De we need additional signaling to make this work reliably for all candidate lists?

[rshpount]

Looking at this problem it is clear that it would be difficult to come up with a timer value which will work for all possible candidate lists. It probably makes sense to have some sort signaling message which requests ICE nomination time out extension. This signaling message will let the other ICE agent know that ICE agent is still sending binding requests and needs nomination time out extended. In other words, ICE should not fail for 32 seconds, but agent should send a signaling message (INFO with current candidates or something similar) every 16 seconds to keep negotiation going unless a binding request from remote was received. This signaling message would extend nomination timeout for another 32 sec from the time the message was received.

[juberti]

This seems like it is expanding the scope of this document a fair amount - can you think of a scenario where this would actually be needed?

[cdh4u]

"This signaling message will let the other ICE agent know that ICE agent is still sending binding requests and needs nomination time out extended."

I am not sure I understand. The draft does not change for how long an ICE agent is sending binding request - it changes for how long an ICE agent waits to RECEIVE binding requests before it triggers ICE failure.

[rshpount]

One thing that is hard to determine for the agent is how long it is supposed to wait for STUN Bind requests before declaring ICE nomination failure. There is no way for the agent to know how many candidate pairs the remote ICE agent is going to try and time out values this agent is going to use. Because of this, there is always a risk that agent stop waiting for incoming bind requests and declares ICE nomination failure before remote agent actually tried all the candidate pairs. One way to work around this is to add a signaling message to indicate that candidate pairs were tried to inform remote agent that it is safe to stop waiting for incoming requests.

Another way of solving this is simply specify that agent waits for some reasonable timer interval before failing. This is what we've chosen for reasons of simplicity.

[juberti]

Another way of solving this is simply specify that agent waits for some reasonable timer interval before failing.

Agreed that that's what we've chosen. Closing accordingly.