Re: [c-nsp] OSPFv3 Adjacency issues between 2921 and 3750
Hi Peter, I would also suggest upgrading to 12.2(55)SE6, that should work. Note that if the switch has not joined the FF02::5 group address, you'll not see the packets destined to ff02::5 in 'debug ipv6 icmp' output. You can check this with 'sh ipv6 int type x/y' command. Best regards, Andras On Fri, Dec 7, 2012 at 3:59 PM, Gabor Ivanszky gaborivans...@gmail.comwrote: Hi Peter, I've seen the same phenomenon on 12.2(50)SE3. After upgrading to 12.2(55)SE6, OSPFv3 started to work(adjacency FULL-FULL and stable) However ping to ff02::5 still doesn't work. regards, Gabor On Tue, Jul 3, 2012 at 1:01 PM, Peter Subnovic cnspmail...@googlemail.com wrote: Hi Peter, thank you for your suggestion. Unfortunately QoS is disabled on the 3750. 3750#show mls qos QoS is disabled QoS ip packet dscp rewrite is enabled 3750#show mls qos interface g1/0/24 GigabitEthernet1/0/24 QoS is disabled. Kind regards, Peter On Tue, Jul 3, 2012 at 1:49 PM, Peter Rathlev pe...@rathlev.dk wrote: On Tue, 2012-07-03 at 05:43 -0400, Chuck Church wrote: Long shot, but don't the 3750s have a certain SDM they need enabled to use IPv6? And on that note, what about MLS QoS on the 3750? I seem to recall having debugged something that turned out to be much (all?) IPv6 traffic ending up in queue 3 or 4, and we have set buffer sizes for these to 0. Does no mls qos make any difference? (If it's enabled of course.) -- Peter ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] OSPFv3 Adjacency issues between 2921 and 3750
Hi Peter, I've seen the same phenomenon on 12.2(50)SE3. After upgrading to 12.2(55)SE6, OSPFv3 started to work(adjacency FULL-FULL and stable) However ping to ff02::5 still doesn't work. regards, Gabor On Tue, Jul 3, 2012 at 1:01 PM, Peter Subnovic cnspmail...@googlemail.com wrote: Hi Peter, thank you for your suggestion. Unfortunately QoS is disabled on the 3750. 3750#show mls qos QoS is disabled QoS ip packet dscp rewrite is enabled 3750#show mls qos interface g1/0/24 GigabitEthernet1/0/24 QoS is disabled. Kind regards, Peter On Tue, Jul 3, 2012 at 1:49 PM, Peter Rathlev pe...@rathlev.dk wrote: On Tue, 2012-07-03 at 05:43 -0400, Chuck Church wrote: Long shot, but don't the 3750s have a certain SDM they need enabled to use IPv6? And on that note, what about MLS QoS on the 3750? I seem to recall having debugged something that turned out to be much (all?) IPv6 traffic ending up in queue 3 or 4, and we have set buffer sizes for these to 0. Does no mls qos make any difference? (If it's enabled of course.) -- Peter ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] OSPFv3 Adjacency issues between 2921 and 3750
Hi Pete, thanks for your time. The 3750 does not respond when i try to ping ff02::5. It does however respond to the all-nodes/all-router ll Multicast address. The authentication configuration is only for ospfv2. So i think it will not make any difference.I cant try it without right now. Thanks, Peter On Tue, Jul 3, 2012 at 5:27 AM, Pete Lumbis alum...@gmail.com wrote: Any difference without authentication? Does the 3750 respond to pings to ff02::5? On Mon, Jul 2, 2012 at 4:16 PM, Peter Subnovic cnspmail...@googlemail.com wrote: Dear list members, i am having a somewhat odd issue (again) and i am out of ideas what is causing it. Let's assume the following very simple topology 2921 3750G-24PS All i want is to establish an OSPFv3 Adjacency between these two devices. Here is where the fun (headache) begins: The 2921 is running 15.1(4)M3 and the 3750G is running 12.2(55)SE5. So i configured OSPv3 on both sides on the respective Interfaces and wondered why the adjacency wont come up. After a little bit of troubleshooting it appears that the 3750G does not recognize the hello packets from the 2921. I can see in the debug output that they are received by the switch, but the OSPFv3 Process does not recognize them as such. If you need any additional input or show commands, please dont hesitate to ask. I appreciate all hints/comments. Thanks in advance, Peter ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] OSPFv3 Adjacency issues between 2921 and 3750
