Version 3 (modified by 13 years ago) (diff) | ,
---|
Notes from End-of-sprint demo 11.27, July 4 2011
Note: incomplete report.
Table of Contents
Report of Team 2, Sprint 7, 11.27
Avaialble 22.8 Planned 26.5
Story: directory structure.
Status: ongoing
Story: LifeLines import into Oracle
Status: done.
Report:
- import into EAV
- would be nice if EAV would be SQL again (for performance)
- made a view that makes it into nice table again (could make materialized view of it).
- importer of MOLGENIS was a bit slow
Question:
- how will the view be generated?? -> group by target, then for each target you
- materialized view takes??
Next isuess
- automatically do when new data set is published
Story: OSGi pilot
- Registry of bundles; bundle exposes service and components
- E.g. service for authorization
- Example using Knopflerfish
- need to define bundles, has Activator to start/stop/register
- need service interface (public) and service implementation (private)
Next stories:
- How to organize bundles properly.
Story: BBMRI catalogue user want simpler search and download
Each of the search options should be listed below each other. Database dummies cannot understand 'CONTAINS' and != IS NOT) Redesign search to show column specific searches per column and only any field on top. Use Excel type of search.
Next issues:
- Download gives weird exception.
Report of team 1
Martijn, George, Freerk, Martijn.
Most work has gone into validation of NGS result.
Question:
- what is different between protocol, workflow and computeapplication?
- whole pipeline is working for 16 minutes??? means manager is kept alive???
Comments:
- move import worksheet en import compute sheet both to lower level.
Next sprint:
- Create overview of data flow and organization in NGS/GCC
- Creating reporting
- Create batch submission
- Abhishek visits Ger
- Need to have smarter way to say that a job is succesful so we can drop intermediate data