Press "Enter" to skip to content

Comparing CSV to Parquet File Loading Performance in Power BI

Chris Webb has a comparison for us:

Earlier in this series on importing data from ADLSgen2 into Power BI I showed how partitioning a table in your dataset can improve refresh performance. In that post I used CSV files in ADLSgen2 as my source and created one partition per CSV file, but after my recent discovery that importing data from multiple Parquet files can be tuned to be a lot faster than importing data from CSV files, I decided to try creating partitions linked to Parquet files instead.

Click through for the experiment and its results.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.