rss.livelink.threads-in-node

Fabric Pipeline Error

odtJitendra — Fri, 18 Sep 2026 19:00:07 GMT

Today, we noticed that our production pipeline started failing with the following error:

Error:

BadRequest Error fetching pipeline default identity userToken{

"code": "LSROBOTokenFailure","message": "AADSTS50173: The provided grant has expired due to it being revoked, a fresh auth token is needed. The user might have changed or reset their password. The grant was issued on '2026-08-24T05:50:05.7907970Z' and the TokensValidFrom date (before which tokens are not valid) for this user is '2026-09-18T16:05:07.0000000Z'. Trace ID: 250ed633-08f0-4dcb-bad0-11596983e901 Correlation ID: 430f0833-fbc4-44f8-b19a-883ada2cad74 Timestamp: 2026-09-18 18:29:30Z",

"target": "PipelineDefaultIdentity-3e591cb6-2704-49f2-9efd-7196287c9feb","details": null,"error": null }FetchUserTokenForPipelineAsync

Has anyone encountered this issue before or have any idea what could be causing it?

Any suggestions on how to resolve this would be appreciated.

updating .sqlproj file (version issue with git)

justfetz — Fri, 18 Sep 2026 18:58:50 GMT

We recently had a couple issues syncing from main with INFORMATION SCHEMA errors.

When we reviewed the documentation Upgrade Fabric Data Warehouse System File Version in a Git Integrated Fabric workspace - Microsoft Fabric | Microsoft Learn, we noticed that we had two errors.

The first is that we were self referencing in a couple PROCS, eg WH.dbo.Table from current WH.

We fixed that problem, and th INFORMATION SCHEMA sync still persisted, but the self ref error went away.

Then we noticed that the sqlproj file is listed as a possible cause. We have the old preview version listed in git.

However, if you export the warehouse project, the actual wh sqlproj file shows up to date.

So there is a difference b/w git and what is actually in the workspace.

We have tried to PR the change to this file, update it directly in git, make a small change in the wh and commit directly to main to see if we could trigger the change. In the one case, we went into ADO, we updated this file directly, committed to it. It actually SHOWED as an update for a hot second, BEFORE converting to a change and the change was showing the old definition.

In that referenced article, we have tried a and b, and we cannot for whatever reason get this file to update in git, which then causes sync issues.

Does anyone have experience with this, a recommendation?

Thanks J.

Allow the target pipeline reference in Invoke Pipeline / ExecutePipeline to be set dynamically

JONATHANHAUN — Fri, 18 Sep 2026 18:29:47 GMT

Problem
Today, the Invoke Pipeline activity (and legacy ExecutePipeline) requires picking the target pipeline statically at design time — a fixed Workspace/Pipeline selection in the UI, which serializes as a literal referenceName GUID in the pipeline's JSON. There is no way to set this reference via an expression or parameter.

This breaks down for any generic, queue-driven orchestration framework where a "dispatcher" pipeline needs to invoke whichever pipeline a config table says to run next. Concretely: our framework maintains a metadata-driven pipeline configuration table (pipeline name plus source/destination/routing attributes) and a queue-processing engine that reads the next pipeline name to run and needs to invoke it generically. Every time a new pipeline is added to that config, we currently have to hand-add a matching case to a Switch activity, each containing its own static Invoke Pipeline/ExecutePipeline targeting that one pipeline by GUID. This is pure maintenance drift — the exact bug we hit was a Switch's defaultActivities firing ("A Proper PipelineName was not passed in") simply because new pipelines were added to the config but nobody remembered to add matching cases to the Switch.

