VS Code Workflow for Microsoft Fabric Development¶ Deep Dive

Last Updated: 2026-04-27 | Version: 1.0.0

Overview¶

VS Code is the recommended local development environment for this POC. The Microsoft Fabric VS Code extension (GA 2025) provides bidirectional sync between your local repository and Fabric workspaces, enabling a professional software engineering workflow for data engineering artifacts that are traditionally edited only in browser-based UIs.

This guide covers installation, workspace binding, the edit-sync-test-commit loop, and how to get the most out of VS Code when developing notebooks, pipelines, and semantic models for Microsoft Fabric.

Why VS Code for Fabric?¶

Fabric Extension Setup¶

Step 1: Install VS Code¶

Download from code.visualstudio.com. Use the System Installer on managed machines so updates apply system-wide.

Step 2: Install the Fabric Extension¶

Open the Extensions panel (Ctrl+Shift+X) and search for Microsoft Fabric:

Extension ID: ms-fabric.fabric-vscode
Publisher:    Microsoft

Alternatively, install from the command line:

code --install-extension ms-fabric.fabric-vscode

Step 3: Sign In¶

1. Open the Command Palette (Ctrl+Shift+P).
2. Run Fabric: Sign In.
3. Complete the browser-based Azure AD authentication flow.
4. Your tenant and workspaces appear in the Fabric sidebar panel.

Tip: If your organization uses Conditional Access policies, ensure your device is compliant before signing in. The extension uses the same token as az login.

Step 4: Verify Connection¶

After sign-in, the Fabric panel displays:

FABRIC WORKSPACES
  Tenant: Contoso (contoso.onmicrosoft.com)
    Casino-POC-Dev
      Notebooks (18)
      Lakehouses (3)
      Pipelines (5)
      Semantic Models (2)
    Casino-POC-Staging
      ...

Workspace Connection¶

Binding a Workspace to Your Repository¶

The Fabric extension maps a workspace to a local folder. This repo uses the following mapping:

E:\Repos\...\Suppercharge_Microsoft_Fabric\
  notebooks/          --> Workspace: Casino-POC-Dev (Notebooks)
  semantic-models/    --> Workspace: Casino-POC-Dev (Semantic Models)
  pipelines/          --> Workspace: Casino-POC-Dev (Pipelines)

To bind:

1. Open the Fabric sidebar panel.
2. Right-click your target workspace.
3. Select Connect to Local Folder.
4. Choose the repository root.
5. The extension creates a .fabric/ metadata folder (already in .gitignore).

Multi-Workspace Setup¶

For environment promotion (Dev/Staging/Prod), bind each workspace to a separate Git branch:

# Switch workspace context by switching branches
git checkout main              # Dev workspace
git checkout release/staging   # Staging workspace

Workspace Settings File¶

The extension stores workspace metadata in .fabric/settings.json:

{
  "workspaceId": "aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee",
  "workspaceName": "Casino-POC-Dev",
  "tenantId": "11111111-2222-3333-4444-555555555555",
  "itemMappings": {
    "notebooks/bronze/01_bronze_slot_telemetry.py": {
      "itemId": "item-guid-here",
      "itemType": "Notebook"
    }
  }
}

Important: The .fabric/ directory is listed in .gitignore. Do not commit workspace-specific bindings to source control.

Syncing Items Between VS Code and Fabric¶

Downloading Items from Fabric¶

Pull the latest version of all items from the connected workspace:

1. Open the Command Palette (Ctrl+Shift+P).
2. Run Fabric: Download All Items.
3. The extension writes Fabric item definitions to your local folder.

Or download a single item:

Fabric Panel > Right-click item > Download to Local

Uploading Local Changes to Fabric¶

After editing locally, push changes back:

1. Fabric: Upload All Items -- syncs everything.
2. Fabric: Upload Current File -- syncs only the active editor file.
3. Right-click in the Fabric panel and select Upload.

Selective Sync¶

For large workspaces, configure which item types to sync in .fabric/settings.json:

{
  "syncScope": {
    "includeTypes": ["Notebook", "SemanticModel"],
    "excludePatterns": ["*_archive_*", "*_deprecated_*"]
  }
}

Conflict Resolution¶

When both local and remote versions have changed:

1. The extension detects the conflict and shows a diff view.
2. Choose Keep Local, Keep Remote, or Merge.
3. For notebook conflicts, prefer Keep Local since your local version is under Git control and can be diffed precisely.

