[ipv6hackers] Some stats on IPv6 fragments and EH filtering on the Internet

[Fernando Gont]

Hi, Marc,

On 11/04/2013 04:55 PM, Marc Blanchet wrote:
> Le 2013-11-04 à 16:50, Fernando Gont <fernando at gont.com.ar> a écrit
> :
> 
>> On 11/04/2013 04:10 PM, Marc Blanchet wrote:
>>> Fernando, clarification question (since I was not at IEPG):
>>> define "failure" in your slides?
>> 
>> Probe packets do not get to the intended destination.
> 
> ok. your presentation says that you used a single vantage point.  Do
> you know where the filtering happened? 

Not exactly


> Is it possible that it could
> be "not far from you", i.e. maybe your upstream provider was
> filtering on some routers to some destination, and not filtering for
> some other routers for some other destinations (i.e. different
> peers).  That might not be the case, I'm just trying to get a better
> understanding of your results.

Certainly more testing (such as identifying who's dropping the packets)
would be useful.

My guess is that' the filtering is not really happening close to the
vantage point.




>>> It might look a dumb question, but I just want to make sure. Is
>>> failure that you don't get back a reply packet? or you get/don't
>>> get ICMP parameter problem or ?
>> 
>> I didn't check whether an ICMPv6 error comes back -- but I'll do.
> 
> That would be very helpful to me. Because I consider differently the
> failures depending on that data.

mm... different in what respect?

Certainly, receiving some sort of error message is useful (e.g. see RFC
5461)... but.. other than that, a packet drop is a packet drop.

Thanks,
-- 
Fernando Gont
SI6 Networks
e-mail: fgont at si6networks.com
PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492