Re: Confirmation to advance: draft-ietf-6man-ipv6only-flag-05

[David Farmer <[e-mail]>]

On Thu, May 9, 2019 at 2:17 PM Sander Steffann <[e-mail]> wrote:

> Hi David,
>
> > Because with this flag now only being a secondary signal and the lack of
> any DHCPOFFERS is now the primary signal, it is safe to allow any RA set to
> 1 to trigger no longer auto-configuring an IPv4 Link-Local address, this
> only has an effect if the host also doesn't receive any DHCPOFFERS.
>
> Making the lack of DHCPOFFERS the primary signal and this flag a secondary
> signal would take away my strongest objections.


Thanks, glad to hear that.


> I still wonder whether it is worth the effort and the added learning curve
> and complexity, and whether other heuristics (see Philip Homburg's message
> for example) aren't as good/good enough. At least the debugging effort for
> help desks won't be harmed anymore by the flag, because as long as
> DHCPOFFERS are working the flag doesn't hurt, and solving problems when
> DHCPOFFERS aren't working is already part of their current procedures.
>

I reviewed Philip's messages, I found several heuristics effectively
relating to when to suppressing periodic IPv4 DHCPDISCOVERS.  However, for
when to not auto-configuring an IPv4 Link-Local address, the only
suggestion I found was RFC 2563. And unfortunately, as that requires an
IPv4 DHCPOFFER in response to a DHCPDISCOVER to be successfully returned to
the host, which is basically incompatible with Layer 2 filtering of
Ethertype 0x0800, I guess you could still filter Ethertype 0x0806 though.

Also, while many older OSes seem to support  RFC 2563, I many newer OSes
seem to have dropped support for it or at least as best as I can tell.

I like the idea a positive signal provided via IPv6 to tell dual-stack
hosts to not auto-configuring an IPv4 Link-Local address. I also think
suppressing periodic DHCPDISCOVERS when all the RA flags agree has the
advantage that when there are no longer any valid RAs, you no longer
suppressing periodic DHCPDISCOVERS.

Thanks.

>
> Cheers,
> Sander
>
>

-- 
===============================================
David Farmer               Email:[e-mail]
Networking & Telecommunication Services
Office of Information Technology
University of Minnesota
2218 University Ave SE        Phone: [phone]
Minneapolis, MN 55414-3029   Cell: 612-812-9952
===============================================