Forum Discussion
Combining Data from Different Semantic Models
DPCCGF - I would suggest the dataflow option is your only viable solution. With one separate dataflow per client and a master dataflow that appends them all. This should ensure if one dataflow fails, your end dataflow will still succeed, it will just have out of date data for the one client until that connection is fixed.
DPCCGF - Whilst I would personally recommend sorting out the "complexity" in my experience this is usually rooted in bad practice, you could (in theory) set up 11 different models and then import them into one model using Composite models - you will not be able to append using this method however, but you could write DAX over the 11 different Fact tables and use some shared dimensions in the composite model - at a guess I would say performance of this would be terrible though.
https://learn.microsoft.com/en-us/power-bi/transform-model/desktop-composite-models
If you need to append things together into one table, I would set up 11 Dataflows and then use a master data flow that appends these into one.
https://learn.microsoft.com/en-us/power-bi/transform-model/dataflows/dataflows-create
If this guidanance helps, please accept as the solution for others with the same issue.
I can rephrase complexity. I have to connect to the souce via API to about 8 different data topics. So if it is one client I have 8 different APIs connections all for that single client and its token. Well, my current model replicates that 11 times as I cannot get to each client in whole. So all the data is brought in, then appended into one table. Then the appended 8 tables are joined to allow for reporting. I am open for suggestions.
DPCCGF - I would suggest the dataflow option is your only viable solution. With one separate dataflow per client and a master dataflow that appends them all. This should ensure if one dataflow fails, your end dataflow will still succeed, it will just have out of date data for the one client until that connection is fixed.