Hi @Johann_32 , Thank you for reaching out to the Microsoft Community Forum.
You need to automate the tracking of refresh status, refresh duration, row counts processed and downstream dependencies. There is no one place in Microsoft where you can do all this and APIs like Get Dataflow Transactions are incomplete. However, you can build a complete solution using Power BI REST APIs, Power Query (M) workarounds, Fabric Lakehouse, and Power Automate.
Start by retrieving metadata for all dataflows across your tenant. If you have admin rights, use the GET /admin/dataflows endpoint to fetch all dataflows at once. If not, you’ll need to iterate through each workspace using GET /groups and GET /groups/{groupId}/dataflows. Store each dataflow’s ID name and workspace ID into a Delta table in a Fabric Lakehouse. This gives you a scalable, persistent structure to hold your monitoring data.
For refresh tracking, call GET /groups/{groupId}/dataflows/{dataflowId}/refreshes to get the latest refresh history. This API returns the status, start time, end time and error message (if any). You can calculate the refresh duration by subtracting start from end time. Store this data in a separate Delta table and append each new refresh record. This lets you track long-running dataflows and identify patterns in failures over time.
Row counts aren’t exposed in any Power BI API, so the only accurate way to track them is by modifying your dataflows Power Query (M) logic. Inside the query, add a small metadata table that uses Table.RowCount to return the number of rows processed. For example:
let
Source = ...,
RowCount = Table.RowCount(Source),
Stats = #table({"Metric", "Value"}, {{"RowCount", RowCount}})
in
Stats
You can output this table alongside your data and read it into another dataflow or Power BI dataset. This is the only reliable and refresh-specific way to track row counts and you should apply it to high-priority dataflows first.
To identify which datasets and reports use each dataflow, use the Scanner API. This is the only method that provides a complete and accurate lineage map. Use POST /admin/scanner/scan to trigger a scan and GET /admin/scanner/scanResult/{scanId} to get the results. This will link dataflow IDs to dataset IDs and report IDs. Store this lineage mapping in another table in your Lakehouse. If you don’t have Premium or admin access, use the upstreamDataflows property from dataset metadata and cross-reference with reports, though this is less complete.
With metadata, refresh logs, row counts and lineage stored in Delta tables, you can now build a Power BI report that connects directly to your Lakehouse. Use a matrix or table to show each dataflow with its latest status, duration, rows processed and the names of linked datasets and reports. Include card visuals for failure counts, line charts for duration over time and filters for workspace and dataflow name. Highlight failed refreshes using conditional formatting.
Finally, use Power Automate to send alerts when a dataflow refresh fails. The “When a dataflow refresh completes” trigger lets you check the status, and if it’s “Failed,” you can send an email or Teams message with the failure details and downstream impact. This ensures your team can respond immediately.
If this helped solve the issue, please consider marking it “Accept as Solution” and giving a ‘Kudos’ so others with similar queries may find it more easily. If not, please share the details, always happy to help.
Thank you.
