📋 Tutorial 13: Enterprise Migration Planning & Delivery¶

Last Updated: 2026-04-15 | Version: 2.0 Status: ✅ Final | Maintainer: Documentation Team

🏠 Home > 📖 Tutorials > 📋 Migration Planning

📋 Tutorial 13: Enterprise Migration Planning & Delivery¶

📊 Progress Tracker¶

┌────────┬────────┬────────┬────────┬────────┬────────┬────────┬────────┬────────┬────────┬────────┬────────┬────────┬────────┐
│   00   │   01   │   02   │   03   │   04   │   05   │   06   │   07   │   08   │   09   │   10   │   11   │   12   │   13   │
│ SETUP  │ BRONZE │ SILVER │  GOLD  │  RT    │  PBI   │ PIPES  │  GOV   │ MIRROR │  AI/ML │TERADATA│  SAS   │ CI/CD  │PLANNING│
├────────┼────────┼────────┼────────┼────────┼────────┼────────┼────────┼────────┼────────┼────────┼────────┼────────┼────────┤
│   ✅   │   ✅   │   ✅   │   ✅   │   ✅   │   ✅   │   ✅   │   ✅   │   ✅   │   ✅   │   ✅   │   ✅   │   ✅   │  🔵   │
└────────┴────────┴────────┴────────┴────────┴────────┴────────┴────────┴────────┴────────┴────────┴────────┴────────┴────────┘
                                                                                                                        ▲
                                                                                                                   YOU ARE HERE
                                                                                                               FINAL TUTORIAL

📖 Overview¶

This comprehensive guide provides everything needed to plan, execute, and deliver a full enterprise migration from existing data platforms (Teradata, Informatica, Oracle, etc.) to Microsoft Fabric. These are mature, capable platforms; references to them here are based on each vendor's publicly available documentation and are not an authoritative assessment of those products. A migration is the right call only when Fabric is a better fit for your specific needs — where the incumbent platform remains the stronger choice, staying put is a valid outcome. The guide includes:

• 6-month migration timeline with 3-week sprint cadence
• POC to Production complete journey
• Enterprise considerations: VPN, ExpressRoute, access control, security
• Center of Excellence establishment
• Organizational change management
• Best practices and lessons learned

This guide is designed for Data Platform Architects, Program Managers, and Technical Leads responsible for large-scale data modernization initiatives.

🎯 Migration Objectives¶

Business Objectives¶

These are example program targets to set and then validate empirically for your estate — not guaranteed outcomes and not a claim about any specific source platform's performance or cost. Actual results depend heavily on your current contracts, workloads, and tuning; set your own baselines.

Technical Objectives¶

🏗️ Migration Architecture¶

Source: What is Microsoft Fabric?

flowchart TB
    subgraph Source["🏢 Source Environment"]
        TD[(Teradata<br/>Data Warehouse)]
        INFA[Informatica<br/>ETL Platform]
        ORACLE[(Oracle<br/>Database)]
        LEGACY_BI[Existing<br/>BI Tools]
    end

    subgraph Network["🔒 Secure Connectivity"]
        EXPROUTE[ExpressRoute /<br/>VPN Gateway]
        FIREWALL[Azure<br/>Firewall]
        PLINK[Private<br/>Link]
    end

    subgraph Azure["☁️ Azure Foundation"]
        AAD[Microsoft<br/>Entra ID]
        KEYVAULT[Azure<br/>Key Vault]
        MONITOR[Azure<br/>Monitor]
        PURVIEW[Microsoft<br/>Purview]
    end

    subgraph Fabric["⚡ Microsoft Fabric"]
        subgraph Ingestion["Data Ingestion"]
            ADF[Data Factory<br/>Pipelines]
            MIRROR[Database<br/>Mirroring]
            STREAM[Eventstream]
        end

        subgraph Storage["Data Storage"]
            ONELAKE[(OneLake<br/>Delta Lake)]
            LH[Lakehouses]
            WH[Warehouses]
        end

        subgraph Analytics["Analytics"]
            NB[Notebooks]
            SM[Semantic<br/>Models]
            PBI[Power BI<br/>Reports]
        end
    end

