Re "If we don't call the connections between bridges links, what do we call them?"
We call them "segments". Yes "link" is an overloaded term, which is why I was sure to clearly define in what context I was using the term. Other documents and contexts may use different terminology. -Dave -----Original Message----- From: Gray, Eric [mailto:[EMAIL PROTECTED] Sent: Wednesday, April 05, 2006 2:12 PM To: Bob Hinden Cc: IPv6 WG; Stephen Sprunk; Dave Thaler Subject: RE: Ethernet as a shared (not broadcast) network? Bob, I wonder if it is likely that some of the confusion relating to "multi-link subnets" that Dave tries to address, in the draft below, comes from a disagreement about the definition of a "link"? In Dave's draft, link is defined as "generally used to refer to a topological area bounded by routers which decrement the IPv4 TTL or IPv6 Hop Limit when forwarding the packet." Outside of a strictly layer three context, this is simply not true. However, it works for router folks because - even if a router does have multiple interfaces into a single subnet (or collection of subnets), the router "bridges" between those (two or more) links. This can be simply modeled as - _________|__________________________________ | | | | | .-----(link-2)---+---- | .-+-. .-+-. | --+-------+ R +--(LINK-1)---+ B +---(link-4)---+---- | `-+-' `-+-' | | | `-----(link-3)---+---- |_________|__________________________________| | In this model, LINK-1 is as Dave's draft defines it. For this reason, it is shown in upper-case. The use of the term "link" for the link-2, link-3 and link-4 (clearly in lower-case in the figure) is just as legitimate a use of the term - but in a layer two sense, only. If we don't call the connections between bridges links, what do we call them? Hopefully, it helps to look at this model, because it should help to highlight the essential layer-two nature of having multiple interfaces connecting to the same subnet(s). By using a bridging logical function to separate the LINK (per Dave's definition) from the physical interfaces connecting to a subnet (set), issues with layer two propagation in that subnet are resolved by that function. -- Eric --> -----Original Message----- --> From: Bob Hinden [mailto:[EMAIL PROTECTED] --> Sent: Wednesday, April 05, 2006 4:24 PM --> To: Stephen Sprunk --> Cc: IPv6 WG --> Subject: Re: Ethernet as a shared (not broadcast) network? --> ... [SNIP] ... --> --> Bob --> --> --> >> Dave Thaler also wrote a draft on the general topic that was --> >> presented at the internet area meeting in Dallas: --> >> --> >> Issues With Protocols Proposing Multilink Subnets --> >> --> >> http://www.ietf.org/internet-drafts/draft-thaler-intarea-multilink-subne t-is sues-00.txt --> > --> > Very good reading; thanks for the pointer. --> > ... [SNIP] ... --> --> --> -------------------------------------------------------------------- --> IETF IPv6 working group mailing list --> [email protected] --> Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --> -------------------------------------------------------------------- --> -------------------------------------------------------------------- IETF IPv6 working group mailing list [email protected] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
