Forum Discussion
Large dataset usage optimization
I'd reconsider "reducing the size is not an option" as there are many best practices to reduce Dataset size. Each time I thought I had reduced my larger datasets to their minimum, I learned or found another way to make it even smaller.
Outside of Dataset size, there are best practices for DAX Measures and Relationships that should also be carefully considered.
Exhaust all best practices provided in the link below. You can also use DAX Studio to help you implement these changes.
https://powerbi.microsoft.com/en-us/blog/best-practice-rules-to-improve-your-models-performance/
hi
Thanks for your quick response.
Something I guess I have forgotten to indicate in my post:
- there is only one table in the dataset (so no relashionship etc...)
- there is no calculations (so far)
I am basicaly just diplaying the data... That's why it is quite complicated to think and reduice the dataset.
Thanks
Okay, well even with 1 table there can be opportunity around field management, such as Data Types (Fixed Decimals, Whole Numbers, Date only vs DateTime), and removing unecessary columns. Across 200m Rows, those things add up signficantly.
Concerning the report, if there are no relationships and no measures, I'd think the only thing left is the request for data is too large to be efficient. What does the visual consist of? Is the visual displaying all 200m Rows?
High performing visuals are typically summaries of data with limited fields and rows, conversely the more fields and rows exposed to the User the worse it would perform.thanks for you reply.
Visual is mainly one table (6 columns) - with 6 filters at the top.
