The context was the discussion in the BFD WG on BFD YANG model. The BFD YANG model is struggling with the need to support multiple BFD sessions between the same endpoints. If 5881 is clear that the BFD session has to traverse different network-layers, then it helps. I do not see the need for different time intervals for different applications as a strong argument to support multiple BFD sessions between the same endpoints. Applications can choose to ignore a DOWN notification if they find it too aggressive.
> On Nov 3, 2015, at 12:28 PM, Manav Bhatia <[email protected]> wrote: > > And pray what is the context? I presume its got something to do with the BFD > WG meeting. Would appreciate some details. > > I have a counter example of when one may need multiple BFD sessions between > the same end-points, but i would like to hear the context before i muddy the > waters. > > Cheers, Manav > > On Tue, Nov 3, 2015 at 7:56 AM, Gregory Mirsky <[email protected] > <mailto:[email protected]>> wrote: > Dear All, > > I think that this paragraph from Section 2 of RFC 5881 prohibits multiple > single-hop BFD sessions between the same pair of end points: > > Each BFD session between a pair of systems MUST traverse a separate > > network-layer path in both directions. This is necessary for > > demultiplexing to work properly, and also because (by definition) > > multiple sessions would otherwise be protecting the same path. > > > > Regards, > > Greg > > > > Mahesh Jethanandani [email protected]