What we tried, and why it's a worse trade-off than fixing this natively
We looked up the target pipeline's item GUID at runtime from a metadata table (populated by a helper pipeline that syncs item names to IDs) and tried invoking it via a WebActivity calling the Fabric REST API's job-trigger endpoint (POST /v1/workspaces/{ws}/items/{itemId}/jobs/instances?jobType=Pipeline) with a manual Until-loop poll to emulate waitOnCompletion. This technically works, but:

It requires setting up a separate OAuth2/Service-Principal-authenticated Web connection just to call Fabric's own API from within Fabric — extra infrastructure and credential management for something that should be native.
It's asynchronous by nature, so we lose native failure propagation (.error.message on the calling activity) and have to hand-roll polling and error-message extraction from failureReason.
Most importantly, we're not confident it preserves Spark high-concurrency session sharing (sessionTag) the way a natively-orchestrated Invoke Pipeline/ExecutePipeline call does, since the REST-triggered run has no recorded parent-child relationship in the orchestration engine's own run lineage — it's an entirely independent top-level run from the platform's point of view.

We also confirmed directly that ExecutePipeline's referenceName genuinely cannot be an expression today — attempting one causes Fabric's Git-integration deploy layer to fail with DiscoverDependenciesFailed, since it needs to statically resolve pipeline dependencies at deploy time.

Ask
Allow Invoke Pipeline's pipeline reference (and/or workspace reference, for cross-workspace scenarios) to be set via a dynamic expression — e.g., bound to a parameter, variable, or Lookup activity output — in addition to the existing static picker. This would let a single generic dispatcher activity invoke any pipeline by name/ID resolved at runtime, with all the native benefits (waitOnCompletion, error propagation, and session-pool participation) that a hardcoded reference gets today. If a fully dynamic dependency graph is a deployment/Git-integration concern, even a documented, git-integration-excluded "dynamic mode" (accepting that dependency discovery can't statically validate it) would be a major improvement over the current all-or-nothing static requirement.

Reset Bookmark Navigator Selection to the Default Published View on Browser Refresh

vivek12344 — Fri, 18 Sep 2026 17:48:55 GMT

In Power BI Service, when a user selects a bookmark through a Bookmark Navigator, the bookmark identifier is appended to the report URL. If the user refreshes the browser page, the report reloads in the previously selected bookmarked state instead of returning to the report's default published view. This occurs even when persistent filters are disabled and report consumers are prevented from saving personalized filter states.

Business Requirement: Customers would like an option that allows report authors or administrators to configure bookmark behavior so that browser refresh always returns users to the default published bookmark/view, regardless of the bookmark selected before the refresh.

Current Behavior: User selects a bookmark from a Bookmark Navigator.

Bookmark state is stored in the URL.
Refreshing the browser reloads the same bookmark.
Report does not return to the default published view.

Expected Behavior: Provide a configurable setting that allows one of the following:

Always reset to the default published bookmark when the browser is refreshed.
Ignore bookmark URL parameters on refresh.
Allow report authors to choose whether bookmark state should persist or reset after refresh.

Value of Enhancemen:Adding a configurable bookmark reset option would:

Improve user experience.
Provide consistent behavior with organizations' default-view requirements.
Reduce support incidents related to bookmark persistence.
Give report authors greater control over navigation behavior in Power BI Service.

Separate Interactive and Background Compute Allocation to Prevent Throttling of Live Reports

TilakArani — Fri, 18 Sep 2026 17:21:01 GMT

Problem Statement

When Fabric capacity utilization reaches or exceeds 100% in the last one-hour window, end users experience significant disruptions when accessing Power BI Fabric live reports. This occurs because scheduled jobs, on-demand jobs, and Spark sessions consume compute resources without differentiation from interactive workloads.

Current Impact:

Live reports fail to open or become unresponsive for end users
Interactive user experience is severely degraded due to background job consumption
Business-critical reporting becomes unavailable during peak processing times
No mechanism exists to prioritize user-facing workloads over batch operations

Proposed Solution

Implement a dedicated compute allocation model that separates interactive and background workloads within Fabric capacity:

Key Features:

Reserved Interactive Compute Pool
Allocate a guaranteed percentage of capacity exclusively for interactive workloads (e.g., live reports, ad-hoc queries, user interactions)
This reserved pool should be protected from background job consumption
Ensure interactive compute never gets throttled regardless of background job activity
Background Job Compute Management
Assign scheduled jobs, testing jobs, on-demand pipelines, and Spark sessions to a separate compute pool
Implement automatic job cancellation or queuing when background compute allocation is exhausted
Prevent background jobs from consuming resources beyond their allocated threshold
Intelligent Throttling Logic
When background compute reaches its allocated limit (e.g., 70-80% of total capacity), automatically:
Queue new job requests
Cancel or pause non-critical Spark sessions
Prioritize critical scheduled jobs based on defined priorities
Interactive workloads should remain unaffected and continue to operate within their protected allocation

Configurable Allocation Settings

Allow capacity administrators to define the split between interactive and background compute (e.g., 30% interactive / 70% background)
Provide flexibility to adjust allocations based on organizational needs and usage patterns
Enable time-based allocation policies (e.g., higher interactive allocation during business hours)

Expected Benefits

Improved User Experience: End users can reliably access live reports without interruption
Predictable Performance: Interactive workloads maintain consistent performance regardless of background activity
Better Resource Management: Clear separation enables more effective capacity planning
Reduced Business Impact: Critical reporting remains available even during heavy batch processing
Enhanced Control: Administrators gain granular control over workload prioritization

Business Justification

Organizations invest in Fabric capacity to support both analytical processing and real-time business intelligence. When background jobs monopolize resources, the primary business value—timely access to insights—is compromised. This feature would ensure that interactive reporting, which directly serves business decision-makers, receives guaranteed compute resources while maintaining efficient batch processing capabilities.

Additional Considerations

Provide monitoring dashboards showing real-time allocation and usage across both compute pools
Include alerting when either pool approaches its limits
Allow emergency override capabilities for critical batch jobs when needed
Consider implementing burst capacity options where interactive workloads can temporarily borrow from background allocation if available

Distance-Based Location Highlighting

Swami_om — Fri, 18 Sep 2026 15:28:14 GMT

Hello,

I’m trying to calculate the distance between the assigned location and the employee’s home location using the latitude and longitude available in this file.

The objective is to identify locations that are within the selected distance range (in kilometres) using a kilometre slicer. I would like the locations that fall within the selected range to be highlighted in green, while all other locations remain grey.

Additionally, I would like the selected location to be highlighted based on the selected City, Employee Location, or StoresDummy[City] value.

I’ve also referred to the following YouTube video for guidance: youtube

Could you please help me implement this logic in Power BI/DAX?

Sample data

Best Practices for Handling Incremental Data Loads in Dataflow Gen2

codeautomation — Fri, 18 Sep 2026 15:26:56 GMT

Hi everyone,

I am exploring different approaches for handling incremental data loads with Dataflow Gen2 in Microsoft Fabric.

For larger datasets, refreshing the entire dataset every time can become inefficient, so I am interested in understanding how others are designing their dataflows to process only new or changed records.

A few questions:

How are you identifying and filtering changed records between dataflow runs?

Is it better to manage incremental logic directly inside Dataflow Gen2, or use a Fabric pipeline to control the process?

How do you handle failed runs or partially processed data without creating duplicate records?

Are there any recommended patterns for maintaining good performance as the volume of historical data grows?

I would appreciate hearing about approaches that have worked well in real Fabric environments.

Designing AI Agent Workflows on Modern Data Platforms

codeautomation — Fri, 18 Sep 2026 15:21:22 GMT

Hi everyone,

I have been exploring how AI agents can work with modern data engineering platforms to automate business processes.

A common architecture I see is that data pipelines collect and prepare information, AI agents analyze the context and determine the next action, workflow services execute tasks through APIs or connected systems, and the results are stored for reporting or further processing.

I am interested in how teams are approaching this with Microsoft Fabric.

Are you using AI agents directly with Fabric workflows, or keeping the AI layer separate?

What patterns work well for connecting AI agents with Lakehouse, Data Pipelines, or notebooks?

