Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!
Context (if you care)
I have built a Power BI report that queries data from many SQL Server databases at my organization. These are refreshed in a data flow. I then have a sharepoint list that i am using as a 'write back' table and this table is synced in another data flow.
My dataset then pulls both of these data flows and creates a merged table with some calculations, i then use this table to do a 'write back' based on certain events. It is all scheuled very carefully with power automate scheuled runs. It is designed so if an item misses a run its ok because it will be picked up in the next run.
My Problem
When i run my refreshes everything is good and my triggered events all work correctly. It will run perfectly for a couple hours. The flow then it will stop working randomly (Refresh works / no errors just my 'write back' stops). What happens is I go look at my sharepoint file and it has say 10 rows of data and then i go look at my flow and it also has the data and then i look at the dataset and it will only have 7 rows, it will continue refreshing correctly updating a bunch of other tables in the dataset. But never updating the table with 7 rows until i do it manually or download the pbix and republish and which point it starts working again like magic.
My thoughts
There is either some bugs with me doing a table in the powerbi data set which combines two datasets. And i should rather read the sharepoint tables directly into the dataset OR merge all the tables into a single dataflow.
Have you ever seen something like this?
Solved! Go to Solution.
It sounds like quite a complex solution and potentially not using the right tool for the job. I would suggest reading the files from SharePoint files or as you suggested merge them all into tables in a dataflow.
I always suggest using the right tool for the job and also make it as simple as possible.
It sounds like quite a complex solution and potentially not using the right tool for the job. I would suggest reading the files from SharePoint files or as you suggested merge them all into tables in a dataflow.
I always suggest using the right tool for the job and also make it as simple as possible.