Forum Discussion
PowerBI Service Dataflow & Datamart & PowerBI Desktop
Super User1) Give Incremental Refresh another try. It is designed to handle your scenario.
2) ditto
3) Datamarts are on the way out. If you really must, use dataflows (but even those don't seem to be needed in your scenario)
Hi Ibendlin,
Happy new year and thank you for your reply. I have consulted it only now. I have to manage this dataset in PowerBI Service (Dataflow or Datamart) in order to share it with others. RLS will be needed later. I just have problems merging huge datasets. I have several tables with 15M one and others around 2M. I need to merge so that we can apply RLS easily and surely. Maybe you can suggest another way ? I just cannot merge because of the heavy datasets. I need to do outer inn but it fails. When you say that neither dataflow nor dtamarts is required in my scenario, what do you mean ? What tool should I use in this case? Also incremental refresh requires a spesific day for saving data. I am afraid that some data will not be refreshed and maybe lost. I have a function which makes an API call for getting data for the day-2 (example: on the 4th of January it will get the data for the 2nd of January. So I don't understand what should I precice in the dataflow for incremental refresh. Thank you for your help.
- lbendlin
I would propose you use datasets. Yes, incremental refresh uses date ranges but you can fake these by mapping years to "first day of the year", for example.
- Hi Ibendlin, I already use datasets. The problem is that my datasets which I upload from another source don't register the update. They only show updated and appended information for today but then the next day, yesterday's information is lost and only a new one is appended and updated. How can I have a dataset registered in powerbi service and not taking it from my desktop source? Or should I created a Azure mysql database for that? Will in this case the information be updated each day and appended ? thank you
- lbendlin
Power BI is a reporting platform, it is not a data storage. You will want to store your data further upstream (for example in that Azure MySQL database you mentioned). The most appropriate architecture solution depends on your business scenario and your growth projections (scalability).