Optimizing VNet Data Gateway CU When Loading Tables from SQL MI

[Shubham_rai955]

I’m currently loading tables from Azure SQL Managed Instance (SQL MI) using Dataflow Gen2 in Fabric, but this setup requires the VNet data gateway and is consuming high Capacity Units (CU). To optimize resource usage, I’m considering switching to the Copy Data activity in a Fabric pipeline to load all tables instead.

• Does the Copy Data activity help reduce or bypass VNet data gateway CU consumption compared to Dataflow Gen2?

• Is there a way to configure the pipeline to minimize VNet gateway usage, or does it always require the gateway for SQL MI private endpoint connections?

• Any practical tips on optimizing pipeline activity for SQL MI sources behind private endpoints?

Thanks for any recommendations or links to related Microsoft documentation!

[v-echaithra]

Hi @Shubham_rai955 ,

We’d like to follow up regarding the recent concern. Kindly confirm whether the issue has been resolved, or if further assistance is still required. We are available to support you and are committed to helping you reach a resolution.

Best Regards,
Chaithra E.

[v-echaithra]

Hi @Shubham_rai955 ,

I hope the information provided is helpful. I wanted to check whether you were able to resolve the issue with the provided solutions. Please let us know if you need any further assistance.

Thank you.

[v-echaithra · Accepted solution]

Hi @Shubham_rai955 ,

To check if the VNet Data Gateway is being used in the Copy Data activity, follow these steps:
Check the Linked Service Configuration. The Linked Service for the SQL Managed Instance in your pipeline should be configured to use the VNet Data Gateway.

Go to Manage.
Select Linked Services.
Find and edit the SQL Managed Instance linked service.
Ensure the Gateway Connection or VNet Integration is set up to use the VNet Data Gateway.

Thank you.

[GilbertQ]

Hi @Shubham_rai955 

Based on your information, it sounds like your SQL Managed instance is secured in the Azure network and that is why you need A VNet Data Gateway because you don't want it to travel over the public Internet. Using a copy data activity will certainly reduce your capacity unit consumption in both your Vnet gateway and your Fabric capacity. Ideally you could use the new incremental feature, or if the data is not large, just copy all the data into your lakehouse or warehouse and then query it from there.

[Shubham_rai955]

From where I can check wether the Vnet data gateway is being used while using copy data activity in fabric pipeline.