Snowflake mirror cost

Hi priyankabis 

When mirroring data from Snowflake to Fabric, there are indeed several operations that can incur compute costs on the Snowflake side. These operations are primarily related to the extraction of data (reading your Snowflake data to replicate it into Fabric's OneLake):
The size of the virtual warehouse and the duration it runs to execute queries for data extraction directly impact the compute costs. Larger warehouses or longer query execution times will increase costs.
Operations performed by Snowflake's cloud services, such as managing and optimizing queries, also contribute to the overall compute costs.
If Fabric frequently queries Snowflake to detect changes (deltas) in the data for replication, each of these queries will consume compute resources on Snowflake's side.

To manage and potentially minimize these costs, consider the following:
Ensure that the queries used for data extraction are optimized for performance. This can reduce the time the virtual warehouse needs to run, thereby reducing costs.
If possible, schedule mirroring operations during off-peak hours or less frequently, based on your data freshness requirements. This can help in managing the compute costs more effectively.
Keep an eye on the virtual warehouse's usage and performance. Snowflake provides tools and features to monitor and analyze warehouse performance, which can help in identifying opportunities for cost optimization.

For more detailed information on configuring Snowflake in a copy activity:

Configure Snowflake in a copy activity - Microsoft Fabric | Microsoft Learn

Best Regards

Zhengdong Xu
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.