MediaWiki

Wikimedia Release Engineering Team/Checkin archive/20141209

< Wikimedia Release Engineering Team | Checkin archive

2014-12-09

edit

Team Business

edit

Team Workboard

edit

https://phabricator.wikimedia.org/project/board/20/?order=priority

Scrum of Scrums

edit

https://phabricator.wikimedia.org/maniphest/query/iYosD4_9SSPW/#R

Vagrant

edit

https://phabricator.wikimedia.org/project/view/627/

  • Dan to email out the survey results to wikitech-l
  • next priority: unit test suite for MW-V

Browser tests

edit

https://phabricator.wikimedia.org/project/view/630/ Zeljko and I discussed making browser tests voting. 

* Would require a mediawiki instance capable of pulling a branch upstream from master *and* have Vagrant-style roles available.
** This might require a full-fledged Vagrant running on the instance (we'd need an LXC or Docker provider)

Deployment tooling

edit

https://phabricator.wikimedia.org/project/board/349/query/open/?order=priority

Continuous Integration

edit

https://phabricator.wikimedia.org/tag/continuous-integration/board/?order=priority

Beta cluster

edit

https://phabricator.wikimedia.org/project/board/497/?order=priority

  • New virt hardware is in and racked!
    We turn out to not know how to work these new HP boxes though, so there's lots of give-and-take with datacenter folks.
    I'll be able to lift quota restricitions just as soon as those boxes are up, which could happen at any minute :/
  • antoine: Could use a quota raise on integration project as well (Jenkins slaves)
    • andrew: yeah doable once we get more hardware (see above)
  • antoine: could use instances teamplates (more core but less mem)
    • andrew: doable, fill a task :)
  • antoine: what about specific virtual images
    • andrew: it is not automatized and adds burden to the OpenStack infrastructure. It is prefered to use base image + puppet.

Hiring

edit

Process discussion

edit

Thinking about process and (possibly) metrics. THIS IS A STRAW MAN BRAIN DUMP DRAFT

TL;DR RelEng team process and metrics should be on an axis of cooperation vs collaboration.

Collaboration is working with others to do a task and to achieve shared goals.
Cooperation is the process of groups of organisms working or acting together for their common/mutual benefit, as opposed to working in competition for selfish benefit.

(from Wikipedia)

Describing process and metrics for the ReEng team is challenging because of a) the diversity of individual activities and b) the nature of a support organization.

  • Apropos of a):
    • How is what Chris does similar to what Mukunda does similar to what Elena does, etc.?
  • Apropos of b)
    • If we work with the VE team, we use their process and conventions, with the Mobile team, their process, with Ops, their process, etc. etc.

Here is something we could say since we are a support organization:

  • We emphasize cooperation over competition. We want to cooperate with others in a useful way.
    • Example of increasing cooperation: not long ago only Gilles had contributed test code to the MultimediaViewer repo. Recently Chris has also begun contributing to MMV.
  • We emphasize collaboration over individual activity. We want to share our areas of expertise with others.
    • Example: Recently only Chris had contributed browser test code to the VisualEditor repo. Now Rummana has also begun contributing.
  • Bonus round: we create new ways to cooperate and collaborate
    • Example: Phabricator of course
    • Jenkins Job Builder
    • Moving from Cloudbees
    • Upcoming: e.g. make browser tests voting, needs collaboration and cooperation to make it happen

Counterexample:

  • If you are always working alone, you are failing at collaboration.
  • If someone else is doing work that you could help with, you are failing at cooperation

Goals: In the next time period:

  • Identify at least X number of projects to which you would contribute. (Cooperation)
  • Identify at least X number of projects projects with which you would share what you know. (Collaboration)

Metrics:

  • Since the last time noted:
    • Where have you contributed to a project or area new to you? (Cooperation)
    • How have you helped others understand the work you do? (Collaboration)

To increase Collaboration:

  • have both Rummana and Elena familiar with Jenkins

Vacations/Confs/Holidays/etc

edit
  • 12/10-12/17 - Elena's on vacation
  • 12/15-16 - Greg in Michigan (mostly out on Monday, somewhat out on Tuesday, will really really really try to be here for the meeting)
  • 12/20 - 1/4 - Greg on vacation
  • 12-22 to 1-06 Željko vacation
  • 12/24 - Chris vacation
  • 12/25 - 1/3 - Andrew B on vacation
  • 12/25-1/4 - Dan vacation
  • 12/25 - Christmas
  • 12/26 - Day after Christmas
  • 1/1 - New Years
  • 1/2 - Day after New Years
  • Antoine has no clue yet. Will be on call at worse.
Retrieved from "https://www.mediawiki.org/w/index.php?title=Wikimedia_Release_Engineering_Team/Checkin_archive/20141209&oldid=3612123"