Santosh, Thanks for the update. A few comments upon my most recent reading of the draft:
- Please consider starting the auth-key-id at 1 rather than 0 and leave 0 reserved. - You don't document the sender timestamp format. :-) Going back to prior discussion from the Working Group, you probably require more clarification as to how you intend to use this mechanism. Since you are building upon authentication, choices include some of these and perhaps others that haven't been raised: - This is fine for the case where no authentication was intended to be used for the session. In this case, state transitions within this BFD session are suitable for standard use. - This is intended only for an adjunct session. In this case, perhaps you want to further document behavior in not taking the session to the Down state in order to permit continuous measurement at rate. - You intend to toggle into and out of this mode? (I suspect not, it complicates key rollover situations which are already a bit weak in our RFCs.) -- Jeff On Thu, Jun 11, 2015 at 06:18:04AM +0000, Santosh P K wrote: > Hello All, > > A new version of draft has been submitted. Below are the changes made. > > > > 1. Added use cases for BFD stability. > > 2. Addressed review comments given in BFD working group. > > 3. Added delay measurement details. > > > > Please do review and get back to us with review comments. > > > > Thanks > > Santosh P K
