Combined Cloud Managed Services

Hands-on Lab 1: Create Azure ML Workspace and Explore Studio Interface

## Learning Objectives By the end of this lab, you will: - Create an Azure Machine Learning workspace - Navigate the Azure ML Studio interface - Understand key components and navigation structure - Explore workspace assets and settings - Configure basic workspace settings

—

## Part 1: Create Azure ML Workspace (15 minutes)

### Step 1: Access Azure Portal 1. Open your web browser and navigate to [portal.azure.com](https://portal.azure.com) 2. Sign in with your Azure account credentials 3. Once logged in, you'll see the Azure portal dashboard

### Step 2: Create Resource Group (if needed) 1. In the Azure portal, click “Resource groups” in the left sidebar 2. Click “+ Create” button 3. Fill in the details:

1. Subscription: Select your subscription
2. Resource group name: `rg-azureml-lab`
3. Region: Choose a region close to you (e.g., East US, West Europe)

4. Click “Review + create” then “Create”

### Step 3: Create Azure ML Workspace 1. In the Azure portal search bar, type “Machine Learning” 2. Select “Machine Learning” from the results 3. Click “+ Create” button 4. Fill in the workspace details:

1. Subscription: Select your subscription
2. Resource group: Select `rg-azureml-lab` (created above)
3. Workspace name: `mlw-lab-workspace` (must be unique)
4. Region: Same as your resource group
5. Storage account: Leave as default (will create new)
6. Key vault: Leave as default (will create new)
7. Application insights: Leave as default (will create new)
8. Container registry: Leave as default (will create new)

### Step 4: Configure Advanced Settings 1. Click “Next: Networking”

1. Leave as “Public endpoint” for this lab

2. Click “Next: Encryption”

1. Leave as default (Microsoft managed keys)

3. Click “Next: Identity”

1. Leave as default (System assigned)

4. Click “Next: Tags”

1. Add optional tags:
   1. Key: `Environment`, Value: `Learning`
   2. Key: `Purpose`, Value: `Lab`

### Step 5: Create the Workspace 1. Click “Review + create” 2. Review all settings and click “Create” 3. Wait for deployment to complete (typically 5-10 minutes) 4. Once complete, click “Go to resource”

—

## Part 2: Launch Azure ML Studio (5 minutes)

### Step 1: Access ML Studio 1. From your workspace resource page, click “Launch studio”

1. Alternative: Navigate directly to [ml.azure.com](https://ml.azure.com)

2. Sign in if prompted 3. Select your workspace from the dropdown if multiple workspaces exist

### Step 2: Studio Welcome Screen 1. You'll see the Azure ML Studio welcome screen 2. Take note of the workspace name in the top-left corner 3. Click “Get started” or “Skip tour” depending on preference

—

## Part 3: Explore Studio Interface (20 minutes)

### Step 1: Navigation Overview Familiarize yourself with the main navigation areas:

Left Sidebar - Main Navigation: - Home: Dashboard and recent activities - Notebooks: Jupyter notebook environment - Automated ML: AutoML experiments - Designer: Visual ML pipeline designer - Data: Datasets and datastores - Jobs: Training jobs and experiments - Components: Reusable pipeline components - Pipelines: ML pipelines - Models: Model registry - Endpoints: Deployment endpoints - Compute: Computing resources - Datastores: Data storage connections

Top Bar: - Workspace selector - Subscription and resource group info - User account menu - Notifications - Help and documentation

### Step 2: Explore Home Dashboard 1. Click “Home” in the left sidebar 2. Observe the dashboard sections:

1. Recent: Recently accessed items
2. Quick actions: Common tasks
3. Learning resources: Tutorials and documentation
4. Workspace overview: Basic workspace information

### Step 3: Explore Compute Section 1. Click “Compute” in the left sidebar 2. You'll see four tabs:

1. Compute instances: Development environments
2. Compute clusters: Training clusters
3. Inference clusters: Deployment clusters
4. Attached compute: External compute resources

3. Note: All sections will be empty as this is a new workspace

### Step 4: Explore Data Section 1. Click “Data” in the left sidebar 2. Explore the two main tabs:

1. Data assets: Datasets registered in the workspace
2. Datastores: Connected data storage locations

3. Note the default datastore (workspaceblobstore) that was created automatically

### Step 5: Explore Models Section 1. Click “Models” in the left sidebar 2. This shows registered models in your workspace 3. Currently empty, but this is where trained models will appear

### Step 6: Explore Jobs Section 1. Click “Jobs” in the left sidebar 2. This shows training experiments and jobs 3. Currently empty, but will populate as you run experiments

