Weiqiang Cheng, On Tue, Nov 12, 2019 at 03:56:23PM +0800, Weiqiang Cheng wrote: > We will cover those points in the presentation.
Thank you. > We just have a quick review on TR-146, where it says that Echo function is > used and the destination IP is set to an IP address of the sender (the > latter is similar to the One-arm idea) for unidirectional failure detection. > > But based on our understanding, BFD Echo requires that the receiver must be > able to recognize and process BFD control packet. One-arm idea assumes that > the receiving end is completely unaware of BFD, it just forwards the packet > based on routing. This is very useful in the scenario (e.g., Data Center or > 4G/5G Telecom Cloud) where large number of BFD sessions may be needed, and > the receiver cannot handle too much BFD sessions with limited resource. TR-146 similarly does not use any BFD signaling, so it is very similar (identical?) to your Internet-Draft. > The idea is straight forward, we will include more details (e.g., session > establishment, state machine, YANG etc.) in the next version. Thank you. -- Jeff
