Power BI和SQL Server索引

Question

I've done some research without getting valuable information about my question.

I'm working on a data warehouse project and one my customer's requirement is that to use power bi pro for data visualisation.

What is not clear to me is if power bi, while acquiring data in its data model, would or not benefit from the indexing structure developed in SQL Server.

Thank you in advance for recommendation/tips on this subject.

Answer 1

It somewhat depends on whether you are using a live connection.

Existing indexes may speed up data loading when using PowerBI in import mode where the data source is a view, query, or stored procedure.

They will also be used in Live mode when connecting to the above sources, and might be used when connecting directly to multiple tables.

As the comments state, if you are bringing entire tables into PowerBI with import mode, then the existing indexes will not benefit you, and the internal SSAS instance that PBI uses is a whole different kettle of fish.

One caveat is that columnstore indexes can be used to get around some of the data size limitations when dealing with the gateway as described here: https://community.powerbi.com/t5/Power-Query/Using-SQL-Server-with-Nonclustered-Columnstore-Index/td-p/563787 , but that's not directly related to your question.

Answer 2

Indexes help with retrieval speed on the server end. The answer to how much it will help depends on the specifics of your situation. If you are doing a lot of data transformation and mashup in the Power BI query editor, indexes will only help where there is a step that selects rows from the SQL Server. It won't help with steps where the processing is being done on the Power BI end (such as merging with data from an Excel file or adding custom columns or some forms of substituting values). However, since you mention a data warehouse rather than a simple database, I'm going to assume you're barely doing any transformation on the Power BI end, relying instead on the server end to do the heavy lifting. In that case indexes will definitely help speed things up if they're done strategically

There are some difference between Import mode and Connect live mode.

Import mode:

Data import can be used against any data source type, it can combining Data from different sources. Current Power BI service limitation published file size is 1 GB.

When using import, data are stored in Power BI file/service. Therefore, there is no need to setup permissions on data source side (service account for load is enough) and you can share data publically or with people outside organization. On the other hand, all data are stored on Power BI. It is supported to implement full DAX expressions and full Power Query transformations.

Connect live mode:

There are more limitations for live connection in place. It doesn't work against all data sources. Current list can be seen here , it cannot combine data from multiple sources.

You are also limited to just one data source/database you selected. You can't combine data from multiple data sources anymore. If you are connected to SQL Database, you can still create logical relationships between objects from that database as well as measures and calculated columns. When you are connected to SQL Server Analysis Services, you are limited just to report layout and even can't make calculated columns ,while you can only create measures currently. When using live connection, users have to have access to underlying data source. This means you can't share outside of your organization or publically. And It is not supported to implement full DAX expressions, only Report Level Measures, to learn more about report level measures, watch this great video from Patrick, and there is no Power Query transformations.

You can learn more: directquery-live-connection-or-import-data-tough-decision