using PBI as data storage/Data repository

[MaryamSe]

Hi, 

My team is using SAS as datastorage. However, the licnces will not be renewed. I suggetsed to use PBI as a data strorage because 1) we do not have a quick solution to replace SAS  2) The amount of data for the last five years is around 300,000. so, not that large 4) PBI can handle all other things like appending all data, cealning and manupulation 5) all the alaysts will be able to connet to one sourec and create their reports 

my plan is to have the pbi in the worksspace that is availabe to anyone and connected to the execl sheets. ( btw, the data gets updated only twice a year)

I am not sure what is the best approch?

1) Should I create a sementice model that includes only the data and then uplaoded to the work space?  I do not think it is the best because then the sementic model should be on someone's computer 

2) Should I created Data flow? 

3) any other options ?

Thanks in advance!

[GilbertQ · Accepted solution]

Hi @MaryamSe 

For what you are suggesting I would suggest using Power BI desktop to consume the data.

The configure the semantic model with Incremental Refresh, here is how to do this:Incremental refresh for semantic models and real-time data in Power BI - Power BI | Microsoft Learn

And then upload and refresh the semantic model in the Power BI Service and then the users can connect via Excel.

[MaryamSe]

Hi, Thank you very much Gilbert. 

May I ask whay do you suggest PBI desktop. My concern is using PBI desktop needs to be on someone computer. I would prefer having a model that is shared on the workspace. 

Data are on several execl sheets ( more than one single source of data - I may need to appned them). The updated data include the new and updatde (overwriten) ones. But, It seems that incremental refersh requires using only one source of data.

Thanks again for your help

From the source you shared "When you configure incremental refresh and real-time data by using Power BI Desktop, or configure an advanced solution by using Tabular Model Scripting Language (TMSL) or Tabular Object Model (TOM) through the XMLA endpoint, all partitions, whether import or DirectQuery, must query data from a single source."