Forum Discussion

p_stathopoulos's avatar
p_stathopoulos
Frequent Visitor
9 months ago
Solved

Power BI Data Loading and Capacity Challenges, file over 1GB

I am currently using a free Power BI license and rely on Power BI Desktop to connect to a MySQL database. I utilize Import Mode with custom SQL queries (e.g., SELECT column1, column2 FROM table) to l...
  • ibarrau's avatar
    9 months ago

    Hi. First of all I would say the best option will always be optimice the data model. You can find amazing practices at youtube or articles talking about autodate time, remove unnecessary columns/tables, or even analyze the vertipaq storage to understand if big columns could be changed. Of course oneof the most impactfull options is thinking the data model again, is that the best star schema possible?

    After saying that, let's talk about the questions.

    1. Direct query will reduce the 1gb to less than 10mb for sure. That's because it won't store data. It will run queries against the source. In order to make it work properly you really really need to optimize the model and make sure your source can respond fast. Otherwise the user experience will be very slow. If you want to try this, please make sure to check a guide for direct query like: https://learn.microsoft.com/en-us/power-bi/guidance/directquery-model-guidance

    2. Premium capacity can't be purchased any more. Depending on the amount of users you would pick PPU or Fabric capacity. PPU is better if you are a small number of users, it can handle more than 1gb model and has many great features. Fabric will help when sharing with more users, you pay for capacity for a workspace. Depending if it's less than F64 SKU you would pay for pro users to view or not. Fabric pricing can be  yearly reserved or pay as you go. You can find prices just searching at web "Fabric Pricing".

     

    I hope that helps,