How do you manage security and permissions when AI applications need access to enterprise data?

Are there recommended approaches for combining Fabric workloads with external AI services?

I would be interested to hear what architecture patterns others are using and what challenges you have encountered.

The permissions granted to user '' are insufficient

eduardserediuc — Fri, 18 Sep 2026 15:18:41 GMT

I installed Power BI Report Server – SQL Server 2025 Standard, product version 15.0.1121.109, on the primary node only of a Windows Failover Cluster SQL Server 2025 (no AG) – OS Windows Server 2025 Datacenter.
I configured the SMTP settings in the Email Settings section of the Report Server Wizard.
I created a shared PostgreSQL data source that connects using a username and password, and then created a paginated report using that data source.
The permissions for my AD user on the Report Server folder containing both the data source and report are all: Browser, Content Manager, My Reports, Publisher, Report Builder.
I created an Email distribution subscription for my report, but when I run the subscription either manually or on schedule I get the below error:

Failure sending mail: The permissions granted to user '' are insufficient for performing this operation.Mail will not be resent.

Can Power Automate Automatically Retry a Power BI Dataset Refresh Until It Succeeds?

manoj_0911 — Fri, 18 Sep 2026 14:32:32 GMT

Hi Power BI Community,

We currently perform a manual refresh of a Power BI dataset in our Production workspace. We would like to automate this using Power Automate.

Our requirement is:

Dataset: Recording
Workspace: Production
Required to be successfully refreshed before 8:00 AM EST every day
Start the refresh automatically, for example around 6:00 AM EST
Sometimes the refresh fails intermittently, so we would like Power Automate to retry the refresh if it fails
Once the refresh successfully completes, the flow should stop immediately and should not trigger any further refreshes
We don't need a success email/notification.
If the refresh continues to fail after a predefined number of retries, we would like to be notified.

Our intended flow is roughly:

Scheduled trigger – 6:00 AM EST ↓ Refresh Power BI Dataset ↓ Check actual refresh status ↓ Completed? ↓ YES ↓ NO / FAILED STOP Wait ↓ Retry ↓ Check status again ↓ SUCCESS → STOP

Questions:

Can this be implemented reliably using the Power BI connector in Power Automate?
Is there an action/API that allows us to check the actual dataset refresh status after triggering the refresh?
How can we make the flow retry only when the refresh fails or does not complete?
Once the refresh succeeds, how can we ensure the flow stops and does not trigger another refresh?
Is there any limitation on repeatedly triggering dataset refreshes through Power Automate that we should be aware of?
Would you recommend using the Power BI REST API instead of the Power BI connector for this scenario?
What Power BI and Power Automate licenses are required for this solution?

Any guidance or examples of a recommended Power Automate design would be appreciated.

Thanks!

Building an AI-powered quote engine on Microsoft Fabric

Avyaktha — Fri, 18 Sep 2026 14:15:20 GMT

The problem

Sales teams for technical products often get free-text enquiries like:

"I need 4 x 5T chain hoists, 10m lift, for offshore use"

Turning that into a quote usually means someone manually matching the product, checking the bill of materials, checking stock, checking supplier lead times, and pricing it up. I wanted to see how far Fabric's native AI capability could get toward automating that whole chain, without reaching for an external LLM service.

What I built

A proof of concept entirely inside one Fabric workspace:

A Lakehouse holding a synthetic but realistic-looking dataset: a product catalogue (chain hoists, wire rope hoists, lever hoists, beam clamps, trolleys), a bill-of-materials per product, inventory levels, supplier lead times, and pricing
A set of notebooks using Fabric's built-in AI Functions to read a customer enquiry and turn it into a structured, priced draft quote — matching the enquiry to a product, exploding the BOM by order quantity, checking stock, and rolling up a price and lead time
A simple interactive demo — type an enquiry into a text box, click a button, get a quote back — running entirely inside a notebook, no external app or Azure OpenAI resource needed

I deliberately made some components tight on stock (gearboxes, motors, brakes — the genuinely long-lead-time parts), so the demo shows a real "can we fulfil this now, or is it pending procurement?" decision, not just a happy path.

