Implement Pentaho CDE Dashboard using Btable and Mondrian

Use the Btable plug-in to create a Pentaho CDE dashboard with a mondrian datasource. Btable is a plugin available at the marketplace. More info: [login to view URL] Attached a screen-shot with the desired dashboard. A little explanation. On top you see multiple-select box's of what we call the 'hub hierarchy'. A Hub belongs to a community, belongs to a project, belongs to a Service provider. One community can have multiple hubs, one hub only belongs to one community. 1) Based on this hub hierarchy select-boxes the table should be filtered. 2) Preferably the Service provider is not shown as column in the table The box with dispensed is also multiple select with things (type logs) that can happen at a hub 3) this type log should be (and is) put as pivotting dimension (no work needed) The from and to select boxes are date select boxes 4) Apply date range selection to the table The date header is a single select box to select the granularity of the date columns. If 'year' only year is show, if 'year-month', year and month is show as table header and so on, desired selections: year, year-quarter, year-month, year-week, year-month-day. I think implementation should be based on year-quarter-month-day hierarchy and year-week hierarchy in date dimension. 5) implement date granularity selection The hub hierarchy select boxes are not correct at this moment. It only gives values in Project if you select a service provider, but it should alwas give the projects restricted to the selection you made on a higher level. So if jou select serviceproviderX and serviceproviderY, only the projects, communities and hubs from serviceproviderX and serviceproviderY could be selected and not those from serviceproviderZ. 6) Adjust the mondrian queries to display the correct projects/communities/hubs in the select boxes It should work on Pentaho CE 5.4 and (with a little effort) 6.x. Input needed?: * sample data * current mondrian model * current dashboard Possible (partly) related projects in the near future: * extend dashboard or create similar dashboard to drill down on user data (a user could be seen as level below hub, also time is needed) * performance improvement (at this moment we have >2.5M rows in fact table, we have some aggregates for another dashboard) Possible solutions: -> splits fact table, create aggregates). But we could also use a clever mechanism to create those aggregates, etl is performed 3hourly, in future each hour.