Re: [IPv6] RFC 6724 shouldn't prefer partial reachability over reachability

[Kyle Rose <[e-mail]>]

On Sun, Nov 26, 2023 at 10:20 AM Michael Richardson <[e-mail]>
wrote:

> But, you as an operator, would be deleting those announcements based upon
> your well configured BCP38.  So, it's really is entirely under your
> control,
> which I agree with you: is a better.
>

I am not referring to leaking routes, but to leaking rendezvous
information, i.e., DNS. In a world in which operators run recursive
resolvers that clients can be assumed to use when on your network, you
maybe can configure yours to drop all external address records referring to
ULA or 1918 space, but we're now in a world in which privacy (via DoH/DoQ)
and end-to-end integrity (via DNSSEC) have been prioritized over
operability, so you have to assume you won't have full control over what
DNS entries a client sees. This has implications for the set of trade-offs
involved in misconfigurations, especially when the naive user blames you
for something not working even when it's mostly not under your control.
With border routers sending ICMP unreachable rejections, working Happy
Eyeballs would hopefully make failover transparent, but "hopefully" is
doing a lot of heavy lifting in that expectation.

Anyway, I consider this an acceptable set of trade-offs, but there is a
discussion to be had about the implications that would benefit from the
experience of big-time operators who have run into a much broader set of
issues than I have.

Kyle