Cogent to Comcast IPv6 broken?

TheIpv6guy . cb.list6 at gmail.com
Fri Apr 18 20:06:51 UTC 2014


On Fri, Apr 18, 2014 at 11:12 AM, Will Orton <will at loopfree.net> wrote:
> Cogent from various west coast USA locations seems to have trouble getting to some Comcast ipv6 addresses:
>
> Cogent LG Los Angeles to www.comcast6.net
> traceroute to 2001:558:fe16:7:69:252:216:215 (2001:558:fe16:7:69:252:216:215), 30 hops max, 80 byte packets
>  1  2001:550:1:317::1 (2001:550:1:317::1)  70.184 ms  70.181 ms
>  2  * *
>  3  * *
>  4  2001:550:4::51 (2001:550:4::51)  15.013 ms  14.981 ms
>  5  * *
>  6  * *
> [ ... ]
>
>
> Cogent LG San Francisco ipv6 trace to www.comcast6.net
>  traceroute to 2001:558:fe16:7:69:252:216:215 (2001:558:fe16:7:69:252:216:215), 30 hops max, 80 byte packets
>  1  2001:550:1:31f::1 (2001:550:1:31f::1)  0.367 ms  0.369 ms
>  2  * *
>  3  * *
>  4  * *
> [ ... ]
>
> Cogent LG San Jose ipv6 ping to www.comcast6.net
> PING 2001:558:fe16:7:69:252:216:215(2001:558:fe16:7:69:252:216:215) 56 data bytes
>
> --- 2001:558:fe16:7:69:252:216:215 ping statistics ---
> 5 packets transmitted, 0 received, 100% packet loss, time 13999ms
>
>
> Cogent LG San Jose ipv6 ping to speedtest.comcast.net (Los Angeles does same thing, but Kansas City works):
> PING 2001:558:fe2a:3:69:241:64:6(2001:558:fe2a:3:69:241:64:6) 56 data bytes
>
> --- 2001:558:fe2a:3:69:241:64:6 ping statistics ---
> 5 packets transmitted, 0 received, 100% packet loss, time 14000ms
>
> from Los Angeles same thing, but from Cogent's LG Kansas City it works:
> PING 2001:558:fe2a:3:69:241:64:6(2001:558:fe2a:3:69:241:64:6) 56 data bytes
> 64 bytes from 2001:558:fe2a:3:69:241:64:6: icmp_seq=1 ttl=55 time=35.6 ms
> 64 bytes from 2001:558:fe2a:3:69:241:64:6: icmp_seq=2 ttl=55 time=35.7 ms
> 64 bytes from 2001:558:fe2a:3:69:241:64:6: icmp_seq=3 ttl=55 time=35.6 ms
> 64 bytes from 2001:558:fe2a:3:69:241:64:6: icmp_seq=4 ttl=55 time=35.6 ms
> 64 bytes from 2001:558:fe2a:3:69:241:64:6: icmp_seq=5 ttl=55 time=35.6 ms
>
> --- 2001:558:fe2a:3:69:241:64:6 ping statistics ---
> 5 packets transmitted, 5 received, 0% packet loss, time 4043ms
> rtt min/avg/max/mdev = 35.620/35.676/35.734/0.241 ms
>
>
> This is similar to issues I see with a Cogent IPv6/BGP-enabled transit connection in Los Angeles to a random
> Comcast customer address (that is otherwise up and reachable from HE.net, NTT, etc):
>
> (from router with ipv6 BGP Cogent transit connection):
> traceroute6 to 2601:7:1300:3d2::1 (2601:7:1300:3d2::1) from 2001:550:2:18::5:2, 64 hops max, 12 byte packets
>  1  2001:550:2:18::5:1 (2001:550:2:18::5:1)  1.903 ms  1.120 ms  10.771 ms
>  2  * * *
>  3  * 2001:550::57 (2001:550::57)  15.654 ms *
>      MPLS Label=19852 CoS=2 TTL=255 S=0
>      MPLS Label=19207 CoS=2 TTL=255 S=1
>  4  * * *
>
>
> I have Cogent ticket #HD0000005706297  on that but I'm being ignored.
> I see the similar things from Cogent IP6 transit out of San Jose.
>
> Anyone at Cogent able to look at this? I basically blackhole big chunks of Comcast ipv6 on my network when I
> enable ipv6 to Cogent.
>
>
> -Will
>

This is probably a better discussion for [email protected] ... but ... i notice
that i could NOT get to anything from IPv6 on Comcast at home this
morning.

And quick look at their comcast route servers shows things are NOT ok

route-server.newyork.ny.ibone>ping www.yahoo.com
Translating "www.yahoo.com"...domain server (68.87.64.164) [OK]

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 2001:4998:F00B:1FE::3001, timeout is
2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 0/0/4 ms
route-server.newyork.ny.ibone>ping www.facebook.com
Translating "www.facebook.com"...domain server (68.87.64.164) [OK]

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 2A03:2880:2130:7F07:FACE:B00C:0:1,
timeout is 2 seconds:
.....
Success rate is 0 percent (0/5)
route-server.newyork.ny.ibone>ping www.google.com
Translating "www.google.com"...domain server (68.87.64.164) [OK]

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 2607:F8B0:400C:C04::6A, timeout is 2 seconds:
.....
Success rate is 0 percent (0/5)




More information about the NANOG mailing list