Hi, the RTGdir has asked me to do a review on draft-ietf-bfd-unsolicited-09 and I think its ready for publication. The concept of passively listening instances reacting on demand should simplify deployment of BDF, especially when switching between active and passive role can be done automatically decided on local demand. As far as I could see from RFC 5880 BDF signals its state (active or passive), so it cannot accidentally have two passive sides just reacting to each other (thinking the other as active). I have two nits with the document... 1st, I would like a clarification which/how many (all?) security measurements you consider mandatory... if you (as an example) run the protocol in a trusted environment, you might be able to skip authentication... but maybe using the TTL to keep the protocol "linklocal" should still be mandatory. 2nd, I would suggest writing out the full name "Bidirectional Forwarding Detection (BFD)" once in the abstract just to make sure nobody confuses the acronym. Henning Rogge