On 20 Apr 2013, at 17:09, Paul Wouters <[email protected]> wrote:

> Imagine your previous DSL router was doing DNSSEC with the previous root key. 
> Your current DSL modem dies, and you power on your old one. What will happen?

You go to the nearest shop and spend $20-30 on new CPE because the old one(s) 
are broken. :-)

> Now repeat, and say that the SHA2 family was broken and SHA4 is now the
> standard. Your old modem talks SHA2 only.
> 
> I don't think we as a group have ignored this problem - we have just not
> found a proper solution yet.

I am not sure there needs to be a proper solution. Or that a one size fits all 
solution is even possible or desirable. This topic is largely a matter of 
implementation and/or local policy after all. Some generalised motherhood and 
apple pie guidelines/advice would probably be enough: don't hard-wire the 
current root KSK as a trust anchor; always design your stuff with the 
assumption the root KSK must change from time to time; provide some out of band 
means for your stuff to recover from a missed key rollover; consider designing 
your firmware to handle changes to the root KSK in a similar way to how it 
deals with a regular software update, etc. Maybe a document along these lines 
is needed?

We can be fairly certain that the URLs IANA publishes for the root TA are not 
going to change. Though perhaps that needs to be explicitly stated by someone. 
So it should be a simple matter of programming to provide some way to slurp 
that stuff, check it and update the locally held TA. Something comparable to 
unbound-anchor could do the job. Suppliers of CPE -- vendor, ISP -- might even 
provide their own copy of that repository in whatever format they choose so 
their stuff can contact the mothership at regular intervals or after a hard 
reset. For bonus points that update mechanism could use a secure out of band 
channel by (say) embedding a cert or public key tied to the IP address of the 
supplier's repository into the firmware.

_______________________________________________
DNSOP mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dnsop

Reply via email to