    TD --> EXPROUTE
    INFA --> EXPROUTE
    ORACLE --> EXPROUTE

    EXPROUTE --> FIREWALL
    FIREWALL --> PLINK
    PLINK --> ADF
    PLINK --> MIRROR

    ADF --> ONELAKE
    MIRROR --> ONELAKE
    STREAM --> ONELAKE

    ONELAKE --> LH
    ONELAKE --> WH

    LH --> NB
    WH --> SM
    SM --> PBI

    AAD --> Fabric
    KEYVAULT --> ADF
    MONITOR --> Fabric
    PURVIEW --> ONELAKE

    style Source fill:#e74c3c,color:#fff
    style Network fill:#f39c12,color:#fff
    style Azure fill:#3498db,color:#fff
    style Fabric fill:#9b59b6,color:#fff

📅 6-Month Migration Timeline¶

Sprint Cadence¶

Timeline Overview¶

gantt
    title 6-Month Migration Timeline
    dateFormat YYYY-MM-DD
    axisFormat %b %d

    section Phase 1: Foundation
    Sprint 1 - Discovery & Planning     :s1, 2024-01-01, 21d
    Sprint 2 - Infrastructure Setup     :s2, after s1, 21d

    section Phase 2: POC
    Sprint 3 - POC Development          :s3, after s2, 21d
    Sprint 4 - POC Validation           :s4, after s3, 21d

    section Phase 3: Core Migration
    Sprint 5 - Dimensions & Reference   :s5, after s4, 21d
    Sprint 6 - Fact Tables & ETL        :s6, after s5, 21d

    section Phase 4: Advanced
    Sprint 7 - BI & Analytics           :s7, after s6, 21d
    Sprint 8 - Real-time & AI/ML        :s8, after s7, 21d

    section Phase 5: Production
    Cutover Preparation                 :co1, after s8, 7d
    Production Cutover                  :milestone, m1, after co1, 0d
    Hypercare                          :hc, after m1, 14d

📊 Sprint-by-Sprint Breakdown¶

Sprint 1: Discovery & Planning (Weeks 1-3)¶

Objectives: - Complete inventory assessment - Establish project governance - Define migration scope and priorities - Set up project infrastructure

Deliverables:

Key Activities:

Week 1:
├── Kickoff meeting with stakeholders
├── Inventory Teradata databases and sizes
├── Catalog Informatica workflows
├── Document current BI landscape
└── Identify data owners and SMEs

Week 2:
├── Complexity scoring of all objects
├── Dependency mapping
├── Performance baseline capture
├── Security and compliance review
└── Infrastructure requirements gathering

Week 3:
├── Prioritization workshop
├── Migration wave planning
├── Risk assessment workshop
├── Project plan finalization
└── Governance model approval

Sprint 1 Checklist:

• Teradata inventory complete (databases, tables, sizes)
• Informatica workflow catalog created
• Dependency matrix documented
• Complexity scores assigned
• Migration waves defined
• Project plan approved
• Governance structure established
• Communication plan distributed

Sprint 2: Infrastructure Setup (Weeks 4-6)¶

Source: Workspaces in Microsoft Fabric

Objectives: - Provision Azure and Fabric infrastructure - Establish secure connectivity - Configure identity and access - Set up development environments

Deliverables:

Infrastructure Components:

flowchart LR
    subgraph OnPrem["On-Premises"]
        DC[Data Center]
        TD[(Teradata)]
        INFA[Informatica]
    end

    subgraph Connectivity["Connectivity"]
        EXPROUTE[ExpressRoute<br/>Circuit]
        VPN[VPN<br/>Gateway]
    end

    subgraph Azure["Azure"]
        VNET[Virtual<br/>Network]
        FW[Azure<br/>Firewall]
        PLINK[Private<br/>Endpoint]
    end

    subgraph Fabric["Fabric"]
        CAP[F64<br/>Capacity]
        WS_DEV[Dev<br/>Workspace]
        WS_TEST[Test<br/>Workspace]
        WS_PROD[Prod<br/>Workspace]
    end

