semantic model
9 TopicsCan't export full .pbix + can't publish after live connection workaround
Can't export full .pbix + can't publish after live connection workaround Hi community, I've been running into a frustrating chain of issues over the past few days and wanted to see if anyone else is experiencing the same. Issue 1: Can no longer export .pbix with data I have reports that I've successfully downloaded as .pbix (full copy with data) many times before. Suddenly, I'm getting the error: "We couldn't export to .pbix format." The only download mode that works now is "A copy of the report with a live connection to the data online" — the full data export fails every time. For context: - The report was originally published from Power BI Desktop (not created in the Service) - Large dataset storage format is OFF - No incremental refresh configured - The tables are not large at all, so dataset size shouldn't be a factor - No deployment pipeline involved Issue 2: Can't publish after editing the live connection .pbix After making report edits in Desktop using the live connection file, I tried to publish back to the service and got this error: "An error occurred while attempting to publish 'Report Name.pbix': The remote server returned an error: (404) Not Found." My questions 1. Is anyone else suddenly unable to export full .pbix files from reports they previously could download without issues? 2. Is there a known issue with the UK South region affecting exports and publishing? 3. Any workaround to publish a live connection .pbix back to the service? Thanks in advance, any help or confirmation that others are seeing this would be really useful!481Views0likes2CommentsShared Semantic Model - Connectivity Issue from different tenants
Hi Everyone, The company that I work, provide the customers workspace in which we publish power bi dashboard and semantics models, using Azure Databricks as the data source. The semantic model is build with Direct Query and the security level has been done in databricks. We grant access to the customers trough the Azure Portal and invite them as external users to our tenant. When the customers log into our tenant they are able to visualize our dashboards and use the semantic model in the power bi service (that works fine), but when they download the power bi and customize the view in power bi desktop they are not able to publish the report back to the workspace, and when they try to publish to their own workspace they are not able to access the data. The error message is that the file is published but the data is disconnected. We have enabled all the options bellow to "entire organization" Enable Cross-Workspace Semantic Model Usage Enable Semantic Model DiscoverabilitySolved3.4KViews0likes20Commentsreport data is not visible
I have created and share the report but end user is not able to see the data. there is no RLS implemented. user have read access to workspace,lakehouse,sql end point,semantic model and report Large semantic model format:yes model lineage: lkh->semantic model->report Underlying Error: Missing_References Activity ID: cdb86897-5cc9-444c-9e0b-70a86cbce09c Correlation ID: 760f05d0-7eb4-e805-cfac-4ce83e7301a7 Request ID: 319ee4a9-fee8-4d82-4186-c08c69663697 Time: Fri Feb 20 2026 08:52:12 GMT+0400 (Gulf Standard Time) Service version: 13.0.27650.51 Client version: 2602.1.27898-train3.9KViews0likes6CommentsRequest for Improved Report Publishing Experience with Direct Lake Semantic Models
I’m encountering a significant limitation when working with Power BI reports connected to Direct Lake semantic models (via Fabric Lakehouse). After migrating a dataset from an Import model to Direct Lake using ALM Toolkit, I can no longer replace the existing report in the Power BI Service by publishing from Power BI Desktop — even though the dataset is correctly updated and the report is connected to the same semantic model. Issue: Publishing from Desktop only replaces the dataset, not the report. The report visuals remain unchanged, even though the PBIX contains updated pages. This behavior is inconsistent with reports using Import or DirectQuery models, where publishing replaces both dataset and report. Manual editing in the Power BI Service is not a viable alternative due to limited functionality and the complexity of the report. Deleting and re-uploading the report breaks bookmarks, links, and Analyze in Excel connections, which is unacceptable in a production environment. Request: Please consider enabling one or more of the following: Allow full report replacement from Desktop for reports connected to Direct Lake models, with a warning if bookmarks or personal views exist. Provide a “safe replace” option that updates report visuals while preserving user bookmarks and links. Improve documentation and tooling around managing reports connected to shared semantic models (especially Direct Lake). Enable Deployment Pipelines or ALM Toolkit to update both dataset and report in a single operation for Direct Lake scenarios. This limitation is a major blocker for teams adopting Fabric and Direct Lake, and it undermines the benefits of centralized semantic models. Thank you for considering this feedback.Solved612Views0likes3CommentsJson error while refreshing semantic model
