Re: [dhcwg] Review of Service-Discovery-Type options in DHCP

>What you propose is perfectly workable, but I don't
>see how it's *better* than what DHCP already provides.

My belief is that one of the benefits of IPv6 is that it does automatic 
address allocation without needing a DHCP server. This is good. This 
means that your network operations people have one less server to install 
and configure and maintain.

If we make our IPv6 hosts dependent on DHCPv6 servers for lots of 
critical information that they cannot operate without, then we still need 
to run a DHCP server and we have lost that IPv6 benefit.

Since (a) hosts still need to find out information like "Where's my time 
server", and (b) in an all-IPv6 world we still plan to run DNS servers 
anyway, then if we can put information (a) into DNS (b), we can eliminate 
the burden of also having to install and configure and maintain a DHCP 
server. Instead of having to communicate arbitrary bootstrap 
configuration information to hosts, we now just have to communicate the 
name server address(es) and the default domain, which is a nice simple 
constrained problem that can be solved in any number of simple ways -- 
e.g. add that information to Router Advertisement packets.

[Of course, I am well aware that this point of view will not be popular 
with all the people working so hard for so long on DHCPv6. Also, if you 
have a company that sells a DHCP server, then eliminating the need for 
DHCP servers is not in your company's best interest. I understand that.]

Stuart Cheshire <cheshire@apple.com>
 * Wizard Without Portfolio, Apple Computer
 * Chairman, IETF ZEROCONF
 * www.stuartcheshire.org

_______________________________________________
dhcwg mailing list
dhcwg@ietf.org
https://www1.ietf.org/mailman/listinfo/dhcwg