Re: [v6ops] Operational Implications of IPv6 Packets with Extension Headers - implications from new development for EHs

[Brian E Carpenter <[e-mail]>]

And it may be another unpopular fact of life, but I believe that RFC8799 is relevant. Certainly some basic rethinking about interoperability is needed.

Regards
   Brian

On 30-Jul-20 14:25, Fernando Gont wrote:
> Hello, Joe,
> 
> On 29/7/20 17:16, Joseph Touch wrote:
> [...]
>>>> If we are merely documenting what happens to be implemented, we cease to be a standards body and become merely reporters.
>>>
>>> If we avoid any introspection or consideration of operational reality, the cease to be a relevant standards body and become an ivory tower.\
>>
>> That’s why I said “merely”. Doing both and appreciating the balance is fine - the point is that “what is implemented/able TODAY” is NOT the only consideration.
> 
> FWIW, I don't think we should limit ourselves to documenting the 
> problem.Indeed, documenting the problem can certainly be a starting 
> point to consider possible ways to mitigate it.
> 
> For a long time, the status quo was assuming that EHs work. More 
> recently, thanks to a number of efforts (Geoff's measurements, what we 
> did in RFC7872, and others), there has been increased awareness about 
> the packet drops.
> 
> I would expect that a common understanding that there are underlying 
> issues that lead to the packet drops (and it's not just folks playing 
> with "firewall" rules at random) can serve as a starting point to 
> consider what can be done to make things better, closing the gap (to the 
> extent that is possible) between what the IETF says IPv6 is, and the 
> operational reality of it.
> 
> Thanks,
>