Thanks to Pete for the pointer, i was curios why the 3750 does not respond to ff02::5 (the 2921), so i activated the debugging for ipv6 packets again and started to ping ff02::5 from the 2921. Interestingly, i can't see the icmp packets at all (well atleast thats what the debug output is telling me on the 3750). The only thing i can see are the ospf hellos from the 2921... 2921#ping ipv6 ff02::5 size 1280 Output Interface: GigabitEthernet0/1 Type escape sequence to abort. Sending 5, 1280-byte ICMP Echos to FF02::5, timeout is 2 seconds: Packet sent with a source address of FE80::1%GigabitEthernet0/1 Request 0 timed out Request 1 timed out Request 2 timed out Request 3 timed out Request 4 timed out Success rate is 0 percent (0/5) 0 multicast replies and 0 errors. And the debug output on the 3750 looks like this: Jul 3 10:41:10.424 CEST: IPV6: source FE80::1 (GigabitEthernet1/0/24) Jul 3 10:41:10.424 CEST: dest FF02::5 Jul 3 10:41:10.424 CEST: traffic class 224, flow 0x0, len 80+14, prot 89, hops 1, forward to ulp Jul 3 10:41:20.096 CEST: IPV6: source FE80::1 (GigabitEthernet1/0/24) Jul 3 10:41:20.096 CEST: dest FF02::5 Jul 3 10:41:20.096 CEST: traffic class 224, flow 0x0, len 80+14, prot 89, hops 1, forward to ulp Regards, Peter On Tue, Jul 3, 2012 at 10:05 AM, Peter Subnovic cnspmail...@googlemail.comwrote: Hi Pete, thanks for your time. The 3750 does not respond when i try to ping ff02::5. It does however respond to the all-nodes/all-router ll Multicast address. The authentication configuration is only for ospfv2. So i think it will not make any difference.I cant try it without right now. Thanks, Peter On Tue, Jul 3, 2012 at 5:27 AM, Pete Lumbis alum...@gmail.com wrote: Any difference without authentication? Does the 3750 respond to pings to ff02::5? On Mon, Jul 2, 2012 at 4:16 PM, Peter Subnovic cnspmail...@googlemail.com wrote: Dear list members, i am having a somewhat odd issue (again) and i am out of ideas what is causing it. Let's assume the following very simple topology 2921 3750G-24PS All i want is to establish an OSPFv3 Adjacency between these two devices. Here is where the fun (headache) begins: The 2921 is running 15.1(4)M3 and the 3750G is running 12.2(55)SE5. So i configured OSPv3 on both sides on the respective Interfaces and wondered why the adjacency wont come up. After a little bit of troubleshooting it appears that the 3750G does not recognize the hello packets from the 2921. I can see in the debug output that they are received by the switch, but the OSPFv3 Process does not recognize them as such. If you need any additional input or show commands, please dont hesitate to ask. I appreciate all hints/comments. Thanks in advance, Peter ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] OSPFv3 Adjacency issues between 2921 and 3750
Long shot, but don't the 3750s have a certain SDM they need enabled to use IPv6? Chuck -Original Message- From: cisco-nsp-boun...@puck.nether.net [mailto:cisco-nsp-boun...@puck.nether.net] On Behalf Of Peter Subnovic Sent: Tuesday, July 03, 2012 4:46 AM To: cisco-nsp@puck.nether.net Subject: Re: [c-nsp] OSPFv3 Adjacency issues between 2921 and 3750 Thanks to Pete for the pointer, i was curios why the 3750 does not respond to ff02::5 (the 2921), so i activated the debugging for ipv6 packets again and started to ping ff02::5 from the 2921. Interestingly, i can't see the icmp packets at all (well atleast thats what the debug output is telling me on the 3750). The only thing i can see are the ospf hellos from the 2921... 