Strange source port filtering by Tinet

Alexander Gall gall at switch.ch
Mon Jul 26 18:04:05 CEST 2010


On Mon, 26 Jul 2010 17:09:20 +0200, Daniel Verlouw <daniel at bit.nl> said:

> Hi,
> On Tue, 2010-07-20 at 14:24 +0200, Alexander Gall wrote:
>> It turned out that Tinet already
>> opened a ticket after my initial report last week but that didn't get
>> through to me.  Anyway, a possible fix will be deployed tomorrow
>> (hopefully).

> seems this is fixed now on their Frankfurt peering router:

> daniel at daniel:~$ traceroute6 -s 2001:7b8:3:1000:202:b3ff:fe1f:2a28
> --sport=35072 -q1 -N1 -w1 ch1.dnsnode.net
> traceroute to ch1.dnsnode.net (2001:67c:1010:2::53), 30 hops max, 80
> byte packets
>  1  fw1.office.bit.nl (2001:7b8:3:1000::1)  0.367 ms
>  2  fw1-gw.jun1.bit-2a.network.bit.nl (2001:7b8:3:1c::3)  1.172 ms
>  3  803.ge-0-1-0.jun1.fra4.network.bit.nl (2001:7b8:0:323::2)  5.742 ms
>  4  de-cix.fra20.ip6.tiscali.net (2001:7f8::cb9:0:1)  19.697 ms
>  5  xe-5-3-0.lon20.ip6.tinet.net (2001:668:0:2::1:1232)  35.260 ms
>  6  xe-1-0-0.lon21.ip6.tinet.net (2001:668:0:2::1:142)  23.470 ms
>  7  xe-5-3-0.lon10.ip6.tinet.net (2001:668:0:2::1:1642)  22.772 ms
>  8  g0-0-123.tr2.tfm7.thn.linx.net (2001:668:0:3::4000:82)  20.262 ms
>  9  2a01:40:1003:2::3 (2a01:40:1003:2::3)  22.757 ms
> 10  ch1.dnsnode.net (2001:67c:1010:2::53)  23.124 ms

It has been fixed this afternoon, or, rather, they verified that a
suggested fix to a Juniper bug works.

> but I'm still seeing issues through their London POP (which is fairly
> odd because we connect directly to the same router as seen in hop 7
> above):

> daniel at daniel:~$ traceroute6 -s 2001:7b8:3:1000:202:b3ff:fe1f:2a28
> --sport=35072 -q1 -N1 -w1 ch1.dnsnode.net
> traceroute to ch1.dnsnode.net (2001:67c:1010:2::53), 30 hops max, 80
> byte packets
>  1  fw1.office.bit.nl (2001:7b8:3:1000::1)  0.290 ms
>  2  fw1-gw.jun1.bit-2a.network.bit.nl (2001:7b8:3:1c::3)  0.402 ms
>  3  806.ge-0-1-0.jun1.thn.network.bit.nl (2001:7b8:0:326::2)  22.418 ms
>  4  *
>  5  *
>  6  *
>  7  *^C

> Anyone knows if they're still working on this?

Yes, the fix has probably not yet been rolled out everywhere.

-- 
Alex




More information about the ipv6-ops mailing list