The consequence is that we don't measure actual WebRTC calls but rather telemetry for failed WebRTC calls. This is better than nothing but it's obviously more interesting to measure real calls.

Note to posterity: I tried the following for perf/benchmarks/webrtc.py:

    options.AppendExtraBrowserArgs('--allow-loopback-in-peer-connection')
    options.AppendExtraBrowserArgs('--enforce-webrtc-ip-permission-check')

Neither of the flags makes a difference. I'm sure we need the topmost one, but it's apparently not enough.

Pri-3 with no updates in 180+ days that are available and have no owner == Archived. Feel free to update status, add context as to why and move this along.

Edward can you look into if this is the case, since the reproduction steps are so easy to perform?

Did you get any result from this?

Just confirmed. 

Thanks. moving to Blink>WebRTC>PeerConnection for triage.

Ned: Do you think this is an issue with WebRTC or with telemetry?

Likely telemetry (see  issue 704782 ). I am swarmed with stuffs right now, so if you can help with that issue, I will really appreciate it. 

How much effort do you think it'll take?

Probably few days for people who are familiar with Telemetry. The tasks would include validate whether chrome flag matches with what shown in the trace, work with perezju@ to fix it, and add test to prevent it from regressing again.

I'll be away for the first two weeks of April, let's talk about this after I come back if that is still an issue :)

Assigning to ehmaldonado@, since he seems to have a plan. Please reassign if necessary.