2921#ping ipv6 ff02::5 size 1280 Output Interface: GigabitEthernet0/1 Type escape sequence to abort. Sending 5, 1280-byte ICMP Echos to FF02::5, timeout is 2 seconds: Packet sent with a source address of FE80::1%GigabitEthernet0/1 Request 0 timed out Request 1 timed out Request 2 timed out Request 3 timed out Request 4 timed out Success rate is 0 percent (0/5) 0 multicast replies and 0 errors. And the debug output on the 3750 looks like this: Jul 3 10:41:10.424 CEST: IPV6: source FE80::1 (GigabitEthernet1/0/24) Jul 3 10:41:10.424 CEST: dest FF02::5 Jul 3 10:41:10.424 CEST: traffic class 224, flow 0x0, len 80+14, prot 89, hops 1, forward to ulp Jul 3 10:41:20.096 CEST: IPV6: source FE80::1 (GigabitEthernet1/0/24) Jul 3 10:41:20.096 CEST: dest FF02::5 Jul 3 10:41:20.096 CEST: traffic class 224, flow 0x0, len 80+14, prot 89, hops 1, forward to ulp Regards, Peter On Tue, Jul 3, 2012 at 10:05 AM, Peter Subnovic cnspmail...@googlemail.comwrote: Hi Pete, thanks for your time. The 3750 does not respond when i try to ping ff02::5. It does however respond to the all-nodes/all-router ll Multicast address. The authentication configuration is only for ospfv2. So i think it will not make any difference.I cant try it without right now. Thanks, Peter On Tue, Jul 3, 2012 at 5:27 AM, Pete Lumbis alum...@gmail.com wrote: Any difference without authentication? Does the 3750 respond to pings to ff02::5? On Mon, Jul 2, 2012 at 4:16 PM, Peter Subnovic cnspmail...@googlemail.com wrote: Dear list members, i am having a somewhat odd issue (again) and i am out of ideas what is causing it. Let's assume the following very simple topology 2921 3750G-24PS All i want is to establish an OSPFv3 Adjacency between these two devices. Here is where the fun (headache) begins: The 2921 is running 15.1(4)M3 and the 3750G is running 12.2(55)SE5. So i configured OSPv3 on both sides on the respective Interfaces and wondered why the adjacency wont come up. After a little bit of troubleshooting it appears that the 3750G does not recognize the hello packets from the 2921. I can see in the debug output that they are received by the switch, but the OSPFv3 Process does not recognize them as such. If you need any additional input or show commands, please dont hesitate to ask. I appreciate all hints/comments. Thanks in advance, Peter ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] OSPFv3 Adjacency issues between 2921 and 3750
Hi Chuck, yes you have to configure a specific SDM template (and reboot the device) in order to use IPv6. I have already activated this, otherwise the 3750 does not recognize any IPv6 commands. 3750#show sdm prefer The current template is desktop IPv4 and IPv6 routing template. Thanks for your time. Regards, Peter On Tue, Jul 3, 2012 at 11:43 AM, Chuck Church chuckchu...@gmail.com wrote: Long shot, but don't the 3750s have a certain SDM they need enabled to use IPv6? Chuck -Original Message- From: cisco-nsp-boun...@puck.nether.net [mailto:cisco-nsp-boun...@puck.nether.net] On Behalf Of Peter Subnovic Sent: Tuesday, July 03, 2012 4:46 AM To: cisco-nsp@puck.nether.net Subject: Re: [c-nsp] OSPFv3 Adjacency issues between 2921 and 3750 Thanks to Pete for the pointer, i was curios why the 3750 does not respond to ff02::5 (the 2921), so i activated the debugging for ipv6 packets again and started to ping ff02::5 from the 2921. Interestingly, i can't see the icmp packets at all (well atleast thats what the debug output is telling me on the 3750). The only thing i can see are the ospf hellos from the 2921... 2921#ping ipv6 ff02::5 size 1280 Output Interface: GigabitEthernet0/1 Type escape sequence to abort. Sending 5, 1280-byte ICMP Echos to FF02::5, timeout is 2 seconds: Packet sent with a source address of FE80::1%GigabitEthernet0/1 Request 0 timed out Request 1 timed out Request 2 timed out Request 3 timed out Request 4 timed out Success rate is 0 percent (0/5) 0 multicast replies and 0 errors. And the debug output on the 3750 looks like this: Jul 3 10:41:10.424 CEST: IPV6: source FE80::1 (GigabitEthernet1/0/24) Jul 3 10:41:10.424 CEST: dest FF02::5 Jul 3 10:41:10.424 CEST: traffic class 224, flow 0x0, len 80+14, prot 89, hops 1, forward to ulp Jul 3 10:41:20.096 CEST: IPV6: source FE80::1 (GigabitEthernet1/0/24) Jul 3 10:41:20.096 CEST: dest FF02::5 Jul 3 10:41:20.096 CEST: traffic class 224, flow 0x0, len 80+14, prot 89, hops 1, forward to ulp Regards, Peter On Tue, Jul 3, 2012 at 10:05 AM, Peter Subnovic cnspmail...