Power BI (Cloud)

1

Connection validation succeeds, but crawling fails because required Power BI APIs cannot be accessed

Standard connection validation checks basic connectivity. It does not necessarily verify every API used during crawling. The Extended Validation option checks the APIs and permissions required by the Power BI connector, including APIs used for report metadata and PBIX processing.

Resolution:

• Enable Extended Validation when creating the connection to validate the required Power BI APIs and permissions.

• Review the validation result for API access or permission failures.

• Resolve the missing Power BI permissions before starting the crawl.

2

Could not establish connection

Connection validation cannot establish communication with Power BI. The issue can result from incorrect Azure AD credentials, incorrect tenant details, missing permissions, or network access restrictions.

Resolution:

• Verify the Client ID, Client Secret, Tenant Name, Tenant ID, and selected authentication method.

• For Username and Password authentication, verify that the account has the required Power BI access.

• For Service Principal authentication, verify that the app does not have any admin-consent required permissions for Power BI of type Application registered in Microsoft Entra ID.

• To check the assigned permissions:

  1. Sign in to the Azure portal.

  2. Select Microsoft Entra ID, then Enterprise applications.

  3. Select the application used to access Power BI.

  4. Select Permissions.

  5. Verify that there are no admin-consent required permissions of type Application registered for the app.

• Verify network or proxy access to the Azure AD and Power BI API endpoints.

3

CLIENT BRIDGE connection error

A Bridge connection error can occur when the Bridge agent is unavailable, the connection is not assigned to the required Bridge, or credentials cannot be resolved through the Credential Manager during crawl execution.

Resolution:

• Verify that the Bridge agent is running and that the Power BI connection is assigned to the correct Bridge.

• Verify that the required Credential Manager secrets are available and can be resolved by the Bridge agent.

• Run the crawl again after the Bridge or credential issue is resolved.

1

PBIX export not completing / "Layout file not found in the path" during regular crawl

The regular crawling job has a default export timeout of approximately 5 minutes. Larger reports (heavy visuals, large semantic models, or files exceeding 1 GB) cannot complete PBIX export within this window, so the layout file is not generated.

Resolution:

• Run the Advanced Job – Download PBIX and increase requestTimeOutInMillis (start with 15 minutes, then 30 minutes if still failing).

• Set "Retry for failed report" = true and increase the retry count (default is 3).

• If export still fails after 30 minutes, the PBIX file must be manually downloaded from Power BI and placed in the configured source-code file path.

• Refer to the System Settings and Advanced Jobs.

2

PowerbiNotAuthorizedException / authorization error while exporting PBIX for a specific report (others in same workspace succeed)

Power BI permissions can only be granted at the workspace level, not at the individual report level. This error typically occurs when a report's underlying semantic model resides in a different workspace than the report itself, and access has not been granted to that second workspace.

Resolution:

• Review the exact error text and compare it with the community-sourced error reference available in the Power BI Limitations documentation.

• Check the report's lineage and dataset details to determine whether the semantic model belongs to a different workspace.

• If the semantic model belongs to a different workspace, ensure access is granted to that workspace as well; if access cannot be granted, the PBIX file must be provided manually.

3

"Export model not found" during PBIX export

This is a known Power BI platform limitation, not an application-level issue, and occurs when the export API is unable to resolve the report's referenced semantic model.

Resolution:

Verify whether the report's semantic model is shared from a different workspace and confirm that access exists.

4

Not able to fetch the domain, report groups, or workspaces during crawling

The Power BI connection can be validated successfully, but crawling fails to retrieve available workspaces or report groups. This can occur when the required security group has not been given access to the required Power BI workspace. The security group must have access to each workspace that needs to be crawled.

Resolution

• Verify that the security group configured for the Power BI connection is added to the required workspace.

• Verify that the security group has Member or Contributor access to the workspace.

• Perform an extended validation to verify all the required Power BI APIs before clicking the Crawl/Profile icon. This helps identify API-related issues upfront and provides the exact error message to the user.

• Resolve any issues reported during extended validation.

• Run the crawl again after the workspace access is updated.

5

Report columns are not displayed after crawling

The Layout file may be successfully downloaded and processed, but report columns are not populated unless the Report Columns option is enabled in the crawler settings. The Layout file contains the visual and field information required to identify report columns.

Layout or report definition is not exported as expected.

Resolution:

• Open the Power BI crawler settings and enable the Report Columns option.

• Verify that the PBIX file and Layout file were successfully processed during crawling.

• Run the crawl again after enabling the Report Columns option.

6

Hidden pages are not available after crawling

Hidden Power BI pages are excluded when the Crawl Hidden Pages option is disabled. A hidden page is not normally visible to standard report viewers in Power BI, but its metadata can still be crawled when the option is enabled.

Resolution:

• Enable the option to crawl hidden pages in the Power BI crawler settings.

• Run the crawl again after enabling the option.

• Verify the hidden page metadata in OvalEdge after the crawl completes.

7

Unexpected automatically generated reports are displayed after crawling

Power BI can contain automatically generated reports that are returned through the REST APIs even when they are not normally visible alongside the other reports in a workspace. These reports can be excluded from crawling by enabling the Exclude Autogenerated Reports option.

Resolution:

• Enable Exclude Autogenerated Reports in the Power BI crawler settings.

• Run the crawl again to exclude the automatically generated reports.

• Keep the option disabled when automatically generated reports are required in the catalog.

8

Curation Status is empty for some Power BI reports

Curation Status values such as endorsement and certified by are obtained from Power BI ReportMetadata. Older reports or reports for which the required administrative metadata could not be retrieved can have empty values.

Resolution:

• Verify that the required administrative ReportMetadata API access is available.

• Enable Extended Validation and confirm that the ReportMetadata API is accessible.

• Re-crawl the affected reports after the required access is available.

1

What is the difference between "Data Set" and "Semantic Model"?

Data Set and Semantic Model refer to the same Power BI object. Microsoft changed the terminology from Data Set to Semantic Model in newer Power BI versions.

• Data Set: Term used in earlier Power BI versions.

• Semantic Model: Term used in newer Power BI versions.

• Both terms refer to the same type of Power BI object.

• Depending on the Power BI version, either term may appear in the application or documentation.

• When reviewing reports or lineage, Data Set and Semantic Model can be treated as the same object.

2

Why does the same column name appear more than once in a report?

The same column can be used in multiple visuals or pages within a Power BI report. As a result, the column may appear more than once.

To identify where each occurrence is used:

• Check the Page Name.

• Check the Sub-Report Name.

• Compare these values for the duplicate column entries.

• Each entry represents the use of the column in a different location within the report.

• Duplicate column names do not necessarily indicate duplicate source columns.

3

What is the Open Power BI Apps option?

The Open Power BI Apps option determines where the report opens when the Open Power BI option is selected from OvalEdge.

When disabled:

• The report opens directly in the Power BI report location.

When enabled:

• The report opens through the corresponding Power BI App.

This option is applicable when reports are published through Power BI Apps. If Power BI Apps are not used, the option can remain disabled.

4

What is the Premium Reports setting?

The Premium Reports setting is related to Power BI capacity used for report preview.

Most environments can initially use the default setting without additional capacity. Additional capacity becomes relevant when report preview reaches Power BI token limitations or when extensive preview activity is required.

Recommended approach:

• Keep the setting disabled during the initial connector setup when additional capacity is not available or required.

• If report preview stops working because the available Embed Token allocation is exhausted, evaluate the required Power BI capacity.

• After capacity is available, assign the capacity to the required workspace and update the applicable connector setting.

5

What metadata objects are crawled by the Power BI Cloud connector?

The Power BI Cloud connector retrieves metadata for the following Power BI objects:

• Workspaces, which are represented as Domains in OvalEdge

• Dashboards and their associated tiles

• Reports, including standard and paginated reports

• Semantic Models, also referred to as Data Sets

• Dataflows

The connector also extracts report fields and source information from the available Power BI source files and metadata.

6

Do we capture Power BI metric/KPI calculations and expressions, similar to what we support for MicroStrategy?

Yes, we are fetching this information at the semantic model level and adding the extracted details to the Formula and Technical Description fields.

7

What is the impact of renaming a Power BI workspace on curated metadata?

Renaming a Power BI workspace does not inactivate or delete curated metadata in OvalEdge.

• The workspace name is updated after re-crawl

• The external reference ID remains unchanged

• Existing objects remain active

• Curated metadata (tags, descriptions, glossary terms, etc.) is retained.

Recommendation: It is safe to rename workspaces at the source. However, it is recommended to validate the change in a development or test environment before applying it in production.

