Introduction
Assignment: Explore how we can integrate the JPivot package with DHIS 2.
*JPivot has a taglib that communicates with and executes MDX queries against Mondrian. It also uses something called WCF (Web Component Framework). That means that the JPivot pages has to be made with JSP. A taglib to write Velocity markup in JSP do exist if we want to use that (do we?).
*The Mondrian server takes care of aggregating data (group by's) for cubes and storing fact-tables using the database. So'll have to talk ''directly with the database''.
If this is undesired, a solution might be to replicate the database (like, perhaps, every night) so we never touch the production database with our code. This also allows for more tuning of the database
dedicated to the Mondrian server (like making materialized views (preaggregated data)).
Also, Øyvind pointed us in the direction of that we might have to make a tool on the web for generation the schemas(xml) Mondrian use. Are we? Or will will predefined schemas for the most common reports suffice?
Development
Developers
| Name |
Mail |
| Thomas Aure (thomasau) |
thomasau.på.ifi.uio.no |
| Odd Christer Brovig (oddcb) |
oddcb.på.ifi.uio.no |
| Tore Vatnan (torevatn) |
torevatn.på.ifi.uio.no |
Developer meetings
Weekly meetings at MML, Wednsdays at 16:00 am.
Progress / roadmap
Our project source code can be checked out with:
Project Documentation / Research
Links
JPivot/Mondrian
Database
