5 min read
Microsoft Fabric Workspace: Organization and Collaboration
Workspaces are the foundation of organization and collaboration in Microsoft Fabric. Today, I will explore how to effectively structure and manage Fabric workspaces for your data platform.
What is a Fabric Workspace?
A workspace in Fabric is a container for your data artifacts that provides:
- Collaboration space for teams
- Security boundary for data access
- Capacity assignment for compute resources
- Git integration for version control
┌─────────────────────────────────────────────────────┐
│ Fabric Workspace │
│ "Sales Analytics" │
├─────────────────────────────────────────────────────┤
│ ┌─────────────────────────────────────────────────┐│
│ │ Items ││
│ ├─────────────┬─────────────┬─────────────────────┤│
│ │ Lakehouses │ Warehouses │ Semantic Models ││
│ │ Notebooks │ Pipelines │ Reports ││
│ │ KQL DBs │ Dataflows │ Spark Job Defs ││
│ └─────────────┴─────────────┴─────────────────────┘│
├─────────────────────────────────────────────────────┤
│ ┌─────────────────────────────────────────────────┐│
│ │ Access Control ││
│ │ Admin │ Member │ Contributor │ Viewer ││
│ └─────────────────────────────────────────────────┘│
├─────────────────────────────────────────────────────┤
│ Assigned Capacity: F64 │
│ Git: Connected to Azure DevOps │
└─────────────────────────────────────────────────────┘Creating Workspaces
Via Fabric Portal
# Workspace creation settings
workspace_settings = {
"name": "Sales Analytics - Production",
"description": "Production workspace for Sales Analytics team",
"license_mode": "Fabric", # or "Premium", "Pro"
"capacity": "F64-Capacity",
"default_storage": "OneLake",
"contact_list": ["sales-analytics-team@contoso.com"],
"workspace_admins": ["admin1@contoso.com", "admin2@contoso.com"]
}Via REST API
import requests
def create_workspace(workspace_name: str, capacity_id: str, access_token: str):
url = "https://api.powerbi.com/v1.0/myorg/groups"
headers = {
"Authorization": f"Bearer {access_token}",
"Content-Type": "application/json"
}
payload = {
"name": workspace_name,
"capacityId": capacity_id
}
response = requests.post(url, headers=headers, json=payload)
return response.json()
# Create workspace
workspace = create_workspace(
"Sales Analytics - Production",
"capacity-guid-here",
access_token
)
print(f"Created workspace: {workspace['id']}")Workspace Roles
# Fabric workspace roles and permissions
workspace_roles = {
"Admin": {
"description": "Full control over workspace",
"permissions": [
"Add/remove members",
"Change workspace settings",
"Delete workspace",
"Create/edit/delete all items",
"Publish apps",
"Manage permissions on all items"
]
},
"Member": {
"description": "Create and edit content, share items",
"permissions": [
"Create/edit/delete all items",
"Share items and manage item permissions",
"Publish apps (with admin approval)",
"Cannot add/remove members",
"Cannot change workspace settings"
]
},
"Contributor": {
"description": "Create and edit content only",
"permissions": [
"Create/edit/delete all items",
"Cannot share items",
"Cannot manage permissions",
"Cannot publish apps"
]
},
"Viewer": {
"description": "View content only",
"permissions": [
"View all items",
"Cannot create or edit",
"Cannot share",
"Can interact with reports (filters, etc.)"