1

Why is Extended Validation recommended when creating the Power BI connection?

Extended Validation checks more than basic connectivity. Standard validation confirms that the connection can communicate with Power BI, while Extended Validation checks the APIs and permissions required during crawling.

Extended Validation is useful because:

• The Power BI connector uses multiple APIs during crawling.

• Report metadata requires additional API access.

• PBIX processing also requires additional API access.

• Missing permissions can therefore remain undetected during basic validation.

• Enabling Extended Validation during the initial connection setup helps identify missing permissions before crawling starts.

2

Does the Power BI Cloud connector require a Bridge?

No. The Power BI Cloud connector does not require a Bridge.

The connector communicates directly with the Power BI service.

• A Bridge does not need to be selected for the Power BI Cloud connector.

• PBIX files can be stored in the configured file path when required.

• The Bridge setting should remain unselected for a standard Power BI Cloud setup.

3

What is the difference between Service Principal and Username and Password authentication?

The Power BI connector supports two authentication methods:

Service Principal

• Uses Client ID, Client Secret, and Tenant ID.

• The application identity is used for authentication.

• This method is generally preferred for production connections.

Username and Password

• Uses a Power BI account username and password in addition to the common connection details.

• This method can be used when account-based authentication is required.

• Some environments may also require Okta-based authentication with this method.

4

When is Okta authentication required for the Power BI connector?

Okta authentication applies to the Username and Password authentication method when the Power BI account uses Okta-based authentication.

Important points:

• Okta is not required for Service Principal authentication.

• Okta is relevant when Username and Password authentication is selected.

• The Okta option should be enabled only when the Power BI account uses Okta authentication.

• The required username and password must also be provided for this authentication method.

5

Why is a Tenant ID required for Service Principal authentication?

The Tenant ID identifies the Microsoft tenant in which the Power BI application and Service Principal are configured.

The Service Principal setup uses:

• Client ID to identify the registered application.

• Client Secret to authenticate the application.

• Tenant ID to identify the Microsoft tenant.

Multiple Service Principals can use the same Tenant ID when they belong to the same Microsoft tenant.

6

Does the Client Secret have an expiration date?

Yes. A Client Secret created in Azure can have a defined expiration period.

During Client Secret creation, an expiration period can be selected, such as:

• One year

• Two years

• Three years

The Client ID and Tenant ID do not have the same expiration behavior as the Client Secret. The Client Secret should therefore be renewed before its expiration to avoid authentication failures.

1

Why can PBIX files not always be exported successfully?

PBIX export can fail because of Power BI platform limitations that are outside the application's control.

Common reasons include:

• The report is very large and requires more time to export.

• The report's semantic model is located in a different workspace.

• The report falls under other Power BI export limitations.

If the PBIX file is not exported:

• Run the Download PBIX Advanced Job.

• Increase the request timeout, such as to 15 or 30 minutes, based on the report size and processing time.

• Enable retries for failed reports if required.

• If the report still cannot be exported, request that the customer manually download the PBIX file from Power BI.

• Place the downloaded PBIX file in the designated source code path for further processing.

2

What permissions are required for the Power BI workspaces that need to be crawled?

The security group configured for the Power BI connection must have access to the required Power BI workspaces.

Workspace access includes:

• The security group is added to the required workspace.

• Member or Contributor access can be provided for the required crawling operations.

• Access must be provided for each workspace that needs to be crawled.

• If a workspace is not included in the security group access, its reports cannot be retrieved during crawling.

3

Why does the Power BI connector download a PBIX file?

The Power BI REST APIs provide information such as workspaces, reports, and semantic models. However, visual information and visual fields are not available through the required REST API responses.

The PBIX file provides the additional information required for report columns and visual metadata.

PBIX processing works as follows:

• The connector downloads the PBIX file through Power BI APIs.

• The PBIX file is extracted during processing.

• The Layout file is retrieved from the extracted PBIX content.

• The Layout file contains metadata such as visuals and fields used in the report.

• Only the required Layout information is retained for processing.

• The PBIX file does not provide report data to the connector for this process. The processing uses the required metadata information from the Layout file.

4

What is the Advanced Job Download PBIX?

The Advanced Job Download PBIX provides an alternative PBIX export process for reports that require more processing time than the regular crawling process allows.

The Advanced Job supports:

• A configurable request timeout in milliseconds.

