Edit

Share via

Troubleshooting digital twin builder (preview)

This article contains troubleshooting suggestions for digital twin builder (preview).

Important

This feature is in preview.

Get logs using Monitor hub

For general troubleshooting of your digital twin builder (preview) item, it can be helpful to view its logs in the Fabric Monitor hub. The hub is accessible through the Monitor button in the Fabric navigation bar on the left.

Screenshot of the monitor hub.

Note

You might have to expand the overflow menu at the bottom of the bar to see the Monitor button.

The Monitor hub shows run history for two types of jobs:

  • Digital twin builder flow items: These entries are the on-demand and scheduled executions for digital twin builder flows. They show a status of either Succeeded or Failed, based on whether the set of operations in the digital twin builder flow completed successfully.
  • Lakehouse item jobs: These entries represent the actual Spark executions associated with the flows. They nearly always say Succeeded, even if the digital twin builder flow itself fails. Errors are captured and written to a log, which digital twin builder flows read to determine whether to report an overall success for the operation.

For more information about the Monitor hub, see Use the Monitor hub.

Troubleshoot data missing from explorer

Entity instances and time series data missing from explorer

What to do when the Explore mode view appears empty after mapping data to entities:

  • Use the Manage operations tab to verify that mapping operations are completed successfully. If they failed, rerun them (starting with failed non-time series operations first, then time series operations).

    Screenshot of completed operations in the Manage operations tab.

  • If mapping operations were successful, there might be a delay in the SQL endpoint provisioning. Look for the SQL endpoint associated with your digital twin builder (preview) data lakehouse. The lakehouse and the SQL endpoint are located at the root of your workspace, and named after your digital twin builder instance followed by dtdm.

    Screenshot of selecting the SQL endpoint from the Fabric workspace.

  • If there's no SQL endpoint associated with your lakehouse, the lakehouse might have failed to provision correctly. Open the lakehouse and follow the error message prompts to recreate the SQL endpoint. For more information, see What is the SQL analytics endpoint for a lakehouse? - SQL analytics endpoint reprovisioning.

Entity instances missing time series data

What to do when an entity instance in the Explore view is missing time series data in the Charts tab:

  • The tab might appear empty because the time series mapping for the entity ran before the non-time series mapping completed. Try creating a new time series mapping with the same source table, and running it with incremental mapping disabled.

  • Time series data might be missing because the link property for the time series mapping doesn't exactly match the entity property. In the time series mapping configuration, verify that the Link with entity property fields have exactly matching values. If they don't, redo the mapping so that the values match.

    Screenshot of defining the link between entity property and time series.

Troubleshoot operation failures

What to do when you see failed operations in the Manage operations tab:

  • Select the Details link to open the operation details.

    Screenshot of the Details for a failed operation.

    The Runs tab shows the run history of the operation, along with the flow that contained each run. Flows titled On demand represent ad-hoc runs of the individual operation. Looking at the run history can help you identify whether it was the operation itself that failed, or a larger flow.

  • In the Runs tab, select the Failed status to see the error message, which gives you more information about the failure. For some error codes, see the following sections for specific troubleshooting steps.

  • If the failure message is empty, you can Create a support ticket for more help.

    Have the job instance ID ready. You can see the job instance ID for each run in the Monitor hub, by adding Job instance ID as a column.

    Screenshot of adding the job instance ID column in the monitor hub.

Error: Concurrent update to the log. Multiple streaming jobs detected for 0

This error message is caused by multiple instances of a mapping operation running concurrently. If you see this error with a Failed status, try rerunning the mapping operation.

Contact support

For information about self-help resources and support options in Microsoft Fabric, see Get help with Microsoft Fabric.

For information about creating a support ticket, see Create a support ticket.