Fabric Ideas Ideas

Expensive wrong click in lakehouse

jp-wair — Wed, 04 Nov 2026 14:08:30 GMT

If you click wrong in Fabric :

and you click on the eventhouse endpoint in a lakehouse, it will spin up the eventhouse which takes a lot of CU capacity which you can see in the metrics app (preview) that is a waste of CU's. would be nice if there is a warning when you want to click on the sql analytics endpoint and by accident press the eventhouse

Microsoft Tornado visual needs tooltip well and a Legend

cstjohn — Wed, 04 Nov 2026 13:56:17 GMT

I'm using the Microsoft Tornado 3.2.1.0 and can't believe it has been available for over 10 years and not updated to have a legend or tooltip well. You also cannot add a tooltip page - it is grayed out. I tried the sentence tooltip and it worked on the first hover but all sequential hovers showed literal text. If I hover outside and come back, the first hover shows the sentence tooltip but all sequential hovers show literal text again. These seem like simple requests, maybe not important but simple.

Add more audience keys: notebookutils.credentials.getToken

frithjof_v — Wed, 04 Nov 2026 12:27:25 GMT

Currently, only 4 audiences are listed in the docs:

Audience key	Resource	Use case
storage	Azure Storage	Access ADLS Gen2 and Blob Storage
pbi	Power BI	Call Power BI and Fabric REST APIs
keyvault	Azure Key Vault	Retrieve Key Vault secrets
kusto	Synapse RTA KQL DB	Connect to Azure Data Explorer

https://learn.microsoft.com/en-us/fabric/data-engineering/notebookutils/notebookutils-credentials?tabs=python#get-token

Please add Azure SQL Database to this list.

That would be very useful for Workspace Identity.

Thanks!

"Explore" setting default should be "disabled" of course

Antonio2026 — Wed, 04 Nov 2026 11:08:03 GMT

Is it possible to disable by default the Explore funcionality? Since we enabled it at tennant level, all datasets got the Explore functionality enabled by default. It should have been disabled by default at every dataset setting, end enabled when required. Do you agree? Otherwise without any alert all viewers becames enabled to self reporting de facto. Thanks

Custom visuals blurry in Power BI Service when report zoom is not 100% (CSS transform scaling)

ADG1999 — Wed, 04 Nov 2026 10:59:13 GMT

Problem

In Power BI Service, all custom visuals appear blurry whenever the report page is displayed at a zoom level other than 100% (e.g. Fit to page / Fit to width, which often results in ~87%). At Actual Size / 100% the same visuals are sharp. The issue does not occur in Power BI Desktop.

Root cause

Power BI Service applies a CSS transform: scale() to the report/page container. The browser composites the custom visual iframe as a bitmap before applying this page-level transform, so the visual content is resampled and becomes blurry. Diagnostics confirm the issue is not in the visual code: HTML, CSS, Canvas and SVG content all behave identically. In Power BI Desktop, zoom changes are handled by passing new viewport dimensions to update() instead of using a CSS transform, so visuals re-render at the correct resolution.

Selecting/clicking the visual may force a repaint in some cases, but it does not fix the issue while the page is rendered below 100%. The only reliable workaround today is keeping the report at Actual Size / 100%, which is not practical for most report consumers.

Requested fix (either option)

Pass the page scale factor inside VisualUpdateOptions, so custom visuals can render at physical pixel resolution; or
Implement page zoom without a CSS transform on the report container (e.g. pass the scaled viewport to update() as Power BI Desktop does, or use an SVG viewBox approach).

Original community thread: https://community.fabric.microsoft.com/discussions/power-bi-web-app/custom-visuals-appear-blurry-in-power-bi-services/5160950

Shape/ Image Tooltip

v-venkrachak — Wed, 04 Nov 2026 05:45:12 GMT

Currently, for Narrator/screen reader to announce “” on an image or shape, the Action property must be enabled to configure the tooltip.

