Juliusz, On 03/02/2014 08:53, Juliusz Chroboczek wrote: >> I happen to have one in mind. It's discussed for the case of >> carrier networks in >> http://tools.ietf.org/html/draft-jiang-config-negotiation-ps-02 > > I've now read this draft. Section 2.4 says things that I never even > considered, and I need to think it over. (In particular, functionality > such as what you describe there would be marvelous for monitoring tools.) > > There's only one point that I disagree with. In Section 4 you say: > > Multi-party negotiations are too complicated to be modeled and > there may be too many dependencies among the parties to converge > efficiently. > > If I read you right, you're disallowing the familiar claim/defend > scheme, where a node claims an address over multicast, and waits for > other nodes to disagree. (The best-known case is probably that of > DHCPv4, where implementations commonly make an ARP query before > assigning an address, but there's no reason why this shouldn't be an > intrinsic part of the configuration protocol.)
I don't think we intended to exclude that, but there may well be scenarios where suitably scoped multicast isn't available. We were concerned about guaranteed convergence in a multi-party scenario, which seems tractable with (reasonably reliable) multicast. Anyway, this aspect needs more thought, thanks! Brian > > -- Juliusz > . > _______________________________________________ homenet mailing list homenet@ietf.org https://www.ietf.org/mailman/listinfo/homenet