• Retry processing for failed reports.

• A configurable retry count.

• Processing of selected workspaces.

• Additional time for reports that take longer to export.

This job should be used when the regular crawl cannot obtain the PBIX or Layout file within the available processing time.

5

What should be done if PBIX files cannot be exported for many reports?

When multiple reports cannot be exported through the Power BI API, manually downloading every PBIX file may not be practical.

Alternative PBIX source options are available in the connector.

Available options include:

• Local Path: Stores PBIX files in a specified local folder.

• OneDrive: Retrieves PBIX files from a configured OneDrive connection.

• Azure DevOps PBIP: Retrieves Power BI project files maintained in an Azure DevOps repository.

The appropriate option depends on where the Power BI files are already maintained.

6

What is the difference between PBIX, PBIT, and PBIP?

These are different Power BI file formats that can be used for report development or source management.

PBIX

• Standard Power BI report file.

• Used as the primary file for extracting report metadata when PBIX export is available.

PBIT

• Power BI template format.

• Can be used in scenarios where a PBIX file does not provide the required Layout information.

PBIP

• Power BI project format.

• Can be maintained in a source control system such as Azure DevOps.

• The Power BI connector can retrieve the required project information from the configured Azure DevOps source.

7

Which visual types provide report columns?

Report columns are available for data-bound visuals that use fields from the semantic model.

Examples discussed in the training include:

• Pie Chart

• Column Chart

• Card

• Area Chart

Visuals that do not use data fields do not provide report columns.

Examples include:

• Text Box

• Image

• Other visual types that contain no semantic model fields

Therefore, the absence of columns for a Text Box or Image is expected. If a data-bound visual does not display its fields, the Layout file and Report Columns setting should be checked.

8

Why are all semantic model tables displayed even when some tables are not used in a particular visual?

A Power BI report can use a semantic model containing multiple tables. A particular visual may use fields from only some of those tables, but the complete semantic model associated with the report is still available.

For example:

• A semantic model can contain eight tables.

• A particular visual may use fields from only two tables.

• The remaining tables are still part of the semantic model.

• OvalEdge retrieves the semantic model information associated with the report.

• Lineage can then identify the relationship between the report, visual fields, semantic tables, and source objects.

9

Can hidden Power BI pages be crawled?

Yes. Hidden pages can be crawled when the Crawl Hidden Pages option is enabled.

When enabled:

• Hidden page metadata is retrieved during crawling.

• Hidden pages can be available in OvalEdge even though they are hidden in Power BI.

• The setting is useful when complete report metadata is required.

When disabled:

• Hidden pages are not crawled.

• Only pages available through the normal report structure are processed.

10

What are autogenerated reports, and can they be excluded from crawling?

Power BI can return automatically generated reports through its APIs even when those reports are not normally visible with the regular reports in a workspace.

The Exclude Autogenerated Reports option controls whether these reports are included.

When enabled:

• Automatically generated reports are excluded from crawling.

When disabled:

• Automatically generated reports can be retrieved along with the other reports.

The option is disabled by default.

11

What is Delta Crawl in the Power BI connector?

Delta Crawl processes only reports that have changed instead of processing the entire Power BI connection again.

The connector compares the report modification information from Power BI with the corresponding information already available in OvalEdge.

Delta Crawl is useful when:

• The connection contains a large number of reports.

• Only recently modified reports need to be processed.

• A complete crawl would require unnecessary processing time.

This can reduce the amount of processing required for subsequent crawls.

12

What is the purpose of the crawl scheduling time zone?

The crawl scheduling time zone controls the time period during which the Power BI crawl runs.

This is useful when Power BI data is refreshed during specific periods and crawling should not run at the same time.

For example:

• Data refresh can be scheduled during business hours.

• Crawling can be scheduled during non-peak hours.

• The configured scheduling time zone determines when the crawl executes.

• Outside the configured crawl schedule, the crawler does not run.

13

Can PBIX files be read from a local path instead of being downloaded from Power BI?

Yes. The connector provides a local file path option for scenarios where PBIX files are already available locally.

When the local path option is selected:

• A folder path is provided for the PBIX files.

• PBIX files are read from that location.

• The regular PBIX download process is not required for those files.

This option is useful when PBIX files are maintained separately and are already available in a designated file location.

14

Can PBIX files be retrieved from OneDrive?

