Direct Lake - Power BI slow performance

[ramiromejia]

Hi everyone,

I've developed a workload where I've created a semantic model in Fabric, comprising three tables, each containing approximately 2 million rows. I'm utilizing Direct Lake to directly connect my semantic model to Power BI Desktop. This approach has streamlined the workload, as analysts can access the same semantic model I've established in Fabric.

According to Microsoft's recommendation, "Direct Lake mode can be the ideal choice for analyzing very large models and models with frequent updates at the data source." Therefore, I'm not concerned about the size of my model, as it's not particularly large.

Recently, analysts have noticed a significant decrease in the performance of some visualizations, with some taking considerable time to reload. My initial suspicion is that either the DAX expressions used are resource-intensive, or that my current license for Fabric may not be adequate for this data scenario. Currently, I'm using an SKU F2 license.

I am seeking advice on how to conduct a detailed performance analysis to isolate the impact of DAX expressions and Direct Lake on my Power BI dashboards. Additionally, I would like to understand if there are any constraints associated with my current license.

Any insights or recommendations on enhancing dashboard performance would be greatly appreciated.

Thank you for your assistance.

Ramiro

[v-shex-msft · Accepted solution]

HI @ramiromejia,

I think these should be related to your license. It not excited enough resource to handle with handle with huge amount of records and complex Dax expression calculations at the same time.

As the document listed, the F2 license should be the minimum level of the license table and only include 2 CU and 0.25 v-core which not able to process with heavy workload.

Microsoft Fabric concepts - Microsoft Fabric | Microsoft Learn

Regards,

Xiaoxin Sheng