GCP Study Hub

Migration

GCP Migration Tools

Practical tools for executing cloud migration

AWS equivalent

AWS MGN / DMS / Snowball equivalents noted

MigrationRound 2 FocusTools

Architecture Diagram

Migration Journey & Tools

1. Assess
→ Migration Center
→ StratoZone
→ Cloud Foundation Toolkit
›
2. Migrate VMs
→ Migrate to VMs
→ Migrate to Containers
→ Transfer Appliance
›
3. Migrate Data
→ Database Migration Service
→ Storage Transfer Service
→ BigQuery Migration Service
›
4. Operate
→ Cloud Monitoring
→ Cloud Logging
→ Active Assist
🔄

AWS → GCP: Key Differences

  • ▸

    Migrate to VMs ≈ AWS Application Migration Service (MGN). Continuous replication, minimal cutover downtime.

  • ▸

    Database Migration Service ≈ AWS DMS. Online database migration with CDC.

  • ▸

    Transfer Appliance ≈ AWS Snowball. Physical device for offline data transfer.

📌

Key Concepts to Know

  • 1

    🔵 Migration Center: free assessment. Discovers on-prem VMs, databases, apps. Maps dependencies.

  • 2

    🔵 Migrate to Virtual Machines: agent-based continuous replication from on-prem/AWS/Azure to GCP. Cutover in minutes.

  • 3

    🔵 Migrate to Containers: analyzes VMs and generates Dockerfiles + Kubernetes manifests.

  • 4

    🟢 Database Migration Service (DMS): online migration with CDC. Supports MySQL, PostgreSQL, SQL Server.

  • 5

    🟢 Storage Transfer Service: scheduled/one-time transfers from S3, Azure Blob, on-prem to GCS.

  • 6

    🟡 Transfer Appliance: for 100TB–1PB offline transfers. Data encrypted on device.

  • 7

    🟡 BigQuery Migration Service: migrate Redshift, Teradata to BigQuery. Translates SQL dialects.

💡

DCE Interview Tips

  • ★

    Security during migration: 'All transfers via DMS are TLS encrypted. For large bulk transfers, Transfer Appliance never touches the public internet. Post-transfer, we validate using checksums.'

  • ★

    Migration assessment pitch: 'We start every engagement with Migration Center — it gives us a full inventory, maps dependencies, and generates a rough cost estimate for GCP.'

  • ★

    Parallel running: 'Before cutover, we run both environments in parallel and compare outputs. Only after sign-off do we cut over and decommission the source.'

⚠️

Common Gotchas

  • !

    DMS requires network connectivity between source DB and GCP.

  • !

    Transfer Appliance has a lead time of several weeks — plan early.

  • !

    SQL dialect translation (BigQuery Migration Service) is not perfect — always validate.