Re: [v6ops] Some stats on IPv6 fragments and EH filtering on the Internet

[Ole Troan <[e-mail]>]

Fernando,

>>> In any case, the interesting (and unfortunate) data if the
>>> chances of success when you use extension headers or
>>> fragmentation are in the scale of "unlikely". :-(
>> 
>> I'm not sure you can draw that conclusion without knowing where
>> the fragments are dropped.
>> 
>> e.g. you are not saying that fragmented packets will be dropped 
>> anywhere on the link between your home and mine, are you?
> 
> I'm certainly not. All tests were against web servers.
> 
> 
>> I'm for example not concerned about a web server or load balancer 
>> that sets TCP MSS to 1220 and then drop fragments.
> 
> Certainly, there's much more testing to be done (this is even stated in
> the slideware :-) ). That said, you might still be concerned about the
> case you meantion -- at least in theory, that case might arise in a
> NAT64 (?) case.
> 
> Also, one of the main points of this slideware is that its not just
> fragments that are dropped, but extension headers in general.
> 
> In any case, my goal of sharing the results is to trigger discussion
> and encourage further testing, rather than coming up with scary or
> bold statements.

splendid.

it would be very interesting to know where filtering is done.
my assumption is that it isn't done in what I call the Internet infrastructure,
but close to the hosts (edge), or in the hosts themselves.

pretty tricky to measure that from afar though.

cheers,
Ole