> On Jul 31, 2018, at 5:44 AM, Philip Homburg <pch-dnso...@u-1.phicoh.com> 
> wrote:
> 
> I wonder if there still is a use case for distributing the root zone. With
> QNAME minimization and NXDOMAIN based on NSEC records, the major use cases
> seem to be gone. Compared to other zones, the root is massively over
> provisioned. So if (from an availability point of view) there is need to have
> a local copy of the root, then you would need a local copy of .com as well.

A local copy of the root zone improves availability in and of itself because of 
its importance as the starting point of all resolution. While the root zone is 
indeed massively over provisioned, the bad guys will always be able to send 
more traffic: that's an un-winnable arms race. And over provisioning doesn't 
help me if reachability is poor from my particular vantage point. Availability 
will therefore always be a concern.

Sure, a local copy of .com would (further) improve availability, but that's 
entirely impractical given the zone's size and rate of change. We're fortunate 
that the critically important root zone is small enough and changes 
infrequently enough that having a local copy is a realistic option.

I don't think we should assume only (or even primarily) AXFR for root zone 
distribution on a massive scale. Building a scalable infrastructure for that is 
a significant expense that I don't think is necessary (for the root operators 
or anyone else) when distributing 2MB files is a problem that's been solved 
other ways many times over. Why not distribute the root zone via, for example, 
multiple CDNs?

To verify the integrity of the downloaded zone one could validate all the 
RRSIGs, but that opens up the DOS and privacy attacks that have been described 
elsewhere in this discussion. And even if we issue admonitions to not have a 
local copy of the root zone without also enabling DNSSEC validation, we know 
that realistically there will be those who do the former without the latter.

For all those reasons, I think a checksum in the zone file itself that can be 
verified with DNSSEC is the best option for this use case, and I like the 
ZONEMD solution.

Matt




_______________________________________________
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop

Reply via email to