Forum Discussion

kintera's avatar
kintera
Helper I
6 years ago

Model Level-of-Effort vs Performance

I have a good understanding of modeling/normalization...I've used Access for nearly 30 years, and currently write SQL.

 

I watched this video

 

2 ways to reduce your Power BI dataset size and speed up refresh

 

But don't think I need to go to that effort. 

 

Currently am getting data from an external system (via excel ...I have no direct connection to tables, etc).  It seems to focus on filesize, and mine is currently 836kb. My data lives in Sharepoint (I use Access to join with other data, and push to Sharepoint). Right now, I'm reporting on ~40k rows and it *seems* to perform well.

 

Is there an estimated number of rows where a flat file needs Modeling (cleansing/normalization)?

 

Any thoughts?

 

Thanks

Need Help

2 Replies

    • kintera's avatar
      kintera
      Helper I
      6 years ago

      Thanks, Cherry.

       

      I'm not having a performance issue. I'm a beginner, and hoping to deliver a product to my Users (~ 10 people) and don't want to get to a point where I AM having performance issues. I'm basically using a flat table, and some of the videos I've seen "guide" to normalize (that is create a "star schema") removing duplicates, etc. I wonder if I ever *could* have performance issues, and *believe this is tied to model size (number of relationships and table size), but am unsure. I am anxious to deliver my visuals, but don't want to get to a point where I have to recreate the model in a rush to deliver a (what is currently) monthly report. I am hoping to have live data (I'm currently moving data 1x/day), with the report being key to their business process.

       

      Am I more clear?... I hope


      Lex