2012/3/15 Masataka Ohta <mohta@necom830.hpcl.titech.ac.jp>:
Even ordinary routers are ends w.r.t. routing protocols, though they also behave as intermediate systems to other routers.
As LS requires less intelligence than DV, it converges faster.
I do believe that's the first time I've heard anybody suggest that a link state routing protocol requires "less intelligence" than a distance vector protocol.
If that isn't clear to you then don't presume to lecture me about the end to end principle.
Here is an exercise for you insisting on DNS, an intermediate system.
What if DNS servers, including root ones, are mobile?
DNS' basic bootstrapping issues don't change, nor do the solutions. The resovlers find the roots via a set of static well-known layer 3 address which, more and more these days, are actually anycast destinations matching diverse pieces of equipment. It makes no particular sense to enhance their mobility beyond this level. Before you jump up and down and yell "Ah ha!" realize that this is true of a mapping function at any level of the stack. ARP doesn't work without knowing the layer 2 broadcast address and IPv6's ND doesn't work without knowing a static set of multicast addresses. Below the roots, the authoritative zone servers are no different than any other node. If you're willing to tolerate the lowered TTL for your NS server's A and AAAA records then when IP address changes and your parent zone is willing to tolerate dynamic updates for any glue, then you can make DNS updates to the parent zone like any other mobile node. The clients find the recursing resovlers via whatever process assigns the client's IP address, e.g. DHCP or PPP. If it is for some reason useful for the server's base address to change then assign a set of VIPs to the DNS service and route them at layer 3. On the other side of the wall, the recursing resolvers don't particularly care about their source addresses for originating queries to the authoritative servers and will move to the newly favored address with nary a hitch. If you want an actually hard question, try this one: what do you do when fewer than all of the authoritative DNS servers for your node's name are available to receive an update? What do you do when those servers suffer a split brain where each is reachable to some clients but they can't talk to each other? How do you stop bog standard outages from escalating into major network partitions? For that matter, how do you solve the problem with your home agent approach? Is it even capable of having multiple home agents active for each node? How do you keep them in sync? Regards, Bill Herrin -- William D. Herrin ................ herrin@dirtside.com bill@herrin.us 3005 Crane Dr. ...................... Web: <http://bill.herrin.us/> Falls Church, VA 22042-3004