On Mar 3, 2014 12:02 AM, "Toby Negrin" <tnegrin@wikimedia.org> wrote:
> The development team is in the middle of the sprint that I sent you an update on last week; At the end of next week I'll update this list with what we accomplished during the sprint.
>
> For this week, here is the current status of the Epics we are working on:
>
> Context
> ======
> List of Epics is here.
> Analytics Mingle board is here
>
> Epics
> ====
> Production Issues
> Resolved on No sampled-1000 tsv file for 2014-02-06 on stat1002
> Wikipedia Zero team investigated ~30% increase of number of lines zero tsvs between 20140218 and 20140220 file
> Wikipedia Zero team investigated on light drop in zero requests around 2014-02-08
> Data for ULSFO Cache performance prepared for Ops blog post
> Metric Definition Standardization
> Current deliverables from Research and Data (New Editor Metrics) and Development (Page Views) internally reviewed and waiting start of community review process
> Accurate Pageviews for Wikipedia Zero
> Dashboards updated to cover all active partners
> Port of legacy processing scripts to Hadoop prioritized and will begin next sprint
> Editor Engagement Vital Signs
> Sync-up meeting confirmed metrics for first release (New Editors) finished by development
> WIkimetrics changes to support scheduled queries under development
> Met with FR Tech to discuss results of visualization library review/Wikimetrics -- loosely coupled collaboration seems likely on both.
> Metrics about Mobile Usage
> Prototype of mobile browser report running on unsampled mobile data on Hadoop
> Decision made to use UA Parser to parse out sanitized UA information for above report
> Worked on implementing Hive unit tests
> Simplify Limn Dashboard Deployment
> Proposal from research to simplify dashboard creation for certain "counting' use cases reviewed; will be assessed for next sprint.
> Page View API
> We have specs for new server from Henrik for stats.grok.se; Currently working with vendor to purchase.
> Worked on implementation of new JVM deployment system
> Worked on puppetizing Camus kafka-HDFS integration tool

> Worked on Kaftatee to support legacy analytics systems

To aid people searching archives for keywords in the future, the card is spelled Kafkatee.

-Jeremy

> Please let me know if you have any questions