The interesting part

Real enquiries aren't always single, clean requests — customers often ask for a few different things in one message ("2 hoists, one standard and one offshore"). The AI extraction naturally returns one value per field, so it initially collapsed multi-item enquiries into a single line and silently dropped part of the request. Getting a reliable split-then-process approach working — without it silently swallowing half the enquiry — was the trickiest part to get right.

A couple of smaller but important lessons: null/missing values from the AI extraction need explicit handling in code rather than relying on prompt instructions alone, and discontinued or not-yet-launched products need to be filtered out at the matching stage so the system doesn't quietly quote something no longer sold.

Happy to go into more detail on the data model or the AI extraction approach if anyone's tackling something similar on Fabric.

Manage Roles - Row level data restrictions

Msampedro — Fri, 18 Sep 2026 14:10:12 GMT

Hi all,

Working on row level data restrictions. I do have a Budget table by Sales representative. Every sales person only can see their own data, plus two managers that they can see the sales representatives reporting to them and finally Director having access to all the data. Somehow it's not working.

I have created a "Security Table" with the users and Sales Rep ID, what indicates which users can see each one (i.e.; meaning Director can see Sales Rep ID 1,2,3 & 4. Manager1 sales rep ID 1 and 2, Manager2 sales rep ID 3 and 4 and finally Sales rep1 only can see Sales rep ID 1 and so on...

Model view below with Relationships are both one to many.

Formula DAX use in the manage roles

Export dashboards created in Power BI

Anas2 — Fri, 18 Sep 2026 13:15:38 GMT

Hello,

is it possible to export dashboards created in Power BI to a PowerPoint file or an interactive PDF ? if it's the case, how to do it ?
Thank's

Data Lineage support for Notebooks

JulianThomas — Fri, 18 Sep 2026 12:55:44 GMT

I would like to be able to utilise and benefit from the Data Catalog Data Lineage, even when I am not utilising the required Pipeline Activities that generate Data Lineage metadata.

I think this should be easy to do, if we can expose at least parts of the the api via python, so that when we are creating notebooks, we can setup similar metadata to our pipelines, allowing notebooks to be included in the data lineage. Possibly even allow this for SQL type code that could be executed as well.

Display Table[Column] instead of Column Name in Conditional Formatting Dialogs

Manickambeeie — Fri, 18 Sep 2026 12:03:31 GMT

In large-scale Power BI semantic models, it is common to have identical column names across multiple fact and dimension tables. Examples include columns such as Status, Category, Region, Employee ID, Created Date, and Priority.

When configuring Conditional Formatting (Icons, Font Color, Background Color, Data Bars, Field Value, Rules, etc.), Power BI often displays only the column name and does not clearly indicate the source table.

Displaying Table[Column] would improve report governance, debugging, and developer productivity.

Native SAP SuccessFactors and Active Directory Connectivity for Simplified Data Integration

João_Cristo — Fri, 18 Sep 2026 11:25:04 GMT

Having native connectors for SAP SuccessFactors and Active Directory would eliminate the need for workarounds such as notebooks, custom scripts, or complex integration setups. Users should be able to create a standard connection and use Copy Data pipelines directly, just like with other supported data sources. This would simplify data ingestion, reduce implementation effort, improve reliability, and make HR and identity data integration much easier to manage in Microsoft Fabric.

Error while exporting Excel using the "Summarize Data" option.

AAM_ — Fri, 18 Sep 2026 11:12:59 GMT

Hi,

When trying to export to Excel, some users are receiving the following error message:

No changes have been made to the report or to the underlying data model. However, starting today, some users have begun encountering this error.

The issue does not occur consistently, and it is not associated with any specific combination of measures and dimensions.

We are unsure what could have caused such a sudden issue.

thanks!

