:30:05:cc:5a:e8>00 :26:55:da:3a:40]:[192.168.0.2>192.168.0.1]) My guess would be, that the vendor code resolution displayed here causes the problem. However, I am not sure wether this is the case and in which library this happens. Any ideas ? Regards, Michael -- Dipl.-Inform. Michael Jarschel Chair of Communication Networks (Lehrstuhl für Informatik III) University of Wuerzburg, Am Hubland, 97074, Wuerzburg, Germany http://www3.informatik.uni-wuerzburg.de/staff/michael.jarschel/ mailto:michael.jarsc...@informatik.uni-wuerzburg.de Tel: +49 931 31-88768 Fax: +49 931 31-86632 ___ nox-dev mailing list nox-dev@noxrepo.org http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org -- Dipl.-Inform. Michael Jarschel Chair of Communication Networks (Lehrstuhl für Informatik III) University of Wuerzburg, Am Hubland, 97074, Wuerzburg, Germany http://www3.informatik.uni-wuerzburg.de/staff/michael.jarschel/ mailto:michael.jarsc...@informatik.uni-wuerzburg.de Tel: +49 931 31-88768 Fax: +49 931 31-86632 ___ nox-dev mailing list nox-dev@noxrepo.org http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org
Re: [nox-dev] Nox + Invalid Ethernet Addr
Hi, I am assuming you are running some Python module. Can you tell us which modules you are running? A tcpdump of the controller traffic might help too. Can you also run just switch or routing that as C? Thanks. Regards KK On 16 August 2010 08:19, Michael Jarschel wrote: > Hi all, > > first off, I'm sorry, if this has been answered before, but Google didn't > turn up any results and I don't have the time to scan all posts. > We have recently switched to OF 1.0 from 0.89 and are running the current > git version of the Nox branch for OpenFlow 1.0 as a controller for our > NetFPGA OpenFlow switch. We have two machines connected to the switch and as > a basic test wanted them to ping each other. Using dpctl to configure flow > rules manually this works flawlessly. However, when the Nox Controller is > connected to the datapath, the intial arp-reply packet is dropped by the > controller because of an invalid ethernet address: > > 00084|openflow-event|DBG:received packet-in event from 0023207f7017 (len:60) > invalid ethernet addr > [(Fujitsu Siemens > Computers):cc:5a:e8>00:26:55:da:3a:40:ARP](REPLY[hw:1p:2048[00:30:05:cc:5a:e8>00 > :26:55:da:3a:40]:[192.168.0.2>192.168.0.1]) > > My guess would be, that the vendor code resolution displayed here causes the > problem. However, I am not sure wether this is the case and in which library > this happens. Any ideas ? > > Regards, > Michael > > -- > Dipl.-Inform. Michael Jarschel > Chair of Communication Networks (Lehrstuhl für Informatik III) > University of Wuerzburg, Am Hubland, 97074, Wuerzburg, Germany > http://www3.informatik.uni-wuerzburg.de/staff/michael.jarschel/ > mailto:michael.jarsc...@informatik.uni-wuerzburg.de > Tel: +49 931 31-88768 > Fax: +49 931 31-86632 > > > ___ > nox-dev mailing list > nox-dev@noxrepo.org > http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org > ___ nox-dev mailing list nox-dev@noxrepo.org http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org
[nox-dev] Nox + Invalid Ethernet Addr
Hi all, first off, I'm sorry, if this has been answered before, but Google didn't turn up any results and I don't have the time to scan all posts. We have recently switched to OF 1.0 from 0.89 and are running the current git version of the Nox branch for OpenFlow 1.0 as a controller for our NetFPGA OpenFlow switch. We have two machines connected to the switch and as a basic test wanted them to ping each other. Using dpctl to configure flow rules manually this works flawlessly. However, when the Nox Controller is connected to the datapath, the intial arp-reply packet is dropped by the controller because of an invalid ethernet address: 00084|openflow-event|DBG:received packet-in event from 0023207f7017 (len:60) invalid ethernet addr [(Fujitsu Siemens Computers):cc:5a:e8>00:26:55:da:3a:40:ARP](REPLY[hw:1p:2048[00:30:05:cc:5a:e8>00 :26:55:da:3a:40]:[192.168.0.2>192.168.0.1]) My guess would be, that the vendor code resolution displayed here causes the problem. However, I am not sure wether this is the case and in which library this happens. Any ideas ? Regards, Michael -- Dipl.-Inform. Michael Jarschel Chair of Communication Networks (Lehrstuhl für Informatik III) University of Wuerzburg, Am Hubland, 97074, Wuerzburg, Germany http://www3.informatik.uni-wuerzburg.de/staff/michael.jarschel/ mailto:michael.jarsc...@informatik.uni-wuerzburg.de Tel: +49 931 31-88768 Fax: +49 931 31-86632 ___ nox-dev mailing list nox-dev@noxrepo.org http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org