MS Fabric workspace and Capacity Planning for Corporate and Consumer Aligned

[prathijp]

I’m working on a scenario that requires setting up multiple Microsoft Fabric workspaces to support both corporate and consumer-aligned needs:

1. Corporate Workspaces: Separate environments for Development (DEV), Testing (TEST), and Production (PROD).
2. Consumer-Aligned Workspaces: Multiple workspaces tailored to different business units or domains for exploration and analysis.

I am looking on how to plan capacity for each workspace, considering the following:

1. Data Migration:

   • Migrate an existing data warehouse (~50GB) to the corporate workspace.
   • Expected annual growth: ~20GB.
   • Target architecture: Medallion (Bronze, Silver, Gold layers).

2. Business Unit Requirements:

   • Each unit should have its own workspace to explore Fabric features and perform ad hoc analysis and reporting.
   • They may need read access to corporate PROD data.
   • Once their solutions are ready for production, they should be able to share or promote them to the corporate workspace.

3. User Base:

   • Estimated number of business users: < 30 across the organization.

4. Future Use Cases:

   • Potential for AI and ML workloads originating from the corporate workspace.

5. Security & Access Control:

   • Need robust access control mechanisms to restrict visibility and access between workspaces.

6. Capacity & Billing Strategy:

   • Should I consider purchasing multiple capacities?
   • What are the recommended billing options for this setup?

[Rufyda]

Hi
Glad that your query got resolved. Please continue using Fabric Community for any help regarding your queries.

[Mauro89]

Hello @prathijp,

I recommend also in addition to @burakkaragoz helpful content to go and check out the documentation here:

 https://learn.microsoft.com/en-us/azure/architecture/analytics/architecture/fabric-deployment-patter...

This one also helped me and my teammates on planning Fabric implementation for a enterprise customer. 

Best regards!

[burakkaragoz · Accepted solution]

Hi @prathijp ,

Here’s a basic setup that should work for your scenario:

Workspace Structure

• Corporate (DEV / TEST / PROD): Use one Fabric capacity, split by workspaces
• Business Units: Give each unit its own workspace under the same capacity

Capacity Planning

• Start with F2 or F4 depending on refresh frequency and AI/ML needs
• 50GB + 20GB/year is manageable under a single capacity for now
• Monitor usage with Fabric metrics — scale up if needed

Access Control

• Use shared datasets from PROD with read-only permissions
• Use workspace roles and OneLake shortcuts to control access
• Avoid cross-workspace editing — promote content instead

Billing

• Stick to one capacity unless you hit performance limits
• Use per-capacity billing for predictable costs
• Track usage per workspace with Fabric admin reports

Let me know if you want help estimating capacity SKUs based on workloads.

If my response resolved your query, kindly mark it as the Accepted Solution to assist others. Additionally, I’d be grateful for a Kudos if you found it helpful.

Translation and formatting supported by AI assistance.

[Mustafa-AlHunai]

Thanks for your comment you help me alot