On Mon, Feb 18, 2013 at 4:12 PM, Behcet Sarikaya <[email protected]> wrote: > would be delighted to hear your comments. > > A new version of I-D, draft-sarikaya-nvo3-vmm-dmm-pmip-00.txt
I have read your draft and I have a number of comments. You indicate a BGP-based mechanism for inter-datacenter IP mobility. Your draft specifies that, "UPDATE message contains one or more ORIGIN path attributes containing the address prefix values in IPv4 or IPv6 of the VM when it was at the source hypervisor." I think you are confused because the ORIGIN attribute contains one of three possible values: IGP, EGP, or INCOMPLETE. These are commonly expressed in router CLI as "I", "E", or "?" and are a component used in the BGP best-path selection algorithm. They do not identify an originating router or node for a prefix. You propose the management control-plane will send routing update messages to the "old hypervisor" ToR or its default router, and also send messages to the old hypervisor itself, causing that old hypervisor to install host routes. Why? How do you know the old hypervisor's ToR or gateway will be capable of forwarding packets to the new hypervisor? If the network contains more than one router, routing information must be propagated to some additional routers, even in absence of a multiple-datacenter scenario, if the "source" and "destination" hypervisors do not share a common default gateway. A person could read this and wonder whether the authors have any experience with, or understanding of, IP networks containing more than one router. -- Jeff S Wheeler <[email protected]> Sr Network Operator / Innovative Network Concepts _______________________________________________ nvo3 mailing list [email protected] https://www.ietf.org/mailman/listinfo/nvo3