]
}
}Assigning Roles
def assign_workspace_role(workspace_id: str, user_email: str, role: str, access_token: str):
url = f"https://api.powerbi.com/v1.0/myorg/groups/{workspace_id}/users"
headers = {
"Authorization": f"Bearer {access_token}",
"Content-Type": "application/json"
}
payload = {
"emailAddress": user_email,
"groupUserAccessRight": role # Admin, Member, Contributor, Viewer
}
response = requests.post(url, headers=headers, json=payload)
return response.status_code == 200
# Add users to workspace
assign_workspace_role(workspace_id, "analyst@contoso.com", "Contributor", token)
assign_workspace_role(workspace_id, "manager@contoso.com", "Member", token)
assign_workspace_role(workspace_id, "executive@contoso.com", "Viewer", token)Workspace Organization Strategies
By Environment
┌─────────────────────────────────────────────────────┐
│ Development │
│ ├── Sales Analytics - Dev │
│ ├── Marketing Analytics - Dev │
│ └── Finance Analytics - Dev │
├─────────────────────────────────────────────────────┤
│ Test/UAT │
│ ├── Sales Analytics - Test │
│ ├── Marketing Analytics - Test │
│ └── Finance Analytics - Test │
├─────────────────────────────────────────────────────┤
│ Production │
│ ├── Sales Analytics - Prod │
│ ├── Marketing Analytics - Prod │
│ └── Finance Analytics - Prod │
└─────────────────────────────────────────────────────┘By Domain
┌─────────────────────────────────────────────────────┐
│ Sales Domain │
│ ├── Sales Raw Data (Lakehouse) │
│ ├── Sales Curated (Lakehouse) │
│ └── Sales Analytics (Reports) │
├─────────────────────────────────────────────────────┤
│ Marketing Domain │
│ ├── Marketing Raw Data │
│ ├── Marketing Curated │
│ └── Marketing Analytics │
├─────────────────────────────────────────────────────┤
│ Shared Data │
│ ├── Enterprise Dimensions │
│ ├── Common Reference Data │
│ └── Master Data │
└─────────────────────────────────────────────────────┘By Layer (Medallion Architecture)
# Medallion architecture workspace structure
medallion_workspaces = {
"Bronze Layer": {
"workspace": "Data Platform - Bronze",
"purpose": "Raw data ingestion",
"items": ["Lakehouses for raw data", "Ingestion pipelines"],
"access": "Data engineers only"
},
"Silver Layer": {
"workspace": "Data Platform - Silver",
"purpose": "Cleaned and conformed data",
"items": ["Lakehouses for curated data", "Transformation notebooks"],
"access": "Data engineers and scientists"
},
"Gold Layer": {
"workspace": "Data Platform - Gold",
"purpose": "Business-ready aggregations",
"items": ["Warehouses", "Semantic models", "Reports"],
"access": "Analysts and business users"
}
}Workspace Settings
Configure Spark Settings
# Workspace-level Spark configuration
spark_settings = {
"spark.executor.memory": "28g",
"spark.executor.cores": "4",
"spark.driver.memory": "8g",
"spark.sql.shuffle.partitions": "200",
"spark.sql.adaptive.enabled": "true"
}
# Apply via workspace settings in Fabric portal
# Settings > Spark settings > Spark propertiesOneLake Settings
# OneLake workspace settings
onelake_settings = {
"storage_format": "Delta", # Default format for tables
"external_data_access": True, # Allow shortcuts to external data
"cross_workspace_access": True, # Allow other workspaces to create shortcuts
"retention_policy": "30 days" # Soft delete retention
}Workspace Governance
Naming Conventions
# Recommended naming convention
naming_convention = {
"pattern": "{Domain} - {Layer/Purpose} - {Environment}",
"examples": [
"Sales - Analytics - Prod",
"Marketing - Raw Data - Dev",
"Finance - Reporting - Test",
"Enterprise - Shared Dimensions - Prod"
],
"rules": [
"Use consistent capitalization",
"Include environment suffix",
"Keep names concise but descriptive",
"Avoid special characters"
]
}Tagging and Metadata
# Use workspace description for metadata
workspace_metadata = {
"name": "Sales Analytics - Production",
"description": """
Owner: Sales Analytics Team
Contact: sales-analytics@contoso.com
Data Classification: Internal
Business Domain: Sales
Last Review: 2023-05-01
SLA: 99.5% availability
""",
"contacts": ["sales-analytics-team@contoso.com"]
}Cross-Workspace Access
# Access data from another workspace via shortcuts
# In Lakehouse: Create shortcut > OneLake > Select workspace > Select item
# Or via Spark with absolute paths
df = spark.read.format("delta").load(
"abfss://other-workspace@onelake.dfs.fabric.microsoft.com/lakehouse.Lakehouse/Tables/shared_dimensions"
)
# SQL endpoint cross-database query
# FROM other_lakehouse.dbo.shared_dimensionsTomorrow, I will cover the comparison between Fabric and Azure Synapse Analytics.