Hi guys, I'm getting the following (not very descriptive) error while refreshing my semantic model in fabric PBI: Data source error: We found an unexpected character in the JSON input.. The exception was raised by the IDataReader interface. Please review the error message and provider documentation for further information and corrective action. Table: Web Site Data. Cluster URI: WABI-EUROPE-NORTH-B-redirect.analysis.windows.net Activity ID: eb32293c-4bc5-41a5-9980-7a28bc630e16 Request ID: c3de247d-980b-4f79-2e53-9cc75b832ec8 Time: 2025-04-30 15:03:41Z I'm not very experienced with PBI and there were no recent changes in this model, so I have no idea where to look... Any directions? Cheers, LuukSolved1.7KViews0likes4CommentsModify DAX function in semantic model
Hello community! I hope that someone can help me to resolve this issue that I'm facing. It seems very easy, but I'm not eable to resolve it. Well, the problem is the update of the new year in the power bi reports. The data is located on the sharepoint and can see the reports on the workscpace (Screenshot1) I duplicated the original report S to the C one. And I'm tryinf to download it, because I ca'nt edit the DAX or the functions in the "server" reports. There is no option to export in .pbix format and if I make a save as I cannot open it locally as it is saved in the workspace (screenshot 2). The second problem is that I have used a copy that I have locally to edit the function but I can't replicate the same function in the original report I created a new measurement but I can't configure it as a sum like this: In brief, what I want to do is very easy. I want to correct the function that is incorrectly filtering the year field. But I have not been able to download the report to edit it locally and republish it. If anyone can help me with this or if there is another way to solve it, I will be very grateful to receive your help. Thank you in advance!Solved671Views0likes2CommentsSemantic Model Refresh Error - "Reported a change in schema version"
I'm getting this error during scheduled refreshes: "The OLE DB provider \"MSOLEDBSQL\" for linked server \"ServerName\" reported a change in schema version between compile time (\"CompileTime\") and run time (\"RunTime\") for table \"\"DatabaseName\".\"dbo\".\"TableName\"\"."}}" Background: TableName is updated and reindexed nightly, but the process completes hours before the refresh. Using SQL Server 2019. Steps I’ve Tried: Restarting the on-prem gateway service after reindexing. Restarting the entire server. Updating Power BI queries from SELECT * to specific column selections. Any advice on resolving this error?Solved1.6KViews0likes3CommentsCalculating weekly running sum
Hello, my semantic model has the following structure: In the fact table FACT_REVENUE, for each snapshot_date/departure_date combination there is a revenue value. It has full revenue value only for Sunday snapshot dates (sunday_delta = 'sunday'). Snapshot dates for the rest of the week (sunday_delta = 'delta') are deltas to the previous day. E.g., the revenue for Wednesday, is a sum of revenue for all snapshot_dates since last Sunday to Wednesday (including Wednesday) for a certain departure_date. Dimension tables DIM_SNAPSHOT_CALENDAR and DIM_DEPARTURE_CALENDAR contain all snapshot_dates and all departure_dates, accordingly. I am using the following formula to calculate the weekly running sum of revenue: revenue_running_sum = VAR maximus = MAX ( dim_snapshot_calendar[snapshot_date] ) VAR totalSplit = CALCULATE ( SUM ( fact_revenue[revenue] ), dim_snapshot_calendar[snapshot_date] <= maximus, dim_snapshot_calendar[snapshot_date] >= maximus + 1 - WEEKDAY ( maximus - 7 ) ) RETURN SUMX ( ADDCOLUMNS ( VALUES ( dim_snapshot_calendar[snapshot_date] ), "@Value", 0 ), totalSplit ) First maximum selected snapsh-date is obtained (maximums). The sum is calculated for snapshot_dates <= maximus and snapshot_dates >= last Sunday. When filters on DIM_SNAPSHOT_CALENDAR[snapshot_date] and DIM_DEPARTURE_CALENDAR[departure_date] are applied, the results are correct: However, if I split the result by FACT_REVENUE[days_before_departure] (it is a difference between departure_date and snapshot_date), the results are split into each day since last Sunday: It seems to happen, because when calculating the running sum these snapshot_date/departure_date combinations are selected and when a filter from FACT_REVENUE table is in scope, then it is dispayed as split into those snapshot_date/departure_date combinations. The expected result is: Thus, the sum should be displayed for days_before_departure = 1/31/2024 - 2/7/2024 = -7. Would appreciate any help or suggestions. Thank you1.4KViews0likes4CommentsAllow measure expressions to be visible when sharing a semantic model
When a semantic model is shared in read only access, the user does not have the ability to to see the details of the dax expression for any measures created in the model. This makes it very difficult to understand how the measure is being used/functioning without seeing the expression. Since users also don't have read only access to the data, understanding how the model works also makes troubleshooting and creating new visuals difficult.1.1KViews0likes0Comments