Allow Azure Databricks streaming tables to be mirrored into Fabric

_taab — Fri, 18 Sep 2026 10:13:41 GMT

Mirroring an Azure Databricks Unity Catalog into Fabric silently drops streaming tables. They are not greyed out, not flagged, not explained. They simply do not appear in the picker, so there is no way to select them. We have tested this and the documentation confirms it: materialized views and streaming tables will not be displayed.

The reason this is hard to accept is that the data already qualifies.

Mirroring for Azure Databricks does not copy anything. It creates OneLake shortcuts to the Delta data in place, which is exactly why Microsoft markets it as unified access without data duplication. So the only question that should matter is whether the object is a Delta table in Unity Catalog.

It is. Databricks documentation states that all Unity Catalog managed tables and streaming tables are Delta tables, and that a streaming table is a Unity Catalog managed table with extra support for incremental processing. Microsoft's own Learn documentation for Azure Databricks describes streaming tables as Lakeflow declarative pipeline datasets backed by Delta Lake. Two sets of Microsoft-published docs say these are Delta tables, and the third filters them out of the mirroring picker with no stated technical reason.

This is not a niche object type. Every output of a Lakeflow or DLT pipeline is a streaming table, and declarative pipelines are the pattern Databricks recommends for building curated silver and gold layers. The practical result is that the tables an organisation most wants to serve to Power BI, semantic models and data science workloads are precisely the ones mirroring refuses to show. A huge share of our own curated estate is unreachable from Fabric for this single reason.

The only workaround is to write jobs that copy every streaming table into a plain Delta table so the picker will accept it. Duplicate storage, ongoing compute to keep the copies current, latency added on top of pipelines built for incremental processing, and governed data existing twice with the second copy outside the lineage of the pipeline that produced it. We are paying, in cash and in engineering time, to reintroduce exactly the duplication that mirroring was built to eliminate.

The ask:
Include streaming tables in the mirrored catalog picker and shortcut them like any other Unity Catalog Delta table. If there is a genuine constraint, for example around managed storage paths or pipeline maintenance operations rewriting files, then say so in the documentation and give us opt-in support with that caveat written down. Right now the limitations page lists streaming tables alongside Delta sharing tables and RLS-protected tables with no explanation of why, and teams only discover the gap after they have committed to the architecture.

Fabric is meant to be where the whole estate comes together. For anyone running declarative pipelines on Databricks, this one exclusion decides whether that is true or not.

References:
https://learn.microsoft.com/en-us/fabric/mirroring/azure-databricks-limitations
https://learn.microsoft.com/en-us/fabric/mirroring/azure-databricks
https://learn.microsoft.com/en-us/azure/databricks/tables/delta-table

Power BI, import report and semantic model: show change connection pop-up

Reitse — Fri, 18 Sep 2026 08:57:51 GMT

When you need to import an existing Power BI report and semantic model in Tenant B that's created in Tenant A, you need to jump through a number of hoops to fix the original connection. It's doable, but overly complex.

Why not, on import, show a pop-up that offers to change the connection in the PBIX file to one that is available in the Tenant B (or is created when importing the report). It will save time, and frustration.

Refill Buffer Error while refreshing a table

ShreyasMungad — Fri, 18 Sep 2026 08:53:28 GMT

The power bi has tables connected to Databricks SQL Server

While refreshing 1 of the tables getting this error: OLE DB or ODBC error: [DataSource.Error] ODBC: ERROR [08S01] [Microsoft][Hardy] (115) Connection failed with error: Could not refill buffer

I have already tried to reduce the size of the output by reducing the date range in SQL Query, cleared cache. The only step applied in power query is converting text to date for a date column. This error has started occuring after I replaced a column with another in Query.

The query loads correctly in Databricks, the power query preview loads as well, the problem occurs only after "Close & Apply"

New to this forum, I have tried looking for solutions on Copilot, nothing helps if any ideas please let me know, I will be very grateful