Re: [rtcweb] IP handling: Using mDNS names for host candidates

[Justin Uberti <[e-mail]>]

On Thu, Jun 14, 2018 at 2:15 PM Harald Alvestrand <[e-mail]>
wrote:

> I have some worries about this proposal. It seems neat, and solves a
> specific problem for a specific use case, but it's not a written-up
> proposal, rather a sketch for one - and I'm afraid of devils in the
> details.
>
> For instance:
>
> - If this technique is used for a computer directly connected to the
> Internet, with a public IP address, it won't communicate - unless it is
> only used on private addresses - because "uuid.local" doesn't resolve,
> whereas a public IP address is globally reachable.


> - The above means that the proposal needs a definition of "private
> address". Do we mean "private" in the RFC 1918 sense? If so, which IPv6
> range is covered by the proposal?
>
> - It will only work if the private address usage is the same scope as
> mDNS resolution. On an unmanaged LAN it works, and on a network with
> explicit mDNS forwarding it works. But on any other deployment, it
> forces traffic to go via public IP addresses learned by STUN.
>
> I think this is worth adding. Perhaps as a new "mode 2m"?
>
> But I'd like a commitment to not adding it until we have a full proposal.
>

I have sketched out the proposal in
https://github.com/juberti/draughts/pull/103, which while not complete,
does address most of your questions.

>
> Den 12. juni 2018 02:40, skrev Justin Uberti:
> > The Safari team has come up with a clever approach to avoid disclosing
> > private IP addresses for host candidates. As discussed in this WebKit
> > bug <https://bugs.webkit.org/show_bug.cgi?id=174500>, the technique
> > works as follows:
> >
> >  1. Register a random UUID-based mDNS name when ICE gathering starts
> >  2. Replace the private IP address by a "{UUID}.local" string in each
> >     host candidate (and set raddr to 0.0.0.0 for other candidates)
> >  3. The other party will do mDNS resolution on any candidate having a
> >     .local suffix, similar to how hostnames in candidates are handled in
> >     RFC 5245, Section 15.1.
> >
> > This technique is relevant to the IP handling document, as it addresses
> > one of the lesser problems (private IP disclosure) from the overall
> > problem statement. While I don't think this will have a large impact on
> > the document, including the default mode selection, incorporating this
> > technique would result in some moderate changes:
> >
> >   * Section 5.1 would mention concealing private IPs in the default case
> >     as an explicit goal.
> >   * In Section 6, Mode 2 would change from handling out private IPs to
> >     handing out mDNS names.
> >   * This document would have to describe the technique or point to
> >     another document that describes the technique. mmusic-ice-sip-sdp,
> >     Section 4.1
> >     <
> https://tools.ietf.org/html/draft-ietf-mmusic-ice-sip-sdp-20#section-4.1
> > seems
> >     like a good option, as it already covers how to handle DNS names in
> >     ICE candidates.
> >
> > This is a significant improvement and I think we will want to
> > incorporate this suggestion. Is this something we could do as part of
> > this WGLC, or should we look for another option?
> >
> >
> > _______________________________________________
> > rtcweb mailing list
> > [e-mail]
> > https://www.ietf.org/mailman/listinfo/rtcweb
> >
>
> _______________________________________________
> rtcweb mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/rtcweb
>