Star vs Snowflake and general help on a dimensional quandry!

• berniesprouster

  SSC Enthusiast

  Points: 124

  More actions

  July 8, 2013 at 2:49 am

  #277097

  Hi,

  I'm relatively new to OLAP/SSAS. I've designed a few cubes/dimensions and they've seemed to work quite well, but I'm in a bit of a quandry at the moment with a solution I'm looking to put in place.

  The solution currently uses SQL Server Change Tracking to iteratively populate a DW with records changed over time. It is a DW for an educational organisation and the solution centers around attendance statistics. The schema of the DW is briefly as follows:

  tblMark - contains the marks for the students (present, absent etc.) along with a key to the tblSession table.

  tblSession - contains details pertaining to the date/time on/at which the marks were taken, along with a key to the tblRegister table.

  tblRegister - contains details of the register taken (Register reference number, title, complete/incomplete).

  tblLecturer - contains 2 keys, linking to the tblRegister table and to the tblStaff table, enabling staff to be linked to registers.

  tblStaff - contains details of staff.

  What I'd like some advice on, if possible, is how best to go about designing an OLAP solution around this schema. I'd like at some point to be able to provide the business with the ability to define attendance statistics by staff, date/time etc.

  I've tried a few times to get to a snowflake design, using referenced dimensions, but seem to fail at after the tblRegister stage. I'm happy to flatten the DW if the best way of achieving it is to use a star schema, but just wanted a little advice on what the seasoned professionals would do.

  Thanks in advance :-),

  Mike.