indexing a dataset in power bi cloud for faster processing

[sandeep540]

If I upload a dataset with 40 columns and 1 million rows, which is output of joining 5-6 tables in db! Once published the dataset resides in cloud for mashup!

Now I want to make sure the report being generated using the dataset is rendered fast without any lag or buffering! For this is it possible to create my own index (or indices) on the dataset for quick processing!

Please share your thoughts!

[ashishrj]

@sandeep540 You cannot create indexes in Power BI dataset as in SQL DB but you can connect your datatset using direct query for faster processing and querying !

[ankitpatira]

@ashishrj DirectQuery doesn't mean faster processing and querying. It is mechanism to keep your data on premise if you can't publish to cloud. Using Import method would be faster as data is is cloud and cached which you don't get with DirectQuery.

[felyoubi]

Hi,

A dataset with 40 columns and 1 million rows is nothing to worry about. The Azure engine what poweres Power BI Service can process your data any level your want. However, there is a concern if you have all of that in our table, aggregating, grouping, etc... will take some processing time as with any tool. If this is the setup you have, I suggest you break the single table into multiple tables and map out the relationship. That will definitely help.

Regards,

Fahd

[ankitpatira]

@sandeep540 I don't think you can create indexes on the dataset that is published in cloud nor can you do that in power bi desktop. You may not have any performance related problems since your data is up in cloud ready to be visualised in cloud. However you will experience longer time in getting that data refreshed from data source.