    DC --> EXPROUTE
    DC -.-> VPN

    EXPROUTE --> VNET
    VPN --> VNET
    VNET --> FW
    FW --> PLINK
    PLINK --> CAP

    CAP --> WS_DEV
    CAP --> WS_TEST
    CAP --> WS_PROD

Connectivity Requirements:

ExpressRoute Setup:

# Create ExpressRoute circuit (Azure CLI)
az network express-route create \
  --name "ER-Fabric-Migration" \
  --resource-group "rg-fabric-connectivity" \
  --location "eastus2" \
  --bandwidth 1000 \
  --peering-location "Washington DC" \
  --sku-family MeteredData \
  --sku-tier Standard

# Create private peering
az network express-route peering create \
  --circuit-name "ER-Fabric-Migration" \
  --resource-group "rg-fabric-connectivity" \
  --peering-type AzurePrivatePeering \
  --peer-asn 65001 \
  --primary-peer-subnet "10.0.0.0/30" \
  --secondary-peer-subnet "10.0.0.4/30" \
  --vlan-id 100

Access Control Matrix:

Sprint 2 Checklist:

• Azure subscription active with Fabric enabled
• Fabric F64 capacity provisioned
• ExpressRoute circuit active OR VPN gateway connected
• Firewall rules configured
• Entra ID groups created and assigned
• Key Vault provisioned with secrets
• Dev workspace created with Git integration
• Test workspace created
• Prod workspace created (empty)
• Self-Hosted Integration Runtime installed (if needed)
• Connectivity tested end-to-end

Sprint 3: POC Development (Weeks 7-9)¶

Objectives: - Migrate 3-5 representative tables - Convert sample ETL workflows - Build initial semantic model - Validate technical approach

POC Scope:

POC Data Selection Criteria:

mindmap
  root((POC Data<br/>Selection))
    Representative
      Various data types
      Different sizes
      Multiple sources
    Complexity Range
      Simple dimension
      Complex fact
      Incremental load
    Business Value
      High visibility
      Active users
      Clear success metrics
    Risk Management
      Non-critical data
      Recoverable
      Parallel available

POC Technical Activities:

Week 7:
├── Create Bronze Lakehouse
├── Configure Teradata connector
├── Migrate DIM_PLAYER table
├── Migrate DIM_MACHINE table
├── Set up initial Data Pipeline
└── Document data mapping

Week 8:
├── Create Silver Lakehouse
├── Develop cleansing notebook
├── Migrate FACT_SLOT_TRANSACTIONS (sample)
├── Convert first Informatica workflow
├── Implement incremental load pattern
└── Performance benchmarking

Week 9:
├── Create Gold Lakehouse
├── Build aggregation notebook
├── Create semantic model
├── Build 2 sample reports
├── End-to-end testing
└── POC documentation

Sprint 3 Checklist:

• Bronze Lakehouse created
• Silver Lakehouse created
• Gold Lakehouse created
• 2+ dimension tables migrated
• 1+ fact table migrated (sample data)
• 2+ ETL workflows converted
• Semantic model created
• 2+ reports functional
• Data validation passed
• Performance baseline captured

Sprint 4: POC Validation & Planning (Weeks 10-12)¶

Objectives: - Validate POC results with stakeholders - Capture lessons learned - Refine migration approach - Plan core migration waves

Validation Activities:

POC Success Criteria:

Sprint 4 Deliverables:

• POC Results Report
• Lessons Learned Document
• Refined Migration Plan
• Updated Timeline and Estimates
• Resource Requirements (refined)
• Go/No-Go Decision
• Core Migration Wave Plan

Sprint 5: Dimensions & Reference Data (Weeks 13-15)¶

Objectives: - Migrate all dimension tables - Migrate reference data - Establish master data patterns - Implement data quality framework

Migration Scope:

Dimension Migration Pattern:

# Fabric Notebook: Dimension Migration Pattern
from pyspark.sql.functions import *
from delta.tables import DeltaTable

def migrate_dimension(
    source_table: str,
    target_table: str,
    key_columns: list,
    scd_type: int = 1
):
    """
    Migrate dimension table with SCD handling.

