Skip to content

Microsoft Fabric

Engine: Spark · Format: Delta · Storage: OneLake (schema-enabled Lakehouse) · Status: ◑ Validated & deploy-ready — assign a Fabric capacity to run.

Assign a capacity, then run

Microsoft Fabric is capacity-based: every Fabric workload runs on an F-SKU (or trial) capacity assigned to the workspace. That assignment is a normal Fabric prerequisite the customer provides — so this provider ships validated and deploy-ready rather than pre-run: the RideFlow (like Uber) contracts are ported verbatim, schema-validated, and the OneLake + REST deploy path is wired. Attach a capacity and the same contracts run unchanged — Spark on Delta into OneLake, both already proven live on other providers.

Reference data mesh lakehouse: lakelogic-microsoft-fabric-data-mesh-lakehouse — the ported RideFlow mesh lives there: the OneLake + REST (az token) deploy path, and the OneLake Delta tables it materializes once a capacity is attached.

The same contract

The RideFlow contracts are ported verbatim from the data-mesh repo — no Fabric-specific edits. Backend choices:

materialization:
  strategy: merge
  format: delta          # OneLake Delta in a schema-enabled Lakehouse

Special configuration

How the LakeLogic framework adapts to this backend (handled for you):

  • Schema-enabled Lakehouse: tables live under schemas (domain-per-schema), matching the mesh layout; OneLake Files holds the landing CSVs.
  • Next step to go Live: attach a Fabric capacity and run the deploy path once to promote this from ◑ to ✅.