1027938 - Timeout for ICE checks should be even longer

Status: RESOLVED FIXED

(Core :: WebRTC: Networking, defect)

Description

[Byron Campen [:bwc]]

After some discussion with TokBox, we're going to try another increase in the ICE check maximum retransmit count.

Comment 1

[Byron Campen [:bwc]]

Attachment: Bump max transmit count for ICE checks to 7. Also, adjust unit-tests. The wait time on ice_unittest has been made easier to adjust case-by-case, and both ice_unittest and signalling_unittests use stun.services.mozilla.com instead of bare IP addresses and

Initial cut. Most of the real work is in the tests.

Comment 2

[Byron Campen [:bwc]]

https://tbpl.mozilla.org/?tree=Try&rev=6220d204cb6a

Comment 3

[Nils Ohlmeier [:drno]]

Comment on attachment 8443151 [details] [diff] [review]
Bump max transmit count for ICE checks to 7. Also, adjust unit-tests. The wait time on ice_unittest has been made easier to adjust case-by-case, and both ice_unittest and signalling_unittests use stun.services.mozilla.com instead of bare IP addresses and

Review of attachment 8443151 [details] [diff] [review]:
-----------------------------------------------------------------

I don't see any unit updated for the increase from 5 to 7 retries. Don't we have anything which verifies the amount of retries?

LGTM.

Comment 4

[Byron Campen [:bwc]]

> I don't see any unit updated for the increase from 5 to 7 retries. Don't we
> have anything which verifies the amount of retries?

Not really, no. We have some tests that require a STUN transaction to time out before they end, and I had to adjust the wait time for gathering on them.

Comment 5

[Byron Campen [:bwc]]

Needinfo self to check back on try push.

https://tbpl.mozilla.org/?tree=Try&rev=4d945b907ec8

Comment 6

[Byron Campen [:bwc]]

Try push looks good.

Comment 7

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/d91074a4d004

Comment 8

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/d91074a4d004