[42338] in SIPB IPv6
Re: wikipedia unreachable
daemon@ATHENA.MIT.EDU (Quentin Smith)
Sat May 24 18:07:30 2014
Date: Sat, 24 May 2014 18:07:25 -0400 (EDT)
From: Quentin Smith <quentin@MIT.EDU>
To: sipbv6@mit.edu
In-Reply-To: <alpine.DEB.2.02.1405221959550.22250@team-rocket.mit.edu>
I shutdown our peering with OCCAID, using
vtysh
configure terminal
router bgp 3
neighbor 2001:4830:e1:8::1 shutdown
and observed that wikipedia was once again reachable.
Most interestingly, en.wikipedia.org *is* reachable from limekiller's
OCCAID tunnel address, 2001:4830:e1:8::2, but is not from any other
address, including when I force traceroute to use 2001:4830:2446:b5::1 as
the source.
This makes me think that there is something wrong with the return route
for our subnets. I'm not sure how to debug this further.
--Quentin
On Thu, 22 May 2014, Quentin Smith wrote:
> jhawk reports en.wikipedia.org is unreachable, and I can confirm:
>
> [quentin@whole-enchilada]:~/web_scripts$ traceroute6 en.wikipedia.org
> traceroute to en.wikipedia.org (2620:0:861:ed1a::1), 30 hops max, 80 byte
> packets
> 1 limekiller.ipv6.mit.edu (2001:4830:2446:b5::1) 0.247 ms 0.269 ms 0.239
> ms
> 2 dcr01-tu100.bstn01.occaid.net (2001:4830:e1:8::1) 1.024 ms 0.995 ms
> 0.961 ms
> 3 bbr01-g0-1.bstn01.occaid.net (2001:4830:ff:b100::1) 1.340 ms 1.309 ms
> 1.279 ms
> 4 * * *
> 5 * * *
> 6 * * *
> 7 * * *
> 8 * * *
> 9 * * *
> 10 * * *
> 11 * * *
> 12 * * *
> 13 * * *
> 14 * * *
> 15 * * *
> 16 * * *
> 17 * * *
> 18 * * *
>
>
> Interestingly, www.google.com also shows bizarre traceroute behavior:
>
> [quentin@whole-enchilada]:~/web_scripts$ traceroute6 www.google.com
> traceroute to www.google.com (2607:f8b0:4009:807::1011), 30 hops max, 80 byte
> packets
> 1 limekiller.ipv6.mit.edu (2001:4830:2446:b5::1) 0.184 ms 0.670 ms 0.656
> ms
> 2 dcr01-tu100.bstn01.occaid.net (2001:4830:e1:8::1) 1.005 ms 0.996 ms
> 1.008 ms
> 3 bbr01-g0-1.bstn01.occaid.net (2001:4830:ff:b100::1) 1.280 ms 1.267 ms
> 1.258 ms
> 4 * * *
> 5 * * *
> 6 * * *
> 7 * * *
> 8 * * *
> 9 * * *
> 10 2001:4860::8:0:5900 (2001:4860::8:0:5900) 35.844 ms 35.843 ms 35.775 ms
> 11 2001:4860::1:0:92e (2001:4860::1:0:92e) 36.235 ms 35.437 ms 36.123 ms
> 12 2001:4860:0:1::7c1 (2001:4860:0:1::7c1) 36.620 ms 36.046 ms 36.557 ms
> 13 2607:f8b0:8000:40::b (2607:f8b0:8000:40::b) 35.336 ms
> 2607:f8b0:8000:40::6 (2607:f8b0:8000:40::6) 35.454 ms 2607:f8b0:8000:40::b
> (2607:f8b0:8000:40::b) 36.094 ms
>
> This makes me think that occaid is broken; should we disable occaid pending
> resolution of this? I don't have a trusted device on me right now.
>
> --Quentin
>