@googlemail.comwrote: Hi Pete, thanks for your time. The 3750 does not respond when i try to ping ff02::5. It does however respond to the all-nodes/all-router ll Multicast address. The authentication configuration is only for ospfv2. So i think it will not make any difference.I cant try it without right now. Thanks, Peter On Tue, Jul 3, 2012 at 5:27 AM, Pete Lumbis alum...@gmail.com wrote: Any difference without authentication? Does the 3750 respond to pings to ff02::5? On Mon, Jul 2, 2012 at 4:16 PM, Peter Subnovic cnspmail...@googlemail.com wrote: Dear list members, i am having a somewhat odd issue (again) and i am out of ideas what is causing it. Let's assume the following very simple topology 2921 3750G-24PS All i want is to establish an OSPFv3 Adjacency between these two devices. Here is where the fun (headache) begins: The 2921 is running 15.1(4)M3 and the 3750G is running 12.2(55)SE5. So i configured OSPv3 on both sides on the respective Interfaces and wondered why the adjacency wont come up. After a little bit of troubleshooting it appears that the 3750G does not recognize the hello packets from the 2921. I can see in the debug output that they are received by the switch, but the OSPFv3 Process does not recognize them as such. If you need any additional input or show commands, please dont hesitate to ask. I appreciate all hints/comments. Thanks in advance, Peter ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] OSPFv3 Adjacency issues between 2921 and 3750
On Tue, 2012-07-03 at 05:43 -0400, Chuck Church wrote: Long shot, but don't the 3750s have a certain SDM they need enabled to use IPv6? And on that note, what about MLS QoS on the 3750? I seem to recall having debugged something that turned out to be much (all?) IPv6 traffic ending up in queue 3 or 4, and we have set buffer sizes for these to 0. Does no mls qos make any difference? (If it's enabled of course.) -- Peter ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] OSPFv3 Adjacency issues between 2921 and 3750
Hi Peter, thank you for your suggestion. Unfortunately QoS is disabled on the 3750. 3750#show mls qos QoS is disabled QoS ip packet dscp rewrite is enabled 3750#show mls qos interface g1/0/24 GigabitEthernet1/0/24 QoS is disabled. Kind regards, Peter On Tue, Jul 3, 2012 at 1:49 PM, Peter Rathlev pe...@rathlev.dk wrote: On Tue, 2012-07-03 at 05:43 -0400, Chuck Church wrote: Long shot, but don't the 3750s have a certain SDM they need enabled to use IPv6? And on that note, what about MLS QoS on the 3750? I seem to recall having debugged something that turned out to be much (all?) IPv6 traffic ending up in queue 3 or 4, and we have set buffer sizes for these to 0. Does no mls qos make any difference? (If it's enabled of course.) -- Peter ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
[c-nsp] OSPFv3 Adjacency issues between 2921 and 3750
Dear list members, i am having a somewhat odd issue (again) and i am out of ideas what is causing it. Let's assume the following very simple topology 2921 3750G-24PS All i want is to establish an OSPFv3 Adjacency between these two devices. Here is where the fun (headache) begins: The 2921 is running 15.1(4)M3 and the 3750G is running 12.2(55)SE5. So i configured OSPv3 on both sides on the respective Interfaces and wondered why the adjacency wont come up. After a little bit of troubleshooting it appears that the 3750G does not recognize the hello packets from the 2921. I can see in the debug output that they are received by the switch, but the OSPFv3 Process does not recognize them as such. Please see the debug output below: I had debugging for IPv6 Packets