IPv6 big packet blackhole in Teleglobe/TATA backbone?

[Pekka Savola]

FYI, the issue has been resolved, with the help of a contact provided 
off-list.

>From the last mail, "... MTU applied at the wrong layer of the 
config."

On Tue, 19 Aug 2008, Ross West wrote:
>> I and Remi have been trying to figure out an IPv6 big packet blackhole
>> that seems to be occurring in Teleglobe network, specifically around
>> IP addresses 2001:5a0:200::5 and 2001:5a0:200::10.  I suspect MTU is
>> not properly configured.
>
> Here's what I see (I'm a Tata customer in Canada, via tunnel to NY/USA)
>
> Got this weird one once:
>
> -=
> [westr at faraday ~]$ traceroute6 -v www.remlab.net
> traceroute6 to poy.remlab.net (2001:41d0:1:a0d6::401:1983) from 2610:38::2, 64 hops max, 12 byte packets
> 1  2610:38::1 68 bytes of data to 2610:38::2  0.490 ms  0.304 ms  0.293 ms
> 2  tunnel13.6bb1.nyy-newyork.ipv6.teleglobe.net 68 bytes of data to 2610:38::2  20.039 ms  19.984 ms  19.725 ms
> 3  2001:5a0:400:300::1 68 bytes of data to 2610:38::2  19.995 ms  19.903 ms  19.934 ms
> 4  2001:5a0:0:300::2d 68 bytes of data to 2610:38::2  96.759 ms  97.458 ms  96.809 ms
> 5  2001:5a0:1900::d 68 bytes of data to 2610:38::2  96.947 ms  97.003 ms  96.882 ms
> 6  2001:5a0:1900::a 68 bytes of data to 2610:38::2  97.347 ms  97.042 ms  97.104 ms
> 7  2001:41d0::591 68 bytes of data to 2610:38::2  100.780 ms
> 157 bytes from 2001:41d0:1:4a86::1 to 2610:38::2: icmp type 1 (Destination Unreachable) code 4
> 0000: 60000000 006d1139 26100038 00000000
> 0010: 00000000 00000002 200141d0 00014a86
> 0020: 00000000 00000001 c0000035 006d1c87
> 0030: b4580000 00010000 00000001 01310139
> 0040: 01350130 01300130 01300130 01300130
> 0050: 01300130 01300130 01300130 01300130
> 0060: 01300130 01300130 01300130 01300164
> 0070: 01310134 01310130 01300132 03697036
> 0080: 04617270 6100000c 00010000 29100000
> 0090: 00800000 00000000 00000000 00
>  100.811 ms  101.340 ms
> 8  2001:41d0:1:a0d6::401:1983 68 bytes of data to 2610:38::2  101.103 ms  100.715 ms  100.754 ms
> -=
>
> Not sure if hop #7 output helps, it seems to be a one off - here's what I normally get:
>
> -=
> [westr at faraday ~]$ traceroute6 -v www.remlab.net
> traceroute6 to poy.remlab.net (2001:41d0:1:a0d6::401:1983) from 2610:38::2, 64 hops max, 12 byte packets
> 1  2610:38::1 68 bytes of data to 2610:38::2  0.519 ms  0.327 ms  0.303 ms
> 2  tunnel13.6bb1.nyy-newyork.ipv6.teleglobe.net 68 bytes of data to 2610:38::2  19.914 ms  19.797 ms  19.803 ms
> 3  2001:5a0:400:300::1 68 bytes of data to 2610:38::2  19.960 ms  19.865 ms  19.931 ms
> 4  2001:5a0:0:300::2d 68 bytes of data to 2610:38::2  96.978 ms  96.758 ms  96.934 ms
> 5  2001:5a0:1900::d 68 bytes of data to 2610:38::2  96.970 ms  96.993 ms  96.885 ms
> 6  2001:5a0:1900::a 68 bytes of data to 2610:38::2  97.142 ms  97.059 ms  97.030 ms
> 7  2001:41d0::591 68 bytes of data to 2610:38::2  100.777 ms  101.015 ms  100.654 ms
> 8  2001:41d0:1:a0d6::401:1983 68 bytes of data to 2610:38::2  100.761 ms  100.674 ms  100.622 ms
> -=
>
> Pings/icmp of any size go through normally.
>
> Cheers,
>  Ross.
>
>

-- 
Pekka Savola                 "You each name yourselves king, yet the
Netcore Oy                    kingdom bleeds."
Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings