864117 - CreateOffer() with no network connection never succeeds or fails

Status: RESOLVED FIXED

(Core :: WebRTC: Networking, defect, P5)

Description

[Jason Smith [:jsmith]]

Attachment: Test Case

Note: No network connection = wireless is disconnected. This will not reproduce with "work offline" in Firefox.

STR:

1. Load the attached test case without a network connection
2. Check console

Expected:

We should get a NS_ERROR_FAILURE indicating that you can't create a peer connection without a network connection.

Actual:

The peer connection is successfully created. We'll fail silently when we reach createOffer, so the developer never knows the connection failed.

Comment 1

[Eric Rescorla (:ekr)]

Why is this marked as TURN? Does it succeed without TURN?

Comment 2

[Jason Smith [:jsmith]]

(In reply to Eric Rescorla (:ekr) from comment #1)
> Why is this marked as TURN? Does it succeed without TURN?

Fails on STUN too.

Comment 3

[Randell Jesup [:jesup] (needinfo me)]

You should be able to create a PeerConnection without a network connection, so that part is incorrect.

CreateOffer (I believe) should eventually time out and return an error if it can't even get local interface IP addresses.  If it never does, that's probably an error.

Please do a run with signaling:5,mtransport:5 logging and upload it (my laptop has a VM that looks like a network interface on it, so it's hard for me to test, and my linux machine normally doesn't have a keyboard or mouse - perhaps I'll just plug some in and try there.)

Comment 4

[Eric Rescorla (:ekr)]

(In reply to Randell Jesup [:jesup] from comment #3)
> You should be able to create a PeerConnection without a network connection,
> so that part is incorrect.
> 
> CreateOffer (I believe) should eventually time out and return an error if it
> can't even get local interface IP addresses.  If it never does, that's
> probably an error.
> 
> Please do a run with signaling:5,mtransport:5 logging and upload it (my
> laptop has a VM that looks like a network interface on it, so it's hard for
> me to test, and my linux machine normally doesn't have a keyboard or mouse -
> perhaps I'll just plug some in and try there.)

Yes, this is bug  827982... unfortunately, the writeup isn't accurate any more
because we now have asynchronous StartGathering.

Comment 5

[Mark Banner (:standard8)]

Attachment: Signalling Log

Not sure if it is still necessary, but here's the logs as requested as I discovered this the other day.

Comment 6

[Mark Banner (:standard8)]

Attachment: Alternate test case

The test case I was using (just in case).

Comment 7

[Randell Jesup [:jesup] (needinfo me)]

may well be fixed now...

Comment 8

[Byron Campen [:bwc]]

Just tested, seems to work.