Local Editing Workflow¶

The recommended development loop for this POC:

                +-------------------+
                |  Edit in VS Code  |
                +---------+---------+
                          |
                +---------v---------+
                | Run local tests   |
                | (pytest + Spark)  |
                +---------+---------+
                          |
                +---------v---------+
                | Upload to Fabric  |
                | (extension sync)  |
                +---------+---------+
                          |
                +---------v---------+
                | Validate in       |
                | Fabric workspace  |
                +---------+---------+
                          |
                +---------v---------+
                | Commit + Push     |
                | (Git)             |
                +-------------------+

Step-by-Step Example: Editing a Bronze Notebook¶

# 1. Pull latest from Git
git pull origin main

# 2. Download latest Fabric state
#    (Command Palette > Fabric: Download All Items)

# 3. Edit the notebook locally
code notebooks/bronze/01_bronze_slot_telemetry.py

# 4. Run local tests
pytest validation/unit_tests/test_generators.py -v

# 5. Upload to Fabric workspace
#    (Command Palette > Fabric: Upload Current File)

# 6. Run the notebook in Fabric to validate against real data

# 7. Commit your changes
git add notebooks/bronze/01_bronze_slot_telemetry.py
git commit -m "fix(bronze): adjust slot telemetry schema for new SAS fields"

# 8. Push and create PR
git push origin feature/slot-schema-update

Notebook Format¶

This POC uses the Databricks notebook source format (.py files with # COMMAND ---------- separators). VS Code treats these as standard Python files, giving full IntelliSense and linting:

# Databricks notebook source
# MAGIC %md
# MAGIC # Bronze Layer: Slot Machine Telemetry

# COMMAND ----------

from pyspark.sql import SparkSession
from pyspark.sql.functions import col, current_timestamp

# Configuration
SOURCE_PATH = "Files/output/bronze_slot_telemetry.parquet"
TARGET_TABLE = "lh_bronze.bronze_slot_telemetry"

# COMMAND ----------

# MAGIC %md
# MAGIC ## Schema Definition

# COMMAND ----------

schema = StructType([
    StructField("machine_id", StringType(), False),
    StructField("event_type", StringType(), False),
    StructField("amount", DoubleType(), True),
    StructField("timestamp", TimestampType(), False),
])

Debugging Notebooks Locally¶

Python Cell Debugging¶

Since notebooks are .py files, you can debug them like any Python script:

1. Set breakpoints in the gutter.
2. Press F5 or use Run > Start Debugging.
3. Select the Python: Current File configuration.

For PySpark cells, you need a local Spark session. See local-spark-debugging.md for full setup.

launch.json Configuration¶

Add this to .vscode/launch.json for notebook debugging:

{
  "version": "0.2.0",
  "configurations": [
    {
      "name": "Debug Notebook (Local Spark)",
      "type": "debugpy",
      "request": "launch",
      "program": "${file}",
      "console": "integratedTerminal",
      "env": {
        "PYSPARK_PYTHON": "${workspaceFolder}/.venv/Scripts/python.exe",
        "FABRIC_POC_HASH_SALT": "local-dev-salt-do-not-use-in-prod",
        "SPARK_LOCAL_IP": "127.0.0.1"
      },
      "justMyCode": true,
      "args": []
    },
    {
      "name": "Debug Generator",
      "type": "debugpy",
      "request": "launch",
      "program": "${file}",
      "console": "integratedTerminal",
      "env": {
        "FABRIC_POC_HASH_SALT": "local-dev-salt-do-not-use-in-prod"
      }
    }
  ]
}

Handling Fabric-Specific APIs¶

Notebook code that references mssparkutils or notebookutils will fail locally. Use conditional imports:

try:
    from notebookutils import mssparkutils
    IS_FABRIC = True
except ImportError:
    IS_FABRIC = False

# Use IS_FABRIC to gate Fabric-specific calls
if IS_FABRIC:
    mssparkutils.credentials.getSecret("keyvault", "secret-name")
else:
    import os
    secret = os.environ.get("SECRET_NAME", "local-default")

Extensions Ecosystem¶

Required Extensions¶

Install these for full Fabric development support:

Recommended Extensions¶

Install All at Once¶

code --install-extension ms-fabric.fabric-vscode \
     --install-extension ms-python.python \
     --install-extension ms-python.vscode-pylance \
     --install-extension ms-azuretools.vscode-bicep \
     --install-extension ms-vscode.azure-account \
     --install-extension eamodio.gitlens \
     --install-extension yzhang.markdown-all-in-one \
     --install-extension redhat.vscode-yaml

Settings Configuration¶

Workspace Settings (.vscode/settings.json)¶

{
  // ── Python ──────────────────────────────────────────────────
  "python.defaultInterpreterPath": "${workspaceFolder}/.venv/Scripts/python.exe",
  "python.analysis.typeCheckingMode": "basic",
  "python.analysis.autoImportCompletions": true,
  "python.analysis.extraPaths": [
    "${workspaceFolder}/data_generation",
    "${workspaceFolder}/notebooks"
  ],

  // ── Pylance ─────────────────────────────────────────────────
  "python.languageServer": "Pylance",
  "python.analysis.diagnosticSeverityOverrides": {
    "reportMissingImports": "warning",
    "reportMissingModuleSource": "information"
  },

