Alternative to AAS Live Connection in Power BI for Weekend Accessibility

Hi rasika_pawar ,

Thank you for reaching out to Microsoft Fabric Community.

Here are few Approaches that will work:

 

1.Switch to Import Mode

Import data from AAS instead of using Live.Schedule refresh on weekdays.Users can view cached data even when AAS is paused.

2.Use Composite Model (Hybrid)

Combine Import and DirectQuery.Use Import for offline/weekend access.Use DirectQuery for real-time weekday data.

3.Migrate AAS Model to Power BI Dataset (Fabric)

Host the model in Power BI Premium/Fabric.Schedule refresh to keep data up to date.Avoids dependency on AAS uptime.

These approaches ensure your reports remain accessible on weekends without keeping AAS online.

 

If this post helps, then please consider Accepting as solution to help the other members find it more quickly.

Thank you.

 

Hi rasika_pawar 

 

Here are a few straightforward options:

• Import mode – Change the report to Import, run a refresh late Friday, pause AAS for the weekend, and users still see the data (it just won’t update until Monday’s refresh).

• Composite model – Keep the tables in Import for weekends, add DirectQuery back to AAS for weekdays. The visuals can swap between “cached” and “live” automatically.

• Move the model to a Power BI Premium / Fabric dataset – Same model, no separate AAS server. The dataset stays viewable even when capacity is paused and only spins up when it needs to refresh.

• Auto-start script – If you must keep AAS, an Azure Function can wake the server on the first weekend request and shut it down after idle time, though that first user will wait a minute or two.

Most people start with Import mode (fastest, zero extra cost) and adopt the composite or Fabric approach if they still need live data on weekdays.

Hi rasika_pawar 

You're facing a common limitation of using a live connection to Azure Analysis Services (AAS): when the AAS server is paused—for example, over weekends to reduce costs—Power BI reports that rely on the live connection become unavailable, as they require real-time access to the underlying data model. Since live connections do not cache data in Power BI, users see errors or blank visuals when the server is offline. To work around this while preserving data integrity and performance, one practical solution is to create an imported version of the AAS model in Power BI. This involves using Power BI Desktop to connect to the AAS model, import the data (instead of using live connection), and publish the report with a scheduled refresh during business hours (e.g., Friday night). This ensures that users can still access the report over the weekend with relatively fresh data, even when the AAS server is paused.

Another option, especially if you're transitioning to Fabric or Power BI Premium, is to migrate your AAS model to a Power BI semantic model (dataset) using tools like Tabular Editor or XMLA. Once hosted in a Premium workspace, the model becomes highly available and supports incremental refresh, scheduled caching, and shared semantic models—all without requiring the AAS server to remain online. This not only solves your weekend access problem but also simplifies architecture and future-proofs your setup.

If you need both live and offline options, consider publishing two versions of the report: one with a live connection for weekday use, and one with imported data as a fallback for weekends. The fallback report could even be embedded on a separate tab or shared via a bookmark, with a message directing users during AAS downtimes.

In summary, live connections depend on AAS being online, so using an imported dataset as a backup, migrating to Power BI Premium, or creating dual-report strategies are the best ways to ensure continuous access without sacrificing performance or data freshness.