Re: [IPsec] Proposed method to achieve quantum resistant IKEv2

["Valery Smyslov" <[e-mail]>]

Hi Paul,

> >>    Adding IKE-level fragmentation to the process adds an additional place
> >>    that DDoS attacks can hit.
> >
> > We have DDoS protection mechanisms. I think it's possible to define
> > IKE_SA_INIT fragmentation so that these mechanisms be still able to work.
> 
> That would be tricky. 

Sure. But not impossible, I think.

> Either a new exchange or an untrusted stream of
> fragments. Either way, a lot of complexity for a rather moving target
> goal that we don't understand yet. I'd personally rather wait until we
> know a bit more about the direction that quantum safe protocols are
> really going to.

Fully agree. And the size of the quantum safe KE payload is really important for IKE.
When I hear that it could be several MB, it makes me nervous - it seems
that transferring so much unauthenticated data will become a good target for DoS attacks.
Several KB is a bit better, at least it seems that we can deal with this amount.

> >>    So, one would have to have some mechanism to know what would get through,
> >>    and if to switch to TCP, etc. before even trying.
> >
> > TCP has a lot of shortcomings. It's a last resort.
> 
> Agreed. It's the emergency backup parachute, not the real parachute.
> 
> People are suggesting a lot of complicated bells and whistles. I wish we
> would not compete with TLS on which can be more complicated :P

Haven't we already won? :-)

> Paul

Regards,
Valery.