E10s/Status/May20: Difference between revisions

no edit summary
(Created page with "thumbnail|left<br> =e10s Update: '''May 6'''= ==Executive Summary== * The biggest news this week is that all Release Criteria that has been signed off...")
 
No edit summary
 
(17 intermediate revisions by the same user not shown)
Line 1: Line 1:
[[File:Yellow-sm.jpg|thumbnail|left]]<br>
[[File:Green-sm.jpg|thumbnail|left]]<br>


=e10s Update: '''May 6'''=
=e10s Update: '''July 1'''=
==Executive Summary==
==Executive Summary==
* The biggest news this week is that all Release Criteria that has been signed off continues to pass in Firefox 47 Beta. There are no new regressions thus far. This is according to our latest [ https://github.com/vitillo/e10s_analyses/blob/master/beta47-ex1/e10s_experiment.ipynb derived data set] which ran from April 27-May 4. We will run these data sets weekly in order to evaluate performance on an ongoing basis. We have at least 469,350 users with e10s enabled on beta.  
* The biggest news is that the stability release criteria for Beta 48 has been signed off on [https://wiki.mozilla.org/Electrolysis/Release_Criteria#Stability]. Slow script count is also within release criteria target.
* We are also at [https://wiki.mozilla.org/E10s/Status/m9#Open_M9_bugs 0 M9's] (and have been) for a few weeks 
* As of this week, 569K users have been exposed to e10s in Beta 48.  
* Given that we have met all of our release criteria and we have not regressed since 47 beta, the project is now designated 'green' or 'on track' to ride the trains to GA and be gradually rolled out (1% at start) to eligible users (41% of total population) or about 800K users for the first 3 weeks.
* '''So what's the release plan?''' We are working on a day-by-day snapshot of the activities that have to occur once Firefox 48 goes to GA. This is basically a small 'throttle team' who will be at helm of throttling e10s up or down (or disabling). This is driven by RelMan but will be supported by e10s subject matter experts so that decisions can be made quickly. You can refer to these docs for more info and WIP details: http://bit.ly/29mIbXs and


* Our #1 risk remains stability. Our [https://sql.telemetry.mozilla.org/queries/271#445 crash rates for beta 1] are still higher (27.2) vs. non-e10s (9.2). A full 33% of e10s content crashes in b1 were caused by {{bug|1255823}} which is fixed for b2, so we expect the crash rate for b2 to stabilize around 20. Getting to 20 or below is our stated goal for next week.
With the approach of Firefox 47 Beta 4 on May 10, stability and unearthing [https://docs.google.com/document/d/150XUlvXi13mliWIixS5x5DoLE14ySxItBhQfCIKg1NI/edit important details we need to line up for a staged rollout to GA] remains our central focus.
==Why yellow? e10s is designated 'yellow' or 'at risk' because: ==
Stability is our #1 risk to our plan to stay enable for the entire beta cycle in Firefox 47. We also need to continue to make progress on an issue with [https://bugzilla.mozilla.org/show_bug.cgi?id=1260769 slow scripts]. All other areas of release criteria are looking good.


==Next Steps==
==Next Steps==
* 04/10: Firefox 47 Beta 3 Ships
* 05/20: Firefox 47 Beta 7 Ships
* 05/11: Second derived data set available (2 weeks of data)
* 05/25: Fourth derived data set available (4 weeks of data)
* 05/12: Quality check point
* 05/24: Release plan review


==Release Criteria==
* See the wiki for up-to-date [https://wiki.mozilla.org/Electrolysis/Release_Criteria e10s release criteria status]


==Add-Ons + WebExtensions==
==Add-Ons + WebExtensions==
Line 25: Line 21:


==Automated tests==
==Automated tests==
* Everything we can have enabled in production is enabled, everything else is enabled on a separate branch called 'Ash'.
* We are at about 90% [https://wiki.mozilla.org/Electrolysis/Test_Coverage Test coverage for all channels. We have touched base with all module owners. Not all of them have signed off specifically but we are calling it within acceptable range for shipping at this point.
* We have all tests be owned
 
* We would like each directory signed-off by test owners week of 06/05
*We're tracking all the tests and who owns them, [https://docs.google.com/spreadsheets/d/10UeyRoiWV2HjkWwAU51HXyXAV7YLi4BjDm55mr5Xv6c/edit#gid=1777180571 here]
*We are tracking tests that are running and passing on some platforms, but are disabled on others.From a feature coverage point of view, it's in pretty great shape. There's very little that is disabled both on opt *and* debug on a same platform (meaning that that test is not running on this platform). Look for the "Fully disabled in a platform" entries in this [https://docs.google.com/spreadsheets/d/1oBnRct8FuTnAoVfbE08m10UHi1_5kOFaOPglJ6ittWc/edit#gid=0 spreadsheet].


==A11y==
==A11y==
* Proposed technical path to encourage a11y clients to communicate directly with the content process {{bug|1258839}}. dbolter is getting some first impressions from clients to see if this is a viable approach.
* It's looking like the initial API support will land in 50
 
* We're also looking into Windows touchscreen/a11y compat, see more information as the plan is taking shape, [https://public.etherpad-mozilla.org/p/a11y+e10s here].
* We're also looking into Windows touchscreen/a11y compat, see more information as the plan is taking shape, [https://public.etherpad-mozilla.org/p/a11y+e10s here].


canmove, Confirmed users, Bureaucrats and Sysops emeriti
5,401

edits