Yes. OneDrive can be used as an alternative PBIX source.

The setup requires:

• A OneDrive connection in OvalEdge.

• The corresponding OneDrive connection ID in the Power BI connector.

• PBIX files maintained in the applicable OneDrive location.

This option can be useful when PBIX files are already maintained in OneDrive instead of being exported during every crawl.

15

Can Power BI project files from Azure DevOps be used instead of PBIX files?

Yes. The connector supports the PBIP source option for Power BI project files maintained in Azure DevOps.

PBIP is useful when Power BI reports are maintained in a Git based development workflow.

With the PBIP option:

• Power BI project files are maintained in Azure DevOps.

• The connector retrieves the required information from the configured Azure DevOps source.

This provides an alternative when PBIX export is not suitable for the environment.

16

What is the purpose of the Read from NFS option?

The Read from NFS option is intended for a specific setup where Power BI report files are read directly from a configured file location instead of connecting to Power BI through the normal Service Principal configuration.

When this option is enabled:

• The regular Client ID and Client Secret-based connection is not used for reading the report files.

• A configured file path is used instead.

• The required PBIX file structure must be available at that location.

For a standard Power BI Cloud connection, this option should remain disabled.

17

What source information is extracted from Power BI reports?

The connector extracts source information required for report metadata and lineage processing.

• DataMashup provides M code information.

• DataModelSchema and DataModel provide semantic model information.

• Layout provides report, visual, and field information.

• Connections information provides source connection details.

• ReportMetadata provides additional Power BI report metadata.

These source artifacts are used to identify report fields and build supported lineage relationships.

1

Why do some tables not display any columns even though lineage is built successfully?

Some tables are created from calculated formulas or expressions instead of a direct source-column mapping.

In these cases:

• Columns may not be populated for the table.

• The exact resulting column name may not always be determined reliably from the formula.

• Showing an incomplete set of columns could cause confusion.

• The lineage can still be built correctly to the original source.

• Once the original source table is cataloged, the lineage can map to the appropriate source object.

Therefore, missing columns in this scenario do not necessarily indicate a lineage issue.

2

How much Power BI lineage is generated automatically?

Approximately 95% of Power BI lineage is generated automatically based on the supported M-code methods and functions.

The remaining lineage depends on the specific M-code patterns used in individual reports.

• Supported M-code patterns are processed automatically.

• Some reports may use less common or unsupported M-code patterns.

• Unsupported patterns can result in lineage gaps.

• Such gaps can be reviewed and addressed through an enhancement process.

• Therefore, a small amount of missing lineage does not necessarily indicate a connector failure.

3

What is Augmented Lineage, and when should it be used?

Augmented Lineage can be used when a report contains an M-code method that is not supported by the standard lineage process.

Standard approach:

• Identify the unsupported M-code method.

• Raise an enhancement request.

• The unsupported pattern can be addressed through a future update.

When immediate lineage is required:

• Augmented Lineage can be used as an alternative.

• It uses AI-based parsing to interpret unsupported code patterns.

• An LLM must be configured and accessible in the relevant environment.

• The LLM analyzes the unsupported code and helps generate the missing lineage.

4

Does Power BI Cloud support sub report lineage?

Sub report lineage is not fully supported by the Power BI Cloud connector.

• Lineage for supported report structures can continue to work normally.

• Missing lineage for sub reports can occur because sub report lineage is not fully implemented.

• The limitation does not indicate that the complete Power BI lineage process has failed.

5

How do we currently parse Power BI lineage - is it only through PBIX files, or also through PowerBI APIs

We are building the lineage using Power BI APIs. For source-system lineage, we fetch the required information from the Power BI Admin APIs, which provide details about the semantic model and M code used in the report. We also use PBIR and PBIX files to extract information about the pages and visuals and build the corresponding lineage.

6

Are there any known API limitations, particularly for column-level lineage? If so, how do we typically address those gaps? The client specifically mentioned that PowerBI API has known limitations for building lineage and asked how OvalEdge can overcome that.

A known limitation is that, while parsing the M code, we may not be able to identify the exact source information at the column level with 100% accuracy automatically. To address any missing lineage gaps, we request customers to identify the missing lineage and export the corresponding source code and share it with us through OE Labs. We then import the source code, analyze the patterns, and enhance our algorithm accordingly to accurately extract the source information and build the corresponding lineage.