Re: [webrtc-stats] How do we handle overflow of metrics (#496)

Local counters are under our control, we should just keep counting up.
Remote counters that are passed in RTCP are more of an issue - what do the IETF specs say?
Back of the envelope: For 1 Kbyte packets in a 10 Mbits/sec video RTP session, we'll have ~1K packets/second, so the 32-bit packet counter will wrap in 49 days - this is not totally unthinkable, but rare enough that it's not likely to show up in casual testing.


-- 
GitHub Notification of comment by alvestrand
Please view or discuss this issue at https://github.com/w3c/webrtc-stats/issues/496#issuecomment-532006628 using your GitHub account

Received on Tuesday, 17 September 2019 00:17:55 UTC