QA/Execution/Web Testing/Meetings/2016-01-14: Difference between revisions

no edit summary
No edit summary
 
(16 intermediate revisions by 3 users not shown)
Line 4: Line 4:


=  Lead / Scribe =
=  Lead / Scribe =
* Lead:  ???
* Lead:  Valentina
* Scribe: ???
* Scribe: Dave
** Scribe is responsible for:
** Scribe is responsible for:
*** adding the action items and takeaways to the current meeting's minutes
*** adding the action items and takeaways to the current meeting's minutes
Line 13: Line 13:


= Last week's action items =
= Last week's action items =
* {{prev|}}[matt][stephen] - The participation team are looking for someone to review their pull request for mozillians-tests - meeting with them during Orlando work week
* {{done|}}[davehunt] Create an issue on addons-tests to update the version of Appium we are using to 1.4
** https://github.com/mozilla/Addon-Tests/issues/806
* {{new|}}[kthiessen] Try to update the default version of Appium we use to 1.4.16
* {{new|}}[stephend] Create an etherpad and send out a note about brainstorming ideas for Q1 goals
* {{new|}}[stephend] Finalize the "year in review" blog post
** team members: please populate the etherpad at https://public.etherpad-mozilla.org/p/webqa-year-in-review-2015


== Read-only ==
== Read-only ==
Line 19: Line 24:


= Discussion items =
= Discussion items =
* [davehunt] Changes to AMO dev caused expected failures, but due to high timeouts and reruns the amo.dev job took over 8 hours!
* [davehunt] Persona.org and related domains will be taken offline on November 30th, 2016
** I've removed reruns from amo.dev.* and set the timeout to 60 seconds
** This will affect any automation using BIDPOM and Persona Test User
** I'd like to do this for all jobs, but it will cause noise due to intermittent failures
** Does this also include restmail?
** Thoughts/opinions? Can we handle the noise and tackle the intermittent failures?
* [davehunt] I have (temporarily?) disabled pytest-timeout to investigate potential Sauce Labs or infrastructure issues
** Note that it is possible to mark individual tests as flaky and to indicate a maximum number of reruns.
** If it turns out we no longer need this plugin, it would remove an upgrade blocker for pytest/pytest-xdist
* [davehunt] [https://github.com/mozilla/marketplace-tests/commit/62295a50902a41cfd90f506368bc9bcff5e9580b Until recently], marketplace-tests waiting for an element to be visible would wait 600 seconds instead of 60
** Please let me know if you see any jobs stalled or taking a very long time
** This is a good example of why it's a bad idea to mix implicit and explicit waits
* [davehunt] Selenium 2.49 is out - any volunteers to work on the upgrade?
* [davehunt] Sauce Labs has announced they're dropping support for Appium 1.3.4 and older
* [davehunt] We can have [http://webqa-ci-staging1.qa.scl3.mozilla.com:8080/job/mozorg.adhoc/20/console colour in our Jenkins console output]. Do we want it?
** We're using 1.4.13 as our default version (latest recommended version is 1.4.16)
** For AMO we're using 1.3.7 due to issues experienced with 1.4.x, so we may want to look into this before support is dropped for 1.3.x entirely


= [https://wiki.mozilla.org/QA/Execution/Web_Testing/Goals/2016/Q1 Goals Check-in] - 1st and 3rd meetings, monthly =
= [https://wiki.mozilla.org/QA/Execution/Web_Testing/Goals/2016/Q1 Goals Check-in] - 1st and 3rd meetings, monthly =
Line 40: Line 43:


= Lightning talks =
= Lightning talks =
*
* [davehunt] LinkChecker


= Project status / goals for next week (keep it brief)  =
= Project status / goals for next week (keep it brief)  =
* High-priority/supported projects:
* High-priority/supported projects:
** Bouncer
*** WinXP aurora-stub problem https://bugzilla.mozilla.org/show_bug.cgi?id=1239685
** Marketplace / AMO
** Marketplace / AMO
** MDN
** MDN
** Mozilla.org
** Mozilla.org
*** All Selenium based tests have been migrated to the Bedrock repository
*** All Sauce Labs jobs in Jenkins for mcom-tests have been deleted
** MozTrap
** MozTrap
** One and Done
** One and Done
** Socorro
** Socorro
*** Stage and prod *should* be stable. The team has been working hard to get these envs back up to spec
*** Dec '15 post-mortem of the outage http://pad.mocotoolsprod.net/p/postmortem-dec-2015-4
** SUMO
** SUMO
* Available for ownership:
* Available for ownership:
Line 58: Line 67:
= CI updates =
= CI updates =
If you've worked on Jenkins or Selenium Grid in the last week, add the necessary info in the Wiki
If you've worked on Jenkins or Selenium Grid in the last week, add the necessary info in the Wiki
*  
* {{bug|1238977}} - Squid proxy timeouts between webqa-ci1.qa.scl3.mozilla.com and saucelabs.com


= Community update =
= Community update =
Line 73: Line 82:


= Action items =
= Action items =
*
* {{prev|}}[kthiessen] Try to update the default version of Appium we use to 1.4.16
* {{prev|}}[stephend] Create an etherpad and send out a note about brainstorming ideas for Q1 goals
* {{prev|}}[stephend] Finalize the "year in review" blog post
** team members: please populate the etherpad at https://public.etherpad-mozilla.org/p/webqa-year-in-review-2015
* {{new|}}[davehunt] Determine who owns http://restmail.net/
* {{new|}}[mbrandt] Upgrade Selenium to 2.49 with help from [davehunt]
* {{new|}}[davehunt] Add colours to Jenkins!
* {{done|}}[bsilverberg] Swap with [stephend] for upcoming buildmaster rotation


= Next Meeting =
= Next Meeting =
* Next owner / scribe:
* Next owner / scribe: Matt
* Next week's meeting notes:
* Next week's meeting notes: Madalin
Confirmed users
2,197

edits