DirectQuery report gets slower with time

[Svet_Dimov]

I have a report that connects to an Azure Databricks database using DirectQuery (as per customer requirements). They want to see real-time data so I have set automatic page refresh. I noticed that the report starts to slow down with time so I fired up the Performance Analyzer. What I noticed is that the Dax Query and Visual Display portions of each visual's refresh time are relatively stable but the Other keeps growing.  I tried to dig into the details of the Performance Analyzer export but I couldn't find anything that can tell me why this happens. 

Any advice will be appreaciated. 

[vanessafvg]

from https://www.sqlbi.com/articles/introducing-the-power-bi-performance-analyzer/

• Other: this is the time spent waiting for other operations to complete. This is typically synchronization time between different visuals in the same page and should not be considered as a bottleneck of the visual being analyzed. It is usually waiting time caused by other pending operations performed by other visuals.

[Svet_Dimov]

Thanks a lot for the reply! The thing is I am not trying to analyze the slowness of a single visual, rather the slowing down of the whole report. For this I cannot really disregard the Other portion.

[vanessafvg]

not sure what you are understanding from this but its saying that the other portion relates to the interconnectness of other visuals on the page. so would be best to understand what you are doing when you refresh it and how those filters impact the other visuals on the page.

[Svet_Dimov]

The report page is quite simple. It consists of the following:
1 table visual with 8 columns. They all come from the same Azure Databricks connection. The SQL for the connection is also very simple select of fields from two joined small streaming tables (inner join 1-1 relationship).

1 background image

8 text boxes/buttons with dynamic text (DAX measures that translate the report in two languages)

2 buttons with Web URLs.

No slicers are present in the page and no filters are applied anywhere. That's what makes things so confusing...

[v-saisrao-msft · Accepted solution]

Hi @Svet_Dimov,

Thank you for reaching out to the Microsoft Fabric Forum Community.

Thank you @vanessafvg, for providing your insights,

You're right to focus on the growing "Other" time in Performance Analyzer while it's usually tied to synchronization between visuals, in your scenario with Direct Query and auto page refresh, it likely reflects cumulative overhead affecting the entire report. 
One possibility is query overlap, where new refresh cycles begin before previous ones complete, especially if the interval is short. Increasing the refresh interval slightly (e.g., from 1s to 5s) may help. 
Additionally, the rise in other times may be due to client-side memory buildup, particularly from dynamic elements like multilingual text boxes. Monitoring Power BI Desktop’s memory usage over time using Task Manager could confirm this.  
I’d also suggest duplicating your report page and testing a simplified version without text boxes, background images, or buttons compare the two using Performance Analyzer to isolate the impact of visuals.  
Lastly, while your back-end tables are small, it’s worth checking with your Databricks admin to ensure sessions aren’t being repeatedly reinitialized or queued, as that could also contribute to latency. Hope this helps narrow down the issue. Feel free to share your findings. 

If this post helps, then please give us ‘Kudos’ and consider Accept it as a solution to help the other members find it more quickly. 

Thank you.