New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 689017 link

Starred by 7 users

Issue metadata

Status: Available
Owner: ----
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Windows , Mac
Pri: 3
Type: Bug



Sign in to add a comment

No DTLS/SCTP shutdown procedure when tab is closed (or Chrome is quit:ed) for PeerConnection with datachannel

Reported by and...@lonelycoder.com, Feb 6 2017

Issue description

UserAgent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_3) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36

Steps to reproduce the problem:
1. Setup a WebRTC PeerConnection to a peer. 
2. Close the Tab

What is the expected behavior?
DTLS Alert and/or SCTP Abort should be sent by Chrome.

What went wrong?
Did not receive and shutdown notification.

See attached file (with protocol dumps). 

Displays packets exchanged both when using Chrome and Firefox.

Did this work before? No 

Does this work in other browsers? Yes

Chrome version: 56.0.2924.87  Channel: stable
OS Version: OS X 10.12.3
Flash Version: Shockwave Flash 24.0 r0

Chrome does shutdown the WebRTC connection when my laptop lid is closed (as expected I suppose)

I've not done any additional testing when using WebRTC with video or audio channels as data is my primary concern.

 
webrtc.txt
13.1 KB View Download
Components: -Blink>WebRTC Blink>WebRTC>Network
Cc: deadbeef@chromium.org
Labels: -Pri-2 Pri-3
Status: Available (was: Unconfirmed)
I confirmed that this is an issue, though I'm putting it at a lower priority since these notifications aren't reliable (being transmitted over UDP).
Labels: WebRTCTriaged
UserAgent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_3) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36

Steps to reproduce the problem:
1. Setup a WebRTC PeerConnection to a peer. 
2. Close the Tab

What is the expected behavior?
DTLS Alert and/or SCTP Abort should be sent by Chrome.

What went wrong?
Did not receive and shutdown notification.

See attached file (with protocol dumps). 

Displays packets exchanged both when using Chrome and Firefox.

Did this work before? No 

Does this work in other browsers? Yes

Chrome version: 56.0.2924.87  Channel: stable
OS Version: OS X 10.12.3
Flash Version: Shockwave Flash 24.0 r0

Chrome does shutdown the WebRTC connection when my laptop lid is closed (as expected I suppose)

I've not done any additional testing when using WebRTC with video or audio channels as data is my primary concern.

Comment 6 by pmfent...@gmail.com, Feb 14 2017

A side-effect of this problem is that the auto-sleep function of host O.S. won't kick in if you closed a tab that had a webRTC peerconnection.  You pretty much have to shut down all chrome tabs.  (Or do not close the tab and instead navigate that tab to a website that doesn't make any webRTC connections.)
Labels: Performance-Power
fwiw I highly disagree with this being p3, as it prevents sleep on devices that should otherwise be sleeping. This was not an issue before either. Looks like the initial report was on OSX, this is happening on Windows 10 as well.
Labels: OS-Windows
Summary: No DTLS/SCTP shutdown procedure when tab is closed (or Chrome is quit:ed) for PeerConnection with datachannel (was: No WebRTC Shutdown when tab is closed (or Chrome is quit:ed) for PeerConnection with datachannel)
The sleep-related issue seems unrelated to not sending DTLS alert/SCTP abort packets. Here's a new issue: https://bugs.chromium.org/p/chromium/issues/detail?id=693804
The sleep issue was recently fixed in  issue 612294 

Comment 12 by gao...@gmail.com, Mar 24 2017

 Chrome version: 59.0.3050.3 canary (64-bit) still have this bug.

you can open https://aixuexi.tmall.com/?ali_trackid=17_802f2a00025e4fd75b683ed4d94c923e 

when this tab open, use "pmset -g assertions" result:

  pid 52034(Google Chrome Canary): [0x0003a64d0001858e] 00:02:59 NoIdleSleepAssertion named: "WebRTC has active PeerConnections"


This bug is specifically about DTLS/SCTP shutdown procedures, not general sleep issues. Let's discuss this on  issue 612294 .
Should we keep this bug open, or archive it?
It still exists as far as I'm aware.
Still exist in version 68.0.3440.106. Running my outlook.com in Edge instead.

Sign in to add a comment