Skip to content
CSA Loom — the Microsoft Fabric experience for Azure tenants where Fabric isn't yet available: lakehouses, warehouses, notebooks, semantic models, Activator rules, Data Agents, across Commercial, GCC, GCC-High, and DoD IL5

Tutorial 08 — Forward-migrate a lakehouse to Fabric (roadmap pattern)

When Microsoft Fabric reaches your audit boundary, you can let Fabric read your existing Loom ADLS data via a OneLake shortcut — zero data movement. 30 minutes (conceptual walkthrough).

No migration tool exists today

There is no Loom migration CLI or automated migration flow. This tutorial documents the manual pattern you will use when Fabric Gov GAs in your boundary. The conceptual model is sound today — your ADLS Gen2 Delta tables stay in place and Fabric reads them via a OneLake shortcut — but you perform the steps by hand in the Fabric portal.

Prerequisites

  • A Loom lakehouse with Bronze / Silver / Gold tables (from previous tutorials), backed by ADLS Gen2
  • A Fabric workspace (Commercial today; Gov when Fabric Gov GAs) bound to F-SKU capacity, where you're a Workspace Admin

Why this matters

The Azure-native Loom estate is intentionally portable: nothing is locked into a Fabric-only format. Because tables live as Delta in ADLS Gen2, Fabric can read them in place via OneLake shortcuts. Testing this pattern against a Fabric Commercial workspace today lets you validate the bridge before your Gov boundary's Fabric GA arrives.

Steps

1. Confirm the Fabric workspace is ready

Open the Fabric portal (https://app.fabric.microsoft.com Commercial, or https://app.fabric.microsoft.us when Fabric Gov GAs):

  • the workspace exists and is bound to F-SKU capacity (F8+), and
  • you're a Workspace Admin.

2. Create a Fabric Lakehouse

In the Fabric workspace, create a Lakehouse item manually.

3. Create a OneLake shortcut to your Loom data

In the Fabric Lakehouse → Files (or Tables) → New shortcut → Azure Data Lake Storage Gen2. Point it at your Loom ADLS Gen2 path:

abfss://<container>@<storage-account>.dfs.core.windows.net/Tables/noaa_silver_daily

Fabric now reads the Delta directly via the shortcut — no copy.

4. Verify the read in Fabric

In a Fabric notebook:

df = spark.table("workspace_name.lakehouse_name.noaa_silver_daily")
df.show(10)

The SQL Analytics Endpoint can read the same table. Results match what you see in Loom.

5. Reconnect the semantic model

For the model from Tutorial 03: in Power BI Desktop, reconnect (or re-author) it against the Fabric Lakehouse SQL Analytics Endpoint using Direct Lake on OneLake storage mode.

6. Compare freshness

  • Loom push dataset (Tutorial 03): rows arrive on an explicit push.
  • Fabric Direct Lake on OneLake: sub-second freshness off the same Delta files.

This is the gap that closes when you forward-migrate.

7. Run side-by-side

Don't decommission Loom immediately. Run both for 30-90 days: Loom keeps serving while Fabric reads the same data via the shortcut. Validate identity passthrough in both, compare results, and migrate reports report-by-report on your own cadence. The ADLS Gen2 data remains the single source of truth throughout.

What's next

Cleanup (if just demoing)

  • In the Fabric portal, delete the OneLake shortcut (and the test Fabric Lakehouse). Your Loom estate is untouched.

Troubleshooting

  • Shortcut creation fails: verify Entra B2B trust between the Loom tenant and the Fabric tenant (only relevant for cross-cloud hybrid; same-tenant works natively)
  • Result mismatch: Fabric reads the same Delta files, so schema and rows should be identical — re-check the shortcut path