Scrum of scrums/2015-05-13
Appearance
2015-05-13
[edit]Analytics
[edit]- Event Logging outage last week due to huge volume spike caused data loss. We're backfilling now.
- Event Logging won't consume your data into mysql until you have 100 events, so low throughput schemas won't show up for a while. A new patch (being tested now) will fix to consume at least every 5 minutes
- People can't test Event Logging in the beta cluster because /var keeps filling up on deployment-eventlogging02.eqiad.wmflabs. Yuvi enabled the /srv mount but we can't run puppet due to T96921. Need some ops help but Andrew is on vacation. In the meantime I dumped the log db, deleted the /var/lib/mysql/ibdata1 file and reloaded the log db backup.
Fundraising Tech
[edit]- New team members Dylan Kozlowski and Casey Dentinger are in SF this week! Come say hi! :)
- Last week:
- Progress on mw-vagrant fundraising role
- More queue and DonationInterface refactoring and minor improvements
- Progress on campaign mixins and banner history for CentralNotice
- This week:
- Onboarding/training
- Continuing above-mentioned Vagrant, DonationInterface and CentralNotice work
- Fixes for issues with some payment processors
- No blockers, no blocking of other teams' work expected
Search Team
[edit]- Updates
- Continued progress on carryovers
- Inching closer to WDQS 0.0.1
- Improvements to CirrusSearch testing
- Development of Maps demo
- Improvements to Graphs
- Interviewing/hiring
- Deployed Elastic plugins (extra, highlighter) to beta
- Continued progress on carryovers
- Blocked by
- Add fluorine rsync connector: https://gerrit.wikimedia.org/r/#/c/209684/1
- Hardware access request for OSM Maps project: https://phabricator.wikimedia.org/T97638
- Need clarity, possible consolidation, of the multiple EventLogging stores orbiting Search
- Blocking
Parsing
[edit]- Low activity week because of vacations, management-related workload, and OCG work.
- Parsoid deployments resume today after skipping a couple cycles.
- We are working through backlogged reviews now.
- Nothing in the way of blockers for other teams.
- Waiting on VE team to help review Marc's patches for css-based customization of cite.
Reading (Clients)
[edit]- What has your team done since we last met?
- Web: Deployed gather to EN Wikivoyage and hewiki
- iOS: Released 4.1.3
- Is anything slowing your team down or getting in their way?
- Apps: Lack of CI, but talking to RelEng about improving support & infra
- Are you about to put something in another team’s way?
- (More to come from Adam Baso, but will start training client devs to do work in PHP/JS for back-end requirements)
Security
[edit]- more patches on the cluster
- T90300 - Aaron?
- T98557 - Hardening for Citoid (ops)
- T98313 - Yuri
Releng
[edit]- Chad had passionate plea to check out 1.25 Mediawiki bugs
- Most of team gone next week for team offsite
Services
[edit]- commons now available through RESTBase: http://commons.wikimedia.org/api/rest_v1/?doc
- remaining public projects to follow
- OCG and cxserver switching to RB, will switch off Parsoid jobs once done
- continuing work with graphoid, mobile app service, hierator
- waiting for Parsoid patch https://gerrit.wikimedia.org/r/#/c/209253/ to be merged for section edit API
- heads-up on https://phabricator.wikimedia.org/T96903: group forming to work on content representation and -storage architecture, ask for input soon
Collaboration
[edit]- No new inter-team dependencies
- Waiting on "Perform schema change to echo_target_page changing from a 1 to 1 mapping between pages and user/notification to a 1 to many." ( https://phabricator.wikimedia.org/T94427 )
- Started actual conversion of LQT to Flow on MediaWiki.org. First couple pages are done. Will continue and get all done
- Continued work on moving header/description from top of board to side rail
- Started refactoring Flow front-end to use OOJS UI
Ops
[edit]- Unblocked LE on a large number of apertium package requests
- PHP-luasandbox is moving on
- Meeting about Maps hardware box and moving forward with a postgres labsdb