The performance of the data pipeline hasn't improved even after upgrading the capacity

[Martin_Pham]

Despite upgrading the workspace capacity from F8 to F64, the execution time of my data pipeline has not improved at all.
Is there any way to check the actual runtime configuration or resource usage of the pipeline?

FYI: My pipeline only includes data copy and stored procedure that connects to a Fabric warehouse.

[spaceman127]

Hello @Martin_Pham ,

As far as the performance of your pipelines is concerned, you could use the Fabric Metric APP or FUAM.

FUAM is a project, you can find the repo here.

https://github.com/renefuerstenberg/fabric-toolbox

Have you ever checked this?

Then there are plenty of views you can use in the warehouse to analyze performance.

Here are a few views with explanations

View / DMV Description

sys.dm_exec_requests	Shows currently running queries, including status, duration, and resource usage
sys.dm_exec_query_stats	Contains execution statistics for cached plans
sys.dm_exec_sessions	Active sessions with login times and host information
sys.dm_exec_sql_text(sql_handle)	Displays the actual SQL text of a query
sys.dm_exec_query_plan(plan_handle)	Returns the execution plan as XML
sys.dm_pdw_request_steps	(Only in Synapse-compatible warehouses) Shows execution steps of a query
sys.dm_pdw_exec_requests	Query history with duration, status, and resource usage
sys.dm_pdw_waits	Details on wait times for individual queries
sys.dm_resource_governor_workload_groups	CPU/memory allocation per workload group

Best regards

[v-aatheeque]

Hi @Martin_Pham   Just following up on your question regarding pipeline performance despite upgrading from F8 to F64 capacity.

 @spaceman127  provided a detailed response covering key tools like the Fabric Metrics App and FUAM (Fabric Usage and Monitoring), including a GitHub link to help you get started with performance diagnostics. 

Additionally, a list of relevant DMVs (Dynamic Management Views) was shared to help you inspect execution stats, query behavior, and resource usage directly within your Fabric Warehouse environment.

To proceed, could you please confirm if you’ve had a chance to review or try any of the suggestions?

Happy to assist further if needed!

[v-aatheeque]

Hi @Martin_Pham 

Just following up on your question about pipeline performance after upgrading from F8 to F64.

@spaceman127  shared guidance on using tools like Fabric Metrics App, FUAM, and DMVs, along with a GitHub link to help with diagnostics.

Have you had a chance to review or try any of the suggestions?
Let us know if you need further assistance!

[v-aatheeque]

Hi @Martin_Pham 

Just checking that your query about verifying runtime configuration or resource usage for a Fabric pipeline was already addressed earlier in the thread.

As previously shared, capacity scaling (from F8 to F64) won’t necessarily reduce execution time for pipelines that are primarily running data copy activities or stored procedures especially when execution is dependent on the Fabric Warehouse’s backend performance or external system latency.

You can monitor your Fabric pipeline runs and capacity metrics via:

• Monitor Hub → Pipeline Runs for detailed activity duration

This helps confirm whether the runtime is truly resource-bound or limited by other factors.

looking forward for your response!!

[Martin_Pham]

It seems that the stored procedure activity in the pipeline does not use Microsoft Fabric's compute.
Is there any way to check the compute configuration and upgrade the compute resources of the Azure SQL, Azure Synapse Analytics, or SQL Server instance that the stored procedure is running on?