Jari Arkko wrote:
Not sure I like either one of these options. Preferrably, the probing frequencywould depend on whether you have something to send or not. For instance, ifwe don't have any payload packets to send, the frequency = n and if we do have packets to send, it would m, m << n. If you combine this with the ULP feedback and packet reception in CUD, we have a mechanism that normally does not need extra probes during traffic and which generates a very small amount of trafficwhen its idle.
We can do better than that. When there is no data traffic neither CUD or FBD should need to send any extra packets. I think we only care about reachability when we have packets to send or receive, so this should be ok. (If we want to detect that a peer might have lost shim state, it might be desirable to send some "context verification" type probe occasionally, but functionally that's different than verifying reachability for the current locator pair.)
Erik