DHCPv6 vs. Stateless Address Autoconfiguration (Part 1)
Since my initiation into the world of IPv6 in 2004, it seemed that a religious war of address management was brewing. Dynamic Host Configuration Protocol for IPv6 (DHCPv6) and Stateless Address Autoconfiguration (RFC 4862) became an issue of security, address management and scalable nodal allocations. Really, what is the big issue? In this blog (and later blog entries) I want to try to address each issue’s pro/con and hopefully discuss any myths behind each of the issues of automatic address issuance.
Issue #1: Why is DHCPv6 implemented differently in multiple operating systems?
This issue is quite a fascinating one in the industry. DHCP in IPv4 worked simply as a server on a network segment issuing IPv4 addresses to hosts that solicit to the broadcast (255.255.255.255). However, in IPv6 it is implemented quite a bit differently. The crux of the debate goes into how multiple IETF RFCs are translated.
M&O Camp: the newer Stateless Address Autoconfiguration standard (RFC 4862) discussed the use of the Managed and Other Configuration flags in Router Advertisements for “DHCPv6 configurations” to instruct a client on the network that there are DHCPv6 servers on the network. Therefore, a client will know to solicit for DHCPv6 servers. Otherwise, they will only accept address prefixes issued by routers on the network. They will then configure their own host identifiers.
Pro: secure address management, current support in Microsoft Vista clients
Con: must implement both solutions, unknown which clients support this approach, inflexible
Server-only camp: In the older address configuration (RFC 2462), the language used for DHCPv6 was a little more vague by using the term “stateful address configuration.” Many operating system distributions in the Linux and Unix world implement DHCPv6 clients very similar to the way they did in IPv4. The client will send out a request on the network using multicast (unlike in broadcast for DHCPv4) for any servers on the network. It will also listen for Router Advertisements and continue to do stateless configuration. However, the DHCPv6 address will be promoted to the primary address.
Pro: Flexible, easily integrated and understood by network operators, support in Linux/Unix systems
Con: No support in Microsoft Vista exists, not as secure to manage nodes on the network
More on the automatic address configuration debate to follow…
August 2nd, 2009 at 9:39 pm
[…] about implementation being “very simple” is rather brash. He’s correct, and explains the differences between Windows and Linux/Unix in this regard in his company blog. With lots of references to “dueling RFCs”, fun. For a broader view […]
August 3rd, 2009 at 2:15 pm
[…] Address Autconfiguration, managing and securing each strategy(as previously blogged here and here), and implementing them efficiently and intelligently. A network engineer/architect cannot […]