Mobile/Testing/08 08 12
From MozillaWiki
Contents
Previous Action Items
- Clint to have a prioritized list of both automation and test failure bugs next week
- Hwine to have a task list for Android PGO
- Jmaher to file a bug for mochitest not running on panda, and for not being able to run debug builds on panda
- Mfinkle to talk with kats about the eideticker regression on CNN bug 777357
- wlach to get mozilla-china person involved in ARMv6 bug for new tablet
Status reports
Dev team
Rel Eng
Tegra
- The reboot issue is fixed, after coord with wlach and jmaher and rest of ATeam.
- Making large progress on linux-foopy (thanks to glandium I think last blocker issue is fixed)
- New Tegras are mostly wired/powered/networked, just have to bring them into production
Panda
Beagle
Foopy-less
Other
- x86 onhold for now bug 750366
- reminder: coop is covering for armenzg until the end of August
IT
- Developing a remote re-imaging process for pandas bug 764534
- Agree on requirements for future panda chassis bug 778221
- Working on a higher density chassis bug 777393 - dependent on 778221
- IT will be expanding in SCL1 Datacenter for mobile device housing bug 774477
- 81 new tegras will be put into production pool bug 767447
- Building 5 new panda chassis for scl1 bug 777359
A Team
- 11.11% total failure rate
- M3 is fixed on inbound, ignoring that would put us under 7% total failure rate.
- Panda Boards
- adjusted resolution [via bootargs] again. This seemed to solve the running of mochitests and talos.
- concerns as if this is the right set of bootargs? no hdmi output, but screecap and ddms work just fine
- burnin script for hardware (based on Autophone) is mostly developed, sanity testing now. bug 776728
- adjusted resolution [via bootargs] again. This seemed to solve the running of mochitests and talos.
S1/S2 Automation
- adb -> SUT change deployed (then fixed properly). This changes timing, so it looks like there's a regression.
- I will rerun some older builds to make sure the "regression" is present there too.
- Design for C++ cross-platform agent up
- Work has started on it but running into some difficulties building it with all the necessary libraries.
- Wondering if, since time is critical, it makes more sense to fix up the C agent first, so the B2G automation folks have something to work with while we build up the C++ agent.
Eideticker
- Still no luck finding ARMv6 tablet, asked mozilla china, no luck
- Looks like we found the original device that we were looking for, not sure if it fits the bill though bug 764491
- Some patches for B2G starting to trickle in
Round Table
- wlach's post on tegra reboot issue resolution: https://groups.google.com/forum/?fromgroups#!topic/mozilla.tools/01jmBM57JXw
- What can we do about documenting our systems end-to-end?
Action Items
- hwine to figure out what the hardware requirement is for library reordering (really panda only??)
- hwine to work with gbrown if emulator is an option
- Mfinkle to talk with kats about the eideticker regression on CNN bug 777357
- wlach to start the process of documenting our automation end-to-end