Forum Discussion
Recommended Approach for Large Project
I had a similar project...about 50 tables, 800 measures, 50M records. I developed the data model in Azure Analysis Services and used that as the source for dozens of individual reports. Simplified maintenance since measures were all centrally located, performance was adequate (used S0 AAS so it was kind of a weakling, but later upgraded to S1). That may/may not be an option for you, but if possible it's worth looking at.
hey,
you in the rigth path actually, pulling as basic stuff, tranformation, filters etc on the data source to get a cleaner data its a good idea,
- having a main model as base for reports make big model mor managable:
this way you have a central model and them live conection to reports, (also in december release there a new capabilities that increase this topic features ( https://docs.microsoft.com/en-us/power-bi/connect-data/desktop-directquery-datasets-azure-analysis-services ) this add a whole new level to composite models),
- give all tables a adecuate and detailed description of what information its in it
- Give all columns a detail description of what means it have the data on the column (or at least the keys columns being used for modeling and visuals)
- Group Columns in folder in the data model relationship power bi desktop interface to have the information type in each table in order, this will help you during visuals creation, modeling and any collaborator working on Excel Conected to BI.
Gracias Stefano (de acuerdo a tu captura al parecer hablas español al igual que yo 😀 ).
Voy a probar el enfoque planteado y además voy a consultar con algún partner de Microsoft si puedo armar el data warehouse en la nube para obtener los mejores resultados.