On Jun 12, 2009, at 1:30 AM, Zach Shelby wrote:
Also, in various places in the ND draft, we say *stateless* address
autoconfiguration - when in fact this is not the case. The
whiteboard
maintains necessary state for all nodes in the network no matter how
you spin it. If that state does not exist, is not maintained
properly,
or cannot be reached by the client node, DAD using the whiteboard
will
fail. At the very least, I think we should drop the word "stateless"
everywhere in the draft.
I think you're very right, that's an excellent point.
I don't agree completely. The node is creating an address using
Stateless Address Autoconfiguration (RFC4862), we are not breaking
that. So in Section 5.2, where we are forming addresses, this is
used correctly.
You have to look at the whole process. RFC 4862 SLAAC is useful
because DAD also does not require any servers or state maintenance.
I'd argue that if RFC 4862 were changed to use 6LoWPAN ND DAD for
*all* networks, then the whole point of SLAAC is missed.
The only thing we are changing, is the mechanism to perform DAD.
I agree that the Whiteboard is not stateless, but it also does not
start with any state, nor does it require configuration from an
administrator. It is a blank slate..
My closest analogy (again from a protocol point of view, not semantic)
of 6LoWPAN ND is the DHCP server in your home linksys box. It requires
no real user-visible network configuration. It doesn't start with any
per-node state. Nodes on the network must communicate with it to join
the network. Let's not kid ourselves here in thinking that the
whiteboard is something that is simpler to understand, use, and manage
than a simple DHCP server is. It's certainly something that Richard is
concerned about and I'm sure others will be too.
--
Jonathan Hui
_______________________________________________
6lowpan mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6lowpan
