article 23823

Linking Queries Run From DAX Studio To Workspace Monitoring And The Capacity Metrics App

A few weeks ago I wrote a blog post about how you can now link event data from Workspace Monitoring to data in the Fabric Capacity Metrics App using OperationId values. In the latest (3.4.0) release of DAX Studio there’s a new feature that you might have missed that link queries run from there to Workspace Monitoring and the Capacity Metrics App in the same way.

I connected DAX Studio to a published Power BI semantic model, turned on Server Timings and ran a DAX query. When the query finished I clicked the Info button at the top of the Server Timings pane, which opened the Query Information dialog shown below, and then I copied the Request ID value from there:

The Request ID is the same thing as the OperationId value in Workspace Monitoring and the Capacity Metrics app. I was then able to use this value in a KQL query in Workspace Monitoring like so:

SemanticModelLogs
| where Timestamp > ago(1h)
| where toupper(OperationId) == "5D4A4F47-370D-4634-B67B-E4B58CB067A8"
| project Timestamp, OperationName, OperationDetailName, EventText, Status, StatusCode
| order by Timestamp asc

[Note that you need to convert the OperationID value in Workspace Monitoring to uppercase to match it to the Request ID from DAX Studio]

What’s the value of doing this? Probably not much for an individual query because the information you see in Workspace Monitoring is exactly the same as you get in Server Timings in DAX Studio, and DAX Studio shows this information in a way that’s much easier to understand. If you’re testing a number of DAX queries, though, then having all this data available in Workspace Monitoring means you can do detailed comparisons of different runs using KQL.

What is really useful is being able to find the CU usage on a capacity of a DAX query run from DAX Studio – something that is very important when tuning DAX queries. While DAX Studio (and Workspace Monitoring) will give you a CPU Time value for a DAX query, for reasons I won’t go into here you won’t be able to reverse engineer the algorithm that converts CPU Time to CUs. However by linking the RequestId from DAX Studio to the OperationIds seen on the Timepoint Detail and the new Timepoint Item Detail page in newer versions of the Capacity Metrics app (as discussed here) you don’t need to care about that. You can simply take the Request ID from DAX Studio and find it in the Capacity Metrics app. For example on the Timepoint Item Detail (preview) page you can either find the value in the OperationId slicer or add the OperationId column to the lower table using the Select Optional Column(s) slicer and look for it there: