Project Stockwell – February 2017

I realized my post for last month was titled “Project Stockwell – January 2016” – that is a fun typo to make 🙂

Last month we focused on triaging all bugs that met our criteria of >=30 failures/week.  Every day there are many new bugs to triage and we started with a large list.  In the end we have commented on all the bugs and have a small list every day to revisit or investigate.

One thing we focus on is only requesting assistance at most once per week- to that note we have a “Neglected Oranges” dashboard that we use daily.

What is changing this month- We will be recommending resolution on priority bugs (>=30 failures/week) in 2 weeks time.  Resolution is active debugging, landing changes to the test to reduce,debug, or fix the intermittent, or in the case where there is a lack of time or ease of finding a fix disabling the test.  If this goes well, we will reduce that down to 7 days in March.

So how are we doing?

Week starting:	Jan 02, 2017	Jan 30, 2017
Orange Factor	13.76	10.75
# priority intermittent	42	61

We have less overall failures, but more bugs spread out.  Some interesting bugs:

• upgrade libxcb
• disabled webvtt cue processing tests which are from web-platform-tests but not implemented in gecko.

In terms of projects underway, here is some status:

• adding BUG_COMPONENTS to all files in m-c (bug 1328351) – slow and steady progress, thanks for the reviews to date!  We expect large majority of this to be completed this month.
• retrigger an existing job with additional debugging arguments (bug 1322433) – main discussion is done, figuring out small details, should see a prototype this month
• add |mach test-info| support (bug 1324470) – landed today!
• add a test-lint job to linux64/mochitest (bug 1323044) – no progress yet, I expect some this month.

Are there items we should be working on or looking into?  Please join our meetings.