When Action is enabled, Narrator announces “Bookmark dot Why this matters” instead of only “Why this matters.” This introduces additional information that is not required for the intended accessibility experience.

Since images and shapes already support tooltips, the enhancement request is to allow the tooltip text to be exposed to Narrator/screen readers without requiring Action to be enabled.

Expected behavior:
Narrator should announce only “Why this matters” when focus is placed on the image or shape, without adding “Bookmark dot” or requiring an Action configuration.

Enhancement request:
Provide support for accessible tooltip announcements on images and shaes independently of the Action property.

Allow Individual Activity Execution Within a Pipeline Without Running the Entire Pipeline

trmei — Tue, 03 Nov 2026 19:30:40 GMT

Problem Statement

Currently, when a Microsoft Fabric Data Pipeline contains multiple Semantic Model Refresh activities, selecting an individual activity and clicking Run triggers the entire pipeline rather than executing only the selected Semantic Model Refresh activity.

This creates an unnecessary limitation when developers or administrators need to refresh, test, troubleshoot, or validate an individual semantic model without triggering refreshes for all other semantic models in the same pipeline.

Current Behavior

For example, consider a pipeline containing five Semantic Model Refresh activities, each targeting a different semantic model.

When I want to refresh only one specific semantic model, there is no direct option to execute that individual activity. Clicking the pipeline's Run button executes all five refresh activities.

The current workaround is to deactivate the other activities, save the pipeline, execute it, and then reactivate the remaining activities. This is cumbersome, requires unnecessary modifications to the pipeline, and introduces the risk of accidentally leaving activities deactivated for subsequent scheduled executions.

Expected Behavior / Feature Request

I would like Microsoft Fabric to provide an option to execute an individual pipeline activity independently without triggering the entire pipeline.

Possible solutions include:

Run Selected Activity: Provide a right-click or context-menu option to execute only the selected activity.
Run From Selected Activity: Allow users to execute a selected activity and its downstream dependencies while skipping unrelated activities.
Debug Individual Activity: Provide an option to test or execute an individual activity without modifying the pipeline configuration.
Parameter-Based Execution: Allow users to select specific activities or semantic models to execute when manually triggering a pipeline.

Business Value

This enhancement would significantly improve the development, testing, troubleshooting, and operational management of Fabric Data Pipelines.

It would be particularly valuable for enterprise environments where a single pipeline manages refreshes for multiple semantic models and dashboards.

It would eliminate the need to deactivate and reactivate activities, reduce accidental refreshes, minimize unnecessary resource consumption, and provide administrators with greater flexibility and control over pipeline execution.

Request: Please consider introducing an option to run an individual activity, particularly Semantic Model Refresh activities, directly from the Fabric Data Pipeline interface without executing the entire pipeline.

Show Last Run, Status, and Next Run for Copy Jobs in the Workspace List View

JDownes — Tue, 03 Nov 2026 15:50:53 GMT

Problem
In the workspace list view, the Status, Refreshed, and Next refresh columns are filled in for Dataflows Gen2 and semantic models. For Copy jobs, they always show "—".

Copy jobs are the main way many teams handle ingestion in Fabric. In one of our workspace folders, 8 of the 10 items are Copy jobs, and the only item that shows a refresh status is the one Dataflow. Right now, to find out whether a Copy job ran or failed, I have to:

Open the Copy job
Click Run history
Wait for it to load
Check the latest run
Go back and repeat for every other Copy job

With dozens of Copy jobs across several folders, a quick daily health check becomes a slow, item-by-item task.

