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

[Fernando Gont <[e-mail]>]

Hi, Michael,

On 11/04/2013 04:08 PM, Mikael Abrahamsson wrote:
> On Mon, 4 Nov 2013, Fernando Gont wrote:
> 
>> Isn't there a v6ops I-D with similar wording?
> 
> I opposed to this wording in there as well.

Then I apologize -- I kind of assumed some sort of intention based on
that I-D. So.. point taken.

That said, the meat is in stats. :-)



> If one deploys a 6500/7600/SUP720 and configures IPv6 on it today, it'll
> slow-path every IPv6 packet with a fragmentation header. Default behaviour.

But slow path != drop.



> I have no idea how much of the behaviour your're seeing is caused by
> this default behaviour, but I'd venture to guess a lot.
> 
>> mm... as far as discussion on a number of forums went, apparently
>> there's intent to do so.
> 
> Oki, I haven't seen these.
> 
>> Data? References?
> 
> http://seclists.org/nanog/2011/Sep/1076

Quickly skimming through the discussion it looks like a decision is made
to drop -- albeit for reasonable reasons. Am I missing something?

Thanks,
-- 
Fernando Gont
e-mail: [e-mail] || [e-mail]
PGP Fingerprint: 7809 84F5 322E 45C7 F1C9 3945 96EE A9EF D076 FFF1