Maybe it's time to talk about rational routing

Brian E Carpenter brian.e.carpenter at gmail.com
Tue Sep 6 23:48:58 CEST 2011


Hi Ignatios,

> Something is strange with UDP filtering at UofAuck - or its upstream(s).

I think that's the KAREN network (our NREN) but it's possible that
my friends across the road at IT Services are also blocking pings on campus.

If you have any way to check on ICMPv6 PTB transparency that would
be nice. www.cs.auckland.ac.nz is dual stacked.

Thanks
   Brian Carpenter

On 2011-09-06 21:09, Ignatios Souvatzis wrote:
> On Tue, Sep 06, 2011 at 01:34:09PM +1200, Brian E Carpenter wrote:
>> Good news down under, it seems that TelstraClear is now supporting
>> native IPv6, announced in Australia but observable in New Zealand too,
>> with much less HE tunnel than before.
>>
>> However, it seems to be time for a discussion about routing policy.
>> The traceroute below reminds me of IPv4 around 1990. Is it really
>> optimal to route from Auckland to Brisbane via Hong Kong and Los
>> Angeles? Of course there are all the usual commercial constraints
>> on policy, but this seems to be worse than that.
>>
>> C:\Documents and Settings\bcar071>tracert www.apnic.net
>>
>> Tracing route to www.apnic.net [2001:dc0:2001:11::211]
>> over a maximum of 30 hops:
>>
>>   1    <1 ms    <1 ms    <1 ms  2001:df0:0:2006::254     % Uni Auckland
> 
> outside view:
> 
> from Uni-Bonn I reach apnic via
> 
> dfn fra - level3 fra/lndn01/nwrk01/asbn01/atln01/lsan01/nsyd
> 
> and Uni Auckland via
> 
> dfn fra - geant2 fra/wash - internet2 atla/hous/losa - 2401:138:2000 -
> 2404:138:0 - then only stars.
> 
> 
>>From home I reach apnic via
> 
> rh-tec fra - HE fra/lon/nyc/lax - reach.com - 2401:38000 - 2001:4403:: - stars
> 
> and U auckland via
> 
> rh-tec fra - HE fra/lon/nyc/lax - 2001:504:13 - nsw.vocusconnect - 
> VOCUS syd - 2402:7800:: - brisbane.pipenetworks - 2001:dc0:2001:11::211
> 
> Something is strange with UDP filtering at UofAuck - or its upstream(s).
> 
> 	-is
> 
> 



More information about the ipv6-ops mailing list