Migration, Disaster Recovery, and Financial Management

Workload Migration Strategy

Migration Center

What it is: A free, unified platform for end-to-end migration planning, discovery, assessment, business case/TCO generation, and migration execution guidance.

!! Key Distinction: Migration Center does not itself move workloads (that is done by tools like Migrate to Virtual Machines, Migrate to Containers, Database Migration Service, Storage Transfer Service, etc.) (brain not hands)

Core Capabilities (Exam-Relevant):

Discovery: Agentless or agent-based scanning of on-prem VMware/physical servers, AWS EC2 and Azure VM discovery, plus SQL Server, MySQL, and PostgreSQL database discovery.
Rapid Cost Estimation: Generate an offline/rapid Google Cloud cost estimate without activating Migration Center; useful for early-stage business cases.
Assessment: Produces a data-driven fit recommendation (e.g., “this VM is a good fit for Compute Engine N2 rightsized to X” or “this workload is a good fit for GKE”), and generates TCO reports based on migration preferences.
Network Dependency Mapping: Generates a network dependencies report to identify which servers/apps must move together during wave planning.
!! Exam pitfall: Migration Center ≠ a migration execution tool. Don’t confuse it with Migrate to Virtual Machines (the actual lift-and-shift VM migration tool) or Migrate to Containers (replatforming to GKE/Cloud Run).

Migration Waves (The “6 Rs” Strategies)

Strategy	Description	When to Use
Rehost (“lift and shift”)	Move VMs as-is, no code changes	Fastest path, large data centres on a deadline, low risk tolerance for change
Replatform (“lift, tinker, and shift”)	Minor optimizations during move (e.g., move DB to Cloud SQL, containerize with minimal refactor)	Want some cloud-native benefit without a full rewrite
Refactor / Re-architect	Rebuild app to be cloud-native (microservices, serverless)	Long-term strategic apps, high ROI justifies more time/cost
Repurchase	Replace with SaaS	Commodity workloads (e.g., CRM)
Retire	Decommission unused workloads	Discovered during Migration Center assessment, always look for retirement candidates first, it is the cheapest migration
Retain	Keep on-prem (hybrid)	Compliance/latency constraints, or not yet ready

!! Exam pattern: Case studies almost always recommend rehost first for speed/cost reduction, then replatform/refactor iteratively post-migration (“migrate then modernize”).

Google Cloud VMware Engine (GCVE) & Landing Zones

GCVE: Dedicated, single-tenant VMware SDDC (vSphere, vCenter, vSAN, NSX, HCX) on Google bare metal. Uses HCX for live migration with near-zero downtime.
Landing Zone: Pre-migration foundational secure multi-project environment (hierarchy, IAM, Shared VPC, org policies).

Disaster Recovery (DR) & Backup

Backup and DR Service

Backup Vault: Immutable, indelible backup storage logically air-gapped in Google-managed projects. Essential for ransomware and insider threat protection.

RTO / RPO & DR Readiness Patterns

Pattern	RTO/RPO	Description	Cost
Backup & Restore	Highest RTO (hours+), RPO = backup interval	Backups stored in GCS; restore on demand	Lowest
Pilot Light	Minutes–hours	Core minimal infrastructure (e.g., DB replica) running in DR region; scale up on failover	Low
Warm Standby	Minutes	Scaled-down but functional copy running in DR region; scale up on failover	Medium
Active-Active (Hot Standby)	Near-zero (seconds)	Full production capacity running simultaneously in 2+ regions; instant traffic shift	Highest

Cost Estimation & FinOps Foundations

Cost Estimation Tools & Framework

Google Cloud Pricing Calculator: Used to construct early-stage architectural cost models prior to deployment.
Migration Center Rapid Cost Estimator: Generates fast, offline GCP cost estimates based on raw CPU/RAