Govern and operate¶
Fabric becomes a durable retail platform when data, workspaces, semantic models, and deployment practices are operated as a connected system. The source material includes Power BI access and refresh guidance, GitHub integration, and deployment-pipeline examples.
flowchart LR
G[Git source control] --> D[Development workspace]
D --> T[Test or validation workspace]
T --> P[Production workspace]
P --> R[Reports and consumers]
R --> M[Monitoring, quality, and access review]
M --> G
Operating controls¶
| Area | Practice |
|---|---|
| Workspace access | Use least privilege, separate administrative and consumption roles, and review access regularly. |
| Data governance | Document business definitions, classify sensitive retail data, define ownership, and monitor data quality. |
| Power BI | Align semantic-model permissions, refresh ownership, incremental refresh policy, and report distribution to business requirements. |
| DevOps | Version workspace artifacts, validate changes before promotion, and use controlled deployment pipelines. |
| Cost and capacity | Monitor capacity consumption, schedule non-production use where possible, and set ownership for capacity decisions. |
Delivery path¶
- Connect a development workspace to source control and define the artifacts that should move together.
- Validate data quality, security, refresh, and semantic-model changes before promotion.
- Use deployment pipelines or an equivalent release process to move through environments intentionally.
- Monitor capacity utilization, refresh failures, data-quality signals, and report behavior after deployment.
- Keep recovery, rollback, and business-owner communication steps in the release process.
Note
A pipeline moves artifacts; it does not replace data-governance, security, or business validation. Keep environment-specific connections, credentials, and access assignments under explicit control.