At  I have close to 100% failure rate of ICE restart mochitests on 32-bit linux. At  I have no failures. The only difference is the patch at  that was applied in the failing case. In essence this changed the logging from > MOZ_LOG=signaling:3,mtransport:4,DataChannel:4,jsep:4,MediaPipelineFactory:4 to > NSPR_LOG_MODULES=timestamp,sync,MediaStream:5,MediaStreamTrack:5,mediapipeline:4,signaling:3,MediaPipelineFactory:5 This makes it look like a timing dependent bug that we should probably fix while we're able to reproduce it.  https://treeherder.mozilla.org/#/jobs?repo=try&revision=fba8ee056e8805d9f3409b7d17ced5ae7f5ca430  https://treeherder.mozilla.org/#/jobs?repo=try&revision=6c9687fc01e1  https://hg.mozilla.org/try/rev/79f4b49f5a5759144007d1c53c0991770a4d49b2
mjf, is this something you'd like to look at?
I could probably look at it after I get done with some ICE TCP stuff.
Mass change P3->P4 to align with new Mozilla triage process.
Priority: P3 → P4
You need to log in before you can comment on or make changes to this bug.