    Args:
        source_table: Source Teradata table
        target_table: Target Fabric table
        key_columns: Business key columns
        scd_type: 1 for overwrite, 2 for history
    """
    # Read from source
    df_source = spark.read \
        .format("jdbc") \
        .option("url", teradata_url) \
        .option("dbtable", source_table) \
        .load()

    # Add metadata
    df_source = df_source \
        .withColumn("_load_timestamp", current_timestamp()) \
        .withColumn("_source_system", lit("Teradata"))

    if scd_type == 1:
        # Type 1: Full overwrite
        df_source.write \
            .mode("overwrite") \
            .format("delta") \
            .saveAsTable(target_table)

    elif scd_type == 2:
        # Type 2: Merge with history
        df_source = df_source \
            .withColumn("_valid_from", current_timestamp()) \
            .withColumn("_valid_to", lit(None).cast("timestamp")) \
            .withColumn("_is_current", lit(True))

        if spark.catalog.tableExists(target_table):
            delta_table = DeltaTable.forName(spark, target_table)

            # Close old records
            delta_table.alias("target").merge(
                df_source.alias("source"),
                " AND ".join([f"target.{k} = source.{k}" for k in key_columns]) +
                " AND target._is_current = true"
            ).whenMatchedUpdate(set={
                "_valid_to": "source._valid_from",
                "_is_current": "false"
            }).execute()

            # Insert new records
            df_source.write.mode("append").saveAsTable(target_table)
        else:
            df_source.write \
                .mode("overwrite") \
                .format("delta") \
                .saveAsTable(target_table)

    # Optimize
    spark.sql(f"OPTIMIZE {target_table}")

    return spark.table(target_table).count()

Sprint 5 Checklist:

• All dimension tables migrated
• Reference data migrated
• Data quality checks implemented
• Row counts validated
• Business key integrity verified
• SCD patterns working correctly
• Documentation updated

Sprint 6: Fact Tables & ETL (Weeks 16-18)¶

Objectives: - Migrate all fact tables - Convert Informatica workflows to Data Factory - Implement incremental load patterns - Establish scheduling and monitoring

Fact Table Migration Strategy:

Informatica to Data Factory Mapping:

ETL Conversion Example:

{
  "name": "pl_daily_slot_transactions",
  "properties": {
    "description": "Converted from Informatica wf_daily_slot_load",
    "activities": [
      {
        "name": "Extract from Teradata",
        "type": "Copy",
        "inputs": [{"name": "ds_teradata_slot_trans"}],
        "outputs": [{"name": "ds_bronze_slot_trans"}],
        "typeProperties": {
          "source": {
            "type": "TeradataSource",
            "query": {
              "value": "SELECT * FROM CASINO_DW.FACT_SLOT_TRANSACTIONS WHERE transaction_date = '@{formatDateTime(pipeline().parameters.RunDate, 'yyyy-MM-dd')}'",
              "type": "Expression"
            }
          },
          "sink": {
            "type": "ParquetSink",
            "storeSettings": {
              "type": "AzureBlobFSWriteSettings"
            }
          }
        }
      },
      {
        "name": "Transform to Silver",
        "type": "SynapseNotebook",
        "dependsOn": [{"activity": "Extract from Teradata", "dependencyConditions": ["Succeeded"]}],
        "typeProperties": {
          "notebook": {"name": "nb_transform_slot_transactions"},
          "parameters": {
            "run_date": {"value": "@pipeline().parameters.RunDate", "type": "string"}
          }
        }
      },
      {
        "name": "Load to Gold",
        "type": "SynapseNotebook",
        "dependsOn": [{"activity": "Transform to Silver", "dependencyConditions": ["Succeeded"]}],
        "typeProperties": {
          "notebook": {"name": "nb_aggregate_slot_metrics"}
        }
      }
    ],
    "parameters": {
      "RunDate": {"type": "string", "defaultValue": "@utcNow('yyyy-MM-dd')"}
    }
  }
}

Sprint 6 Checklist:

