The QUALIFY keyword is becoming more adopted in the SQL dialects of other cloud hosted analytic platforms and would be nice to see the same addition introduced into Fabric. QUALIFY simplifies syntax and structure when dealing with filtering on window function results. ... View more

If Fabric pipelines support array and object parameters without issue and a Spark Notebook within the pipeline can leverage the pipeline parameters for configurable notebook execution than it would be nice for both to support the same range of data types for their parameters.   This would be a QOL adjustment instead of having to pass strings that are parsed into json objects in the notebook. It would also allow, in a more complex workflow, the ability to cleanly use an array in a forEach loop in the pipeline and potentially handle it in a seperate fashion in a notebook. ... View more

Problem As of July 2025,  26 of 145 tenant settings are in Preview-and 12 ship Enabled by default.  Preview featuresfall outside Microsoft’s normal SLA, support and data-residency guarantees .  Some previews  are already scheduled for retirement yet still activate automatically, exposing organisations to dead-end features.  Admins have no master switch for turning off default features. So they must re-audit the portal after every monthly release. This is a  gap called out repeatedly in the community community.fabric.microsoft.com.  Impact Production workloads lose contractual SLAs the moment they touch an auto-enabled preview. Compliance and data-sovereignty promises can be broken without any conscious action by the customer. Security posture erodes: new, unvalidated endpoints light up before risk teams can review them. Admin overhead skyrockets; scarce BI teams spend time chasing settings drift instead of delivering value. Proposal Ship every preview feature with Default = OFF for all tenants and capacities. Expose a single “Enable Previews” master toggle so early adopters can opt-in deliberately. Push an Admin Center notification each time a new preview becomes available or an existing one flips GA/Retired. Benefits Preserves Microsoft’s valuable innovation pipeline while respecting enterprise governance. Eliminates accidental loss of SLA, support and compliance for mission-critical solutions. Saves administrators countless audit hours and reduces avoidable help-desk tickets. Builds customer trust. Admins remain in full control of when experimental code enters production. Ask Vote to set Previews = OFF by default and add a master opt-in switch. The change is minimal for Microsoft but delivers maximum risk reduction for every customer running Power BI and Fabric in production. ... View more

Currently service princaipals can't read and write to OneLake storage when Tenant Setting:  "Allow apps running outside of Fabric to access data via OneLake" is disabled   Enabling this setting to able access is enabling to much access to OneLake and a more restrcitive option is required. We need a similar setting to this one for OneLake storage access:  "Service principals can call Fabric public APIs"   eg "Service principals can access OneLake storage"   ... View more

Description: In Power Query Web, there's no graphical interface to edit the Table.ExpandTableColumn step — unlike in Power BI Desktop, where a gear icon lets users easily select or change the expanded columns. I propose adding the same feature in the web version to allow users to select columns without manually editing M code. Benefits: Consistent experience between Desktop and Web Easier use for non-technical users Fewer errors and improved workflow efficiency ... View more

Let's say I schedule a Notebook to run (either by Notebook schedule or Data Pipeline schedule).   However, someone else with edit permission on the Notebook can subsequently alter the source code of the Notebook.   The new code will be executed the next time the notebook runs on my schedule.   But, it will still run under my user identity, able to utilize all my permissions, even if the code was altered by someone else and I might not even be informed about this.   To avoid this source of potential confusion and security risk:   Please make it possible to "lock" a scheduled notebook run or data pipeline to a specific version of the Notebook.   This way, I can know exactly which source code gets executed when the notebook is run on my schedule (or as part of my data pipeline).   I also want the ability to easily update which version of the notebook that gets run. And an option to "always run the latest version". ... View more

I have a page with three visuals. The last visual has a drill through.  When the drill through is selected a new page appears.  When selecting the back button I would like it to return the third visual and not the top of the page.  ... View more

How are we supposed to do version control if there is no option to export to JSON like we had with Gen1? I don't want to commit changes using git integration. ... View more

It's fairly minor, but can the width and height boxes be swapped round? I always think of width first. Also the boxes below are horizontal and then vertical so it would be consistent. Sometimes I make mistakes because I naturally assume it's width first. Many thanks.     ... View more

WHY is the Ribbon not customizable in Power-BI? WHY is there no option to hide the AI if one does not want to use it?! ... View more

To have the ability to organize reports into folders within a single workspace and restrict access so that specific user groups can only view their designated folders, similar to how permissions are managed in SharePoint ... View more

Admins need to be able to see a summary of CU usage for Fabric Trial capacities. Just a maximum % used figure would be enough to plan for sizing a new Fabric capacity purchase.   Thanks. ... View more

Implementing this would make Direct Lake on OneLake a true superset of onSQL, combining its unique benefits with greater query reliability and flexibility. Currently, when using Direct Lake on OneLake, guardrails are enforced for each table (such as the 1.5 billion rows/F64 limit). When these limits are reached, queries cannot be executed due to memory constraints. This is a significant restriction, especially given the many advantages of building models on OneLake—such as applying OneLake Security and leveraging multi-artifact model creation. Although this is a common limitation of Direct Lake, in onSQL-based models, this issue was effectively mitigated through Direct Query (DQ) fallback—meaning that when Direct Lake could not handle a query, it would automatically switch to DQ and allow the query to continue. I propose enabling the same Direct Query fallback mechanism for Direct Lake on OneLake. If a table hits the memory or row limit and the query cannot be executed in Direct Lake mode, the system should automatically fall back to Direct Query. ... View more

To have an option to add in bulk columns from the dataset schema in bulk when in 'personalize this visual' mode, either by having 'select all' option in each table or by clicking multiple columns with control key (or any other control button)  ... View more

Even though Office 365 Outlook and Teams Activity are in preview, they should use parameterized Fabric Connections instead of the "signed in as" option. This removes the dependency on manually editing settings and allows the use of a service account. ... View more

While creating an app and adding the audience, all people who are added can view all the reports in single app curretly. It would be great if there is an ability to select audience to view only some reports in app and other audience can view other reports not all in same app. ... View more

When sharing reports and dashboards, I would like to have the option to control whether to display the option 'Allow recipients to share this report' based on administrative privileges. ... View more

Even users with read-only permissions can open the report and click “View semantic model” from the ellipsis menu at the top, which allows them to see the data content. From a security standpoint, we would like you to consider adding a feature that lets us disable “View semantic model” from the ellipsis menu. ... View more

Please add HA support for Azure Database for postgreSQL mirroring! We definitely will not be mirroring potsgreSQL to fabric from a production application database unless HA support exists. We can not have a critical application database with 99.9% uptime. HA brings SLA to 99.99% and is necessary for production applications.   For reference, High Availability is not supported.  https://learn.microsoft.com/en-us/fabric/database/mirrored-database/azure-database-postgresql-limitations   ... View more