  // ── Formatting ──────────────────────────────────────────────
  "editor.formatOnSave": true,
  "editor.defaultFormatter": "ms-python.python",
  "[python]": {
    "editor.defaultFormatter": "ms-python.python",
    "editor.codeActionsOnSave": {
      "source.organizeImports": "explicit"
    }
  },

  // ── Files ───────────────────────────────────────────────────
  "files.exclude": {
    "**/__pycache__": true,
    "**/.pytest_cache": true,
    "**/node_modules": true,
    ".fabric/": true
  },
  "files.associations": {
    "*.bicep": "bicep",
    "*.bicepparam": "bicep"
  },

  // ── Search ──────────────────────────────────────────────────
  "search.exclude": {
    "**/data_generation/output": true,
    "**/.venv": true,
    "**/temp": true
  },

  // ── Testing ─────────────────────────────────────────────────
  "python.testing.pytestEnabled": true,
  "python.testing.pytestArgs": [
    "validation/unit_tests",
    "-v"
  ],

  // ── Bicep ───────────────────────────────────────────────────
  "bicep.experimental.deployPane": true,

  // ── Terminal ────────────────────────────────────────────────
  "terminal.integrated.defaultProfile.windows": "Git Bash",
  "terminal.integrated.env.windows": {
    "FABRIC_POC_HASH_SALT": "local-dev-salt-do-not-use-in-prod"
  }
}

User Settings Recommendations¶

Add these to your user-level settings.json (Ctrl+, > Open Settings JSON):

{
  "editor.minimap.enabled": false,
  "editor.bracketPairColorization.enabled": true,
  "editor.guides.bracketPairs": "active",
  "editor.stickyScroll.enabled": true,
  "workbench.editor.wrapTabs": true,
  "git.autofetch": true,
  "git.confirmSync": false,
  "diffEditor.ignoreTrimWhitespace": false
}

Keyboard Shortcuts and Productivity¶

Essential Shortcuts¶

Fabric-Specific Shortcuts¶

Configure these in keybindings.json for common Fabric operations:

[
  {
    "key": "ctrl+shift+u",
    "command": "fabric.uploadCurrentItem",
    "when": "editorTextFocus"
  },
  {
    "key": "ctrl+shift+d",
    "command": "fabric.downloadCurrentItem",
    "when": "editorTextFocus"
  }
]

Multi-Cursor Editing for Notebook Cells¶

When refactoring across multiple notebook cells:

1. Ctrl+D to select next occurrence.
2. Ctrl+Shift+L to select all occurrences.
3. Alt+Click to place cursors manually.

This is particularly useful when renaming a column across Bronze/Silver/Gold notebooks.

Comparison: VS Code vs Fabric Web UI vs Tabular Editor¶

When to Use Each Tool¶

VS Code + Fabric Extension: - Day-to-day notebook and pipeline development - Code reviews and pull requests - Refactoring across multiple files - Local debugging and testing - Bicep infrastructure editing

Fabric Web UI: - Running notebooks against live data - Monitoring pipeline executions - Visual pipeline design - Lakehouse data exploration - Spark UI and job monitoring

Tabular Editor: - Complex DAX measure authoring - Semantic model schema changes - Calculation group management - Performance tuning (vertipaq analyzer)

Troubleshooting¶

Extension Not Showing Workspaces¶

Problem: Fabric panel shows "No workspaces found"
Cause:   Token expired or wrong tenant
Fix:     Ctrl+Shift+P > "Fabric: Sign Out" > "Fabric: Sign In"

Upload Fails with 403¶

Problem: "Insufficient permissions" when uploading
Cause:   User lacks Contributor role on the workspace
Fix:     Ask workspace admin to assign Contributor or Member role

Notebook Sync Creates Duplicate¶

Problem: Upload creates a new item instead of updating existing
Cause:   Item mapping lost (missing .fabric/settings.json)
Fix:     Delete .fabric/ and rebind workspace

Local Spark Session Fails¶

Problem: "Java not found" when debugging notebooks locally
Cause:   JDK not installed or JAVA_HOME not set
Fix:     See local-spark-debugging.md for JDK setup

Python Import Errors in Notebooks¶

Problem: Pylance reports missing imports for pyspark, mssparkutils
Cause:   Virtual environment missing PySpark
Fix:     pip install pyspark delta-spark in your .venv
         Add stubs for mssparkutils (see notebook-unit-testing.md)

References¶

• Microsoft Fabric VS Code Extension
• VS Code Python Development
• Fabric Git Integration Overview
• fabric-cicd Python Library
• Tabular Editor Documentation
• This POC: fabric-cicd Deployment Guide
• This POC: Local Spark Debugging
• This POC: Notebook Unit Testing