Proposed solution
Fill in the existing workspace list columns for Copy jobs the same way they're filled in for Dataflows:

Status: icon for the last run (Succeeded / Failed / In progress / Cancelled), with the error message shown on hover
Refreshed: start or end time of the last run
Next refresh: next scheduled run, if a schedule is set

Nice to have:

Click the status icon to go straight to that run's details
Show the same information for Pipelines and Notebooks, so the whole ingestion layer can be checked from one screen
Add a filter or sort by status, so failed items rise to the top

Benefit

Lets you open the workspace and see at a glance that everything ran successfully
Catches failures faster, before downstream reports show stale data
Makes Copy jobs consistent with how Dataflows and semantic models already appear in the workspace
Cuts down on extra clicks and page loads for admins and data engineers who manage many ingestion items

Adding component to Invoke Power Automate flow in Fabric

in_kkumar3 — Tue, 03 Nov 2026 15:39:46 GMT

hi,

we are using power automate for most of the report schedules.

it would be great in pipelines to provide power automate component so we can invoke the schedule post an event.

So on successful completion of an event (DB or File events) , we can invoke a report to run and share the results.

Thanks

FireWatch Connector Support

taylorc — Tue, 03 Nov 2026 11:35:48 GMT

Would it be possible to add support for a connector to the FireWatch Data Intelligence service?

The Power Apps/Automate custom connector for FireWatch's Data Intelligence service is rolling out. Microsoft Fabric / Power BI having the same capability would be very useful for FireWatch customers, as it would streamline their data analytics process.

Add a run-level failed jobs view to the new Monitoring hub

v-varunvv — Tue, 03 Nov 2026 10:13:08 GMT

The legacy Monitoring hub allows users to filter individual pipeline runs by Status = Failed within a selected period. In the new Monitoring hub, pipelines with an earlier failure does not appear if their latest run succeeded as the document says "The new Monitoring hub provides a table view of your pipeline items and their recent execution health. "

Please add a run-level view or an “Any failed run within the selected period” filter. This would help users quickly identify failed jobs without opening the Run history of every pipeline

Display Report ID and Report Name in the Fabric Capacity Metrics App for semantic model operations.

amanivannan — Tue, 03 Nov 2026 09:56:03 GMT

Currently, the Capacity Metrics App shows the Semantic Model ID when checking capacity usage.

If one semantic model is used by many reports and across workspaces, it is difficult to know which report caused the capacity usage or spike.

It would be helpful if the Capacity Metrics App could also show the Report Name and Report ID. This would make it much easier to identify the report responsible for the activity.

Fix incorrect datetime format in "Filter by last modified" tooltip

fti — Tue, 03 Nov 2026 09:48:35 GMT

It seems the datetime format in the Filter by last modified of some activities (Copy Data, Get Metadata and possibly others?) is incorrect (see screenshot below). Announced format is said to be:

yyyy-mm-ddThh:mm:ss.fffZ

(Notice the duplicated "mm"). I guess it is rather:

yyyy-MM-ddThh:mm:ss.fffZ

Scope "Copilot and Microsoft's MCP tools" semantic model setting to security groups

peterkach — Tue, 03 Nov 2026 09:33:47 GMT

The semantic model setting "Allow any person with only read permissions to use AI (Copilot, Microsoft's MCP tools)" is on/off for all read-only users. We need to allow AI/MCP access (Fabric IQ, Copilot, Data Agent) only for selected Entra security groups, like tenant settings allow.
The current workaround (turn it off and grant Build) gives users broader permissions than needed and hides the model from AI search even for allowed users.
Uncontrolled agent-generated DAX from many read-only users can put load on shared capacity, so per-group control is needed for governance and capacity protection.

Simplify Dataflow Gen1 to Gen2 migration for Viewer users

v-kkalvala — Tue, 03 Nov 2026 08:31:45 GMT

Enable viewer users to consume Dataflow Gen2 directly through the Dataflows connector or simplify Dataflow Gen1 to Gen2 migration from viewer perspective. This would provide a smoother Gen1-to-Gen2 migration, preserve existing consumption patterns, and reduce the need to repoint downstream solutions or manage additional permissions across Lakehouse and Warehouse objects.

Show "last refreshed" and "data freshness" status on every Power BI visual, with no DAX workaround

kotaruravitejas — Tue, 03 Nov 2026 08:30:14 GMT

PROBLEM:
Report viewers often can't tell whether the numbers they're looking at are current. Today, report authors have to build a workaround:

a hidden table with NOW() or a refresh-time query in Power Query
a card visual with a custom measure
manual placement on every page

Those workarounds are easy to get wrong. NOW() shows the time the report was opened, not the time the data was refreshed. A visual built on a DirectQuery or Direct Lake source can also be fresher or staler than the rest of the page.

