0

What I do:
I built ETL processes with power query to load data (production machine stop history) from multiple Excel files directly into PowerBI. On each new shift (every 8 hrs.) there is a new excel file generated by the production machine that need to be loaded to the data model too.

How I did it:
To do so, power query is processing all files found in a specific folder.

The problem:
During query refresh it need to process all the data files again and again (old files + new files).
If I remove the old files from the folder, power query removes the data also from the data model during the next refresh cycle.

What I need / My question:
A batch process copies new files into the folder while removing all the old files.
Is there a possibility to configure powery query in a way that it keeps the existing data inside the data model and just extend it with the data from the new files?

What I would like to avoid:
I know building a database would be one solution but this requires a second system with new ETL process. But power query does already a very good job for preprocessing the data! Therefore and if possible, it would be highly appreciated if this problem could be solved directly inside power query / power bi.

1 Answer 1

0

If you want to shoot sparrows with a cannon gun, you could try incremental refresh, but it's Premium feature.

In Power BI refreshing a dataset reloads it, so first it is cleared, and second - you will need all the files to re-load them and recalculate everything. If you don't want this, you have to either change your ETL to store the data outside of the report's dataset (e.g. a database would be a very good choice), or to push the data from the new files only to a dataset (which I wouldn't recommend in your case).

To summarize - the best solution is to build ETL process and put the data in a datawarehouse, and then to use it as a datasource for your reports.

Sign up to request clarification or add additional context in comments.

Comments

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.