DHCPv6 route option (was Re: Android (lack of) support for DHCPv6)

Masataka Ohta mohta at necom830.hpcl.titech.ac.jp
Thu Jun 11 03:44:36 UTC 2015


Doug Barton wrote:

> No, you're not. Some of us have been saying that requiring RA is a bad 
> idea, and that adding features to it is a bad idea, for over 15 years now.

Need a DHCPv6 route option?

> Unfortunately the anti-DHCP crowd hasn't budged, no matter how many 
> operators have told them that they cannot manage an IPv6 network with 
> the current state of the protocol.

FYI, the operators suffering from a lack of feature of some standard
are free to have an agreement on how to use the private use part of
a number range in the standard controlled by someone else.

It is legal, harmless and IETF did so, for example, to map IPv6
multicast addresses to local (that is, not assigned by IEEE and
for private use) Ethernet group MAC addresses in section 7 of
rfc2464.

Thus, interested operators can have an agreement to use some
private use option values (224-254) of DHCPv6 for the route option.

Moreover, the agreement can be published as a NANOG/RIPE/APRICOT/...
recommendation or a some (newly formed) forum standard.

Then, some implementers are happily follow the recommendation/standard
in addition to IETF standards.

At least, ISC has done so, already.

	https://www.isc.org/blogs/routing-configuration-over-dhcpv6-2/
	The presented examples use values 242 for NEXT_HOP and 243
	for RTPREFIX option codes.

Don't you want to increase the number of operators endorsing the
private assignments?

						Masataka Ohta





More information about the NANOG mailing list