and OSPFv3 Hellos enabled. FE80::1 = 2921, FE80::2 = 3750G Jun 30 17:00:38.347 CEST: IPV6: source FE80::1 (GigabitEthernet1/0/24) Jun 30 17:00:38.347 CEST: dest FF02::5 Jun 30 17:00:38.347 CEST: traffic class 224, flow 0x0, len 80+14, prot 89, hops 1, forward to ulp Jun 30 17:00:41.057 CEST: OSPFv3: Send hello to FF02::5 area 0 on GigabitEthernet1/0/24 from FE80::2 interface ID 1040 Jun 30 17:00:41.057 CEST: IPV6: source FE80::2 (local) Jun 30 17:00:41.057 CEST: dest FF02::5 (GigabitEthernet1/0/24) Jun 30 17:00:41.057 CEST: traffic class 224, flow 0x0, len 76+0, prot 89, hops 1, originating Jun 30 17:00:41.057 CEST: IPv6-Fwd: Sending on GigabitEthernet1/0/24 As you can see, the hello packet arrives on the interface, but somehow the OSPFv3 process on the 3750 does not recognize them as such. After digging around i found the following bug and thought it could be related: CSCtr55645 IPv6 multicast packets for routing protocols (such as OSPFv3 hello messages) are not enqueued to CPU queue 3 as expected. So i upgraded to 12.2(55)SE5 (where this bug should be fixed) this weekend in the hope to resolve issue, but still the same. I am really kinda lost here. The 2921 does see the hello packets from the 3750 and the adjacency is in the INIT State as i would have it expected. show ipv6 ospf neighbor Neighbor ID Pri State Dead Time Interface IDInterface 1.2.3.4 0 INIT/ -00:00:361040 GigabitEthernet0/1 So has anybody seen this strange behavior or has an idea what is causing the issues? I rebooted the 3750, cleared the ospf processes on both sides, no luck I dont wanna rule out that i maybe just missed something. Please see below the relevant configuration of the respective Interfaces: 2921: interface GigabitEthernet0/1 description ===XXX=== ip address XXX no ip redirects no ip proxy-arp ip flow ingress ip flow egress ip ospf authentication message-digest ip ospf message-digest-key 1 md5 7 ip ospf network point-to-point ip ospf 1 area 0 load-interval 30 duplex auto speed auto ipv6 address FE80::1 link-local ipv6 enable ipv6 nd ra suppress all no ipv6 redirects ipv6 ospf network point-to-point ipv6 ospf 1 area 0 no cdp enable no mop enabled ipv6 router ospf 1 router-id 1.1.1.1 === 3750G interface GigabitEthernet1/0/24 description ===xxx=== no switchport ip address xxx 255.255.255.252 no ip proxy-arp no ip redirects ip ospf authentication message-digest ip ospf message-digest-key 1 md5 7 xxx ip ospf network point-to-point ip ospf 1 area 0 no keepalive ipv6 address FE80::2 link-local ipv6 enable ipv6 nd ra suppress no ipv6 redirects ipv6 ospf network point-to-point ipv6 ospf 1 area 0 no cdp enable ipv6 router ospf 1 router-id 1.2.3.4 log-adjacency-changes If you need any additional input or show commands, please dont hesitate to ask. I appreciate all hints/comments. Thanks in advance, Peter ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] OSPFv3 Adjacency issues between 2921 and 3750
Hi Mikael, thanks for the input. Unfortunately i can not set the IPv6 MTU on the 3750. According to the show ipv6 int xx output, both sides have an mtu of 1500 3750: === GigabitEthernet1/0/24 is up, line protocol is up IPv6 is enabled, link-local address is FE80::2 No Virtual link-local address(es): No global unicast address is configured Joined group address(es): FF02::1 FF02::2 FF02::5 FF02::1:FF00:2 MTU is 1500 bytes 2921 GigabitEthernet0/1 is up, line protocol is up IPv6 is enabled, link-local address is FE80::1 No Virtual link-local address(es): No global unicast address is configured Joined group address(es): FF02::1 FF02::2 FF02::5 FF02::6 FF02::1:FF00:1 MTU is 1500 bytes Thanks for your time. Kind regards, Peter On Mon, Jul 2, 2012 at 10:33 PM, Mikael Abrahamsson swm...@swm.pp.sewrote: On Mon, 2 Jul 2012, Peter Subnovic wrote: If you need any additional input or show commands, please dont hesitate to ask. MTU issue? Try setting ipv6 mtu 1500 at both ends and see if that changes anything? -- Mikael Abrahamssonemail: swm...@swm.pp.se ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] OSPFv3 Adjacency issues between 2921 and 3750