• All fact tables migrated
• Historical data backfill complete
• Incremental patterns working
• Informatica workflows converted (80%+)
• Pipelines scheduled
• Monitoring alerts configured
• Data quality gates implemented
• Performance optimized (OPTIMIZE, Z-ORDER)

Sprint 7: BI & Analytics (Weeks 19-21)¶

Source: Semantic models in Microsoft Fabric

Objectives: - Build production semantic models - Migrate all reports and dashboards - Establish self-service analytics - Train business users

Semantic Model Design:

erDiagram
    DIM_DATE ||--o{ FACT_SLOT : "date_key"
    DIM_PLAYER ||--o{ FACT_SLOT : "player_id"
    DIM_MACHINE ||--o{ FACT_SLOT : "machine_id"
    DIM_PLAYER ||--o{ FACT_SESSION : "player_id"
    DIM_DATE ||--o{ FACT_SESSION : "date_key"
    DIM_PLAYER ||--o{ FACT_CAGE : "player_id"
    DIM_DATE ||--o{ FACT_CAGE : "date_key"

    DIM_DATE {
        int date_key PK
        date calendar_date
        int year
        int quarter
        int month
    }

    DIM_PLAYER {
        string player_id PK
        string name
        string loyalty_tier
        date registration_date
    }

    DIM_MACHINE {
        string machine_id PK
        string name
        string game_type
        decimal denomination
    }

    FACT_SLOT {
        bigint transaction_id PK
        int date_key FK
        string player_id FK
        string machine_id FK
        decimal coin_in
        decimal coin_out
    }

    FACT_SESSION {
        bigint session_id PK
        int date_key FK
        string player_id FK
        int duration_minutes
        decimal total_spend
    }

    FACT_CAGE {
        bigint transaction_id PK
        int date_key FK
        string player_id FK
        decimal amount
        string type
    }

Report Migration Approach:

Sprint 7 Checklist:

• Semantic models created
• All reports migrated
• Dashboards rebuilt
• Row-level security configured
• Report schedules configured
• User training completed
• Self-service enabled
• Documentation updated

Sprint 8: Real-time & AI/ML (Weeks 22-24)¶

Objectives: - Implement real-time streaming - Deploy ML models - Enable advanced analytics - Final integration testing

Real-time Architecture:

flowchart LR
    subgraph Sources["Event Sources"]
        SLOT[Slot Machines]
        POS[POS Systems]
        SENSOR[IoT Sensors]
    end

    subgraph Ingest["Ingestion"]
        EH[Event Hub]
        ES[Eventstream]
    end

    subgraph Process["Processing"]
        KQL[KQL Database]
        RT[Real-time<br/>Analytics]
    end

    subgraph Consume["Consumption"]
        DASH[Live<br/>Dashboard]
        ALERT[Alerts]
        ML[ML Scoring]
    end

    SLOT --> EH
    POS --> EH
    SENSOR --> EH

    EH --> ES
    ES --> KQL
    ES --> RT

    KQL --> DASH
    RT --> ALERT
    KQL --> ML

ML Model Deployment:

Sprint 8 Checklist:

• Eventstream configured
• KQL database created
• Real-time dashboards working
• ML models deployed
• Alerting configured
• Integration testing complete
• Performance testing complete
• Security testing complete

🚀 Production Cutover¶

Cutover Timeline¶

gantt
    title Production Cutover - 7 Day Plan
    dateFormat YYYY-MM-DD
    axisFormat %a %d

    section Preparation
    Final sync from source       :prep1, 2024-06-24, 1d
    Freeze source changes        :prep2, after prep1, 1d
    Pre-cutover validation       :prep3, after prep2, 1d

    section Cutover
    Stop source ETL              :cut1, after prep3, 4h
    Final data sync              :cut2, after cut1, 8h
    Switch connections           :cut3, after cut2, 2h
    Validation testing           :cut4, after cut3, 4h
    Go-Live                      :milestone, m1, after cut4, 0d

    section Hypercare
    Intensive monitoring         :hyp1, after m1, 3d
    Issue resolution             :hyp2, after hyp1, 4d
    Knowledge transfer           :hyp3, after hyp2, 3d
    Steady state                 :milestone, m2, after hyp3, 0d