SOLUTION:
Add a built-in, no-code option to the report canvas:

A "Data freshness" card or badge in the Visualizations pane that reads the last successful refresh time from the semantic model automatically.
A per-visual toggle in Format > General, "Show data freshness", which shows a small timestamp tooltip or footer on that visual.
Threshold alerts. The author sets a limit, such as "older than 24 hours", and the badge turns amber or red when the data goes stale.
Support for all storage modes: Import, DirectQuery, Direct Lake and composite models, so each source shows its own freshness.

Add a native "Data freshness" visual and per-visual toggle that automatically shows the last successful refresh time of the semantic model, with optional stale-data warnings, so authors no longer need DAX or Power Query workarounds.

Enhance Microsoft Fabric Learning with Real-World Projects and Hands-On Practice

ADABOINA_SAIRAM — Mon, 02 Nov 2026 21:37:07 GMT

I would like to suggest introducing more hands-on learning resources, real-world projects, and practical exercises across Microsoft Fabric.

It would be beneficial to have:

More end-to-end Microsoft Fabric projects.
Advanced SQL, PySpark, and Data Engineering challenges.
Real-world Power BI and Direct Lake scenarios.
More DP-600, DP-700, and DP-800 exam preparation resources.
Interactive exercises with step-by-step solutions.
Dedicated learning paths for beginners and experienced professionals.

These improvements would help community members strengthen their practical skills, prepare for Microsoft certifications, and become industry-ready.

Thank you to the Microsoft Fabric team for continuously improving the community and learning experience.

PowerBI Mailer Subscription

Jeff_Bennett10 — Mon, 02 Nov 2026 21:26:02 GMT

Received Therapy Emailer from powerBI and though the data was accurate (productivity, margin, co-treatment, etc) the comparison group data was incorrect which rendered the email info and suggestions invalid as it thought that my margin went up 933% which it obviously didn't do anything like that as well as a lot of negative % as there was no correct comparison data. This needs to pull from YTD values especially with newly acquired buildings that don't have data from over a year ago so we can truly see how things are going within our facility. If these changes are made, the power bi therapy mailer could be useful.

Multilingual Voice AI & Real-Time Lakehouse Integration for Elder Care & Accessibility (Saathi)

siddhant1407 — Mon, 02 Nov 2026 16:41:09 GMT

Idea Summary / Pitch: Millions of senior citizens and non-English speakers face a steep digital divide when using modern apps for healthcare, cab booking, meal ordering, and daily reminders. We propose Saathi — a voice-first, multilingual AI companion (Marathi, Hindi, Gujarati, English) that connects seniors to everyday services via natural speech, and streams interaction, medication adherence, and safety telemetry directly into Microsoft Fabric. Try our live prototype on Telegram by searching for @saathi12_bot. Experience voice-first elder care, daily medicine reminders, and emergency support in Marathi, Hindi, Gujarati, and English.

Proposed Solution & Fabric Integration:

Front-End Voice Assistant: Operates over accessible messaging channels (Telegram) with full speech-to-text and text-to-speech in regional languages. Handles medicine schedules, cab/food links, instant photo-to-PDF conversion, and built-in OTP security shields.
Fabric Real-Time Intelligence: Medication confirmations and alarm events stream directly into Microsoft Fabric Eventstream and KQL Database / Lakehouse.
Power BI Caregiver Dashboards: Families and clinics view real-time adherence rates, timeline trends, and proactive alerts triggered by Data Activator.

Unimatic Usage Snapshot

u055489 — Mon, 02 Nov 2026 14:13:19 GMT

In the Unimatic Usage Snapshot report, Can we that ability to drill down to the User file number? This way we can ensure that they are using UNI...