Any difference without authentication? Does the 3750 respond to pings to ff02::5? On Mon, Jul 2, 2012 at 4:16 PM, Peter Subnovic cnspmail...@googlemail.com wrote: Dear list members, i am having a somewhat odd issue (again) and i am out of ideas what is causing it. Let's assume the following very simple topology 2921 3750G-24PS All i want is to establish an OSPFv3 Adjacency between these two devices. Here is where the fun (headache) begins: The 2921 is running 15.1(4)M3 and the 3750G is running 12.2(55)SE5. So i configured OSPv3 on both sides on the respective Interfaces and wondered why the adjacency wont come up. After a little bit of troubleshooting it appears that the 3750G does not recognize the hello packets from the 2921. I can see in the debug output that they are received by the switch, but the OSPFv3 Process does not recognize them as such. Please see the debug output below: I had debugging for IPv6 Packets and OSPFv3 Hellos enabled. FE80::1 = 2921, FE80::2 = 3750G Jun 30 17:00:38.347 CEST: IPV6: source FE80::1 (GigabitEthernet1/0/24) Jun 30 17:00:38.347 CEST: dest FF02::5 Jun 30 17:00:38.347 CEST: traffic class 224, flow 0x0, len 80+14, prot 89, hops 1, forward to ulp Jun 30 17:00:41.057 CEST: OSPFv3: Send hello to FF02::5 area 0 on GigabitEthernet1/0/24 from FE80::2 interface ID 1040 Jun 30 17:00:41.057 CEST: IPV6: source FE80::2 (local) Jun 30 17:00:41.057 CEST: dest FF02::5 (GigabitEthernet1/0/24) Jun 30 17:00:41.057 CEST: traffic class 224, flow 0x0, len 76+0, prot 89, hops 1, originating Jun 30 17:00:41.057 CEST: IPv6-Fwd: Sending on GigabitEthernet1/0/24 As you can see, the hello packet arrives on the interface, but somehow the OSPFv3 process on the 3750 does not recognize them as such. After digging around i found the following bug and thought it could be related: CSCtr55645 IPv6 multicast packets for routing protocols (such as OSPFv3 hello messages) are not enqueued to CPU queue 3 as expected. So i upgraded to 12.2(55)SE5 (where this bug should be fixed) this weekend in the hope to resolve issue, but still the same. I am really kinda lost here. The 2921 does see the hello packets from the 3750 and the adjacency is in the INIT State as i would have it expected. show ipv6 ospf neighbor Neighbor ID Pri State Dead Time Interface IDInterface 1.2.3.4 0 INIT/ -00:00:361040 GigabitEthernet0/1 So has anybody seen this strange behavior or has an idea what is causing the issues? I rebooted the 3750, cleared the ospf processes on both sides, no luck I dont wanna rule out that i maybe just missed something. Please see below the relevant configuration of the respective Interfaces: 2921: interface GigabitEthernet0/1 description ===XXX=== ip address XXX no ip redirects no ip proxy-arp ip flow ingress ip flow egress ip ospf authentication message-digest ip ospf message-digest-key 1 md5 7 ip ospf network point-to-point ip ospf 1 area 0 load-interval 30 duplex auto speed auto ipv6 address FE80::1 link-local ipv6 enable ipv6 nd ra suppress all no ipv6 redirects ipv6 ospf network point-to-point ipv6 ospf 1 area 0 no cdp enable no mop enabled ipv6 router ospf 1 router-id 1.1.1.1 === 3750G interface GigabitEthernet1/0/24 description ===xxx=== no switchport ip address xxx 255.255.255.252 no ip proxy-arp no ip redirects ip ospf authentication message-digest ip ospf message-digest-key 1 md5 7 xxx ip ospf network point-to-point ip ospf 1 area 0 no keepalive ipv6 address FE80::2 link-local ipv6 enable ipv6 nd ra suppress no ipv6 redirects ipv6 ospf network point-to-point ipv6 ospf 1 area 0 no cdp enable ipv6 router ospf 1 router-id 1.2.3.4 log-adjacency-changes If you need any additional input or show commands, please dont hesitate to ask. I appreciate all hints/comments. Thanks in advance, Peter ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