Cutover Checklist¶

T-7 Days: Preparation

• Final UAT sign-off obtained
• Cutover runbook reviewed and approved
• Rollback plan documented and tested
• Communication sent to all stakeholders
• Support team briefed
• Monitoring dashboards ready
• War room scheduled

T-3 Days: Final Preparation

• Source system change freeze confirmed
• Full backup of source completed
• Full backup of target completed
• All team members confirmed available
• Escalation contacts verified
• Test cutover completed (dry run)

T-0: Cutover Day

T+1 to T+14: Hypercare

• 24/7 monitoring active
• Daily health check meetings
• Issue triage and resolution
• Performance monitoring
• User feedback collection
• Documentation updates
• Knowledge transfer sessions
• Transition to BAU support

🏢 Center of Excellence¶

CoE Structure¶

flowchart TB
    subgraph CoE["Center of Excellence"]
        LEAD[CoE Lead]

        subgraph Pillars["Pillars"]
            ARCH[Architecture<br/>& Standards]
            ENABLE[Enablement<br/>& Training]
            GOV[Governance<br/>& Security]
            OPS[Operations<br/>& Support]
        end

        LEAD --> ARCH
        LEAD --> ENABLE
        LEAD --> GOV
        LEAD --> OPS
    end

    subgraph Teams["Delivery Teams"]
        T1[Team 1<br/>Core Analytics]
        T2[Team 2<br/>Real-time]
        T3[Team 3<br/>Self-service]
    end

    ARCH --> T1
    ARCH --> T2
    ARCH --> T3

    style CoE fill:#9b59b6,color:#fff

CoE Responsibilities¶

CoE Artifacts¶

🔒 Security & Compliance¶

Source: Microsoft Fabric security overview

Security Architecture¶

flowchart TB
    subgraph Identity["Identity & Access"]
        ENTRA[Microsoft<br/>Entra ID]
        COND[Conditional<br/>Access]
        PIM[Privileged<br/>Identity Mgmt]
    end

    subgraph Network["Network Security"]
        PLINK[Private<br/>Link]
        FW[Firewall]
        NSG[Network<br/>Security Groups]
    end

    subgraph Data["Data Security"]
        ENCRYPT[Encryption<br/>at Rest]
        TLS[Encryption<br/>in Transit]
        RLS[Row-Level<br/>Security]
        DDM[Dynamic Data<br/>Masking]
    end

    subgraph Compliance["Compliance"]
        PURVIEW[Microsoft<br/>Purview]
        AUDIT[Audit<br/>Logging]
        DLP[Data Loss<br/>Prevention]
    end

    ENTRA --> COND --> PIM
    PLINK --> FW --> NSG
    ENCRYPT --> TLS --> RLS --> DDM
    PURVIEW --> AUDIT --> DLP

    style Identity fill:#3498db,color:#fff
    style Network fill:#e74c3c,color:#fff
    style Data fill:#27ae60,color:#fff
    style Compliance fill:#f39c12,color:#fff

Access Control Matrix¶

Compliance Requirements¶

📊 Success Metrics¶

KPIs Dashboard¶

🔧 Troubleshooting Guide¶

Common Issues¶

Escalation Path¶

Level 1: Self-service (documentation, FAQ)
    ↓
Level 2: CoE Support (office hours, tickets)
    ↓
Level 3: Platform Team (urgent issues)
    ↓
Level 4: Microsoft Support (product issues)

📚 Resources¶

Documentation¶

Training¶

🎉 Summary¶

This comprehensive migration guide provides:

• ✅ 6-month timeline with 8 sprints
• ✅ POC to Production complete journey
• ✅ Enterprise connectivity (ExpressRoute, VPN)
• ✅ Security and compliance framework
• ✅ Center of Excellence structure
• ✅ Detailed checklists for each phase
• ✅ Production cutover runbook
• ✅ Hypercare and steady-state operations

🧭 Navigation¶

⬆️ Back to Top | 📚 Tutorials | 🏠 Home