Re: [dhcwg] Lifetime draft: refresh time should never be more thanIRT_DEFAULT

On Nov 16, 2004, at 7:50 AM, Joe Quanaim wrote:
> A rogue
> dhcpv6 server could do much worse than to set the lifetime option to an
> excessive value.

Or it could do both, for a triple word score.   This is my concern - 
clients that don't restrict the length of the lifetime option simply 
never recover from an attack like this, where as clients that do 
restrict it do recover.

> And if a client does not implement the lifetime option
> (which will probably be true for the near term), it is still 
> vulnerable to
> this attack whatever we decide the maximum value should be.

I think such clients are unlikely to be an issue in practice at this 
stage of deployment.

I think that it should be sufficient to say that clients SHOULD 
restrict the lifetime to a day, and explain why, and let the 
implementor decide how to handle it.   Or maybe just say "clients MAY 
limit the lifetime", not specify any particular limit, and explain the 
tradeoffs.

_______________________________________________
dhcwg mailing list
dhcwg@ietf.org
https://www1.ietf.org/mailman/listinfo/dhcwg