—

## Part 4: Workspace Settings and Configuration (10 minutes)

### Step 1: Access Workspace Settings 1. Look for the gear icon (⚙️) in the top-right area or 2. Click your workspace name in the top-left to access workspace menu 3. Select “View all properties in Azure portal”

### Step 2: Review Workspace Properties In the Azure portal workspace page, explore: 1. Overview: Basic information and metrics 2. Access control (IAM): User permissions and roles 3. Properties: Workspace configuration details 4. Locks: Resource protection settings 5. Monitoring: Workspace activity and diagnostics

### Step 3: Review Associated Resources 1. In the workspace Overview page, note the “Associated resources”:

1. Storage account
2. Key vault
3. Application Insights
4. Container Registry

2. Click on each to understand their role in the ML workspace

### Step 4: Configure Workspace Access (Optional) 1. Go to “Access control (IAM)“ 2. Click ”+ Add” → “Add role assignment” 3. Review available roles:

1. Owner: Full access
2. Contributor: Can create and manage resources
3. Reader: Read-only access
4. AzureML Data Scientist: ML-specific permissions

4. Don't actually add users in this lab - just review the options

—

## Part 5: Verification and Exploration Tasks (10 minutes)

### Task 1: Verify Workspace Creation 1. Return to Azure ML Studio ([ml.azure.com](https://ml.azure.com)) 2. Confirm you can access your workspace 3. Check that the workspace name appears correctly in the top-left

### Task 2: Explore Sample Notebooks 1. Click “Notebooks” in the left sidebar 2. Look for “Samples” folder 3. Expand the folder to see sample notebooks 4. Click on any notebook to preview (don't run yet)

### Task 3: Check Default Datastore 1. Go to “Data” → “Datastores” 2. Click on “workspaceblobstore” 3. Review the connection details 4. Note the Azure Blob Storage connection

### Task 4: Explore Learning Resources 1. Return to “Home” 2. Scroll down to “Learning resources” 3. Click on “Azure Machine Learning documentation” 4. Bookmark useful resources for future reference

—

## Troubleshooting Common Issues

### Issue 1: Workspace Creation Fails Symptoms: Deployment fails or times out Solutions: - Check subscription limits and quotas - Try a different region - Ensure unique workspace name - Verify resource group permissions

### Issue 2: Cannot Access ML Studio Symptoms: Studio doesn't load or shows access denied Solutions: - Clear browser cache and cookies - Try incognito/private browsing mode - Check if workspace deployment completed successfully - Verify you have appropriate permissions

### Issue 3: Studio Interface Loads Slowly Symptoms: Pages take long to load Solutions: - Check internet connection - Try different browser - Disable browser extensions temporarily - Use recommended browsers (Chrome, Edge, Firefox)

—

## Lab Completion Checklist

Mark each item as complete:

- [ ] Successfully created Azure ML workspace - [ ] Launched Azure ML Studio - [ ] Explored main navigation sections - [ ] Reviewed workspace dashboard - [ ] Examined compute, data, and model sections - [ ] Accessed workspace settings in Azure portal - [ ] Identified associated Azure resources - [ ] Verified workspace functionality - [ ] Bookmarked important resources

—

## Next Steps

After completing this lab, you should:

1. Keep your workspace: You'll use it for subsequent labs 2. Explore documentation: Familiarize yourself with Azure ML concepts 3. Review pricing: Understand costs for compute resources 4. Plan next lab: Prepare for Lab 2 (Compute Resources and Datasets)

—

## Key Takeaways

✅ Workspace is the Foundation: Everything in Azure ML happens within a workspace ✅ Studio is the Interface: Web-based interface for most ML tasks ✅ Multiple Entry Points: Can access via Azure portal or directly via ml.azure.com ✅ Associated Resources: Workspace automatically creates supporting Azure services ✅ Role-Based Access: Proper permissions are crucial for collaboration

—

## Additional Resources

- [Azure ML Workspace Documentation](https://docs.microsoft.com/azure/machine-learning/concept-workspace) - [Azure ML Studio Overview](https://docs.microsoft.com/azure/machine-learning/overview-what-is-machine-learning-studio) - [Azure ML Pricing Calculator](https://azure.microsoft.com/pricing/calculator/) - [Azure ML Service Limits](https://docs.microsoft.com/azure/machine-learning/resource-limits-quotas-capacity)

Estimated Time to Complete: 45-60 minutes Cost Impact: Minimal (workspace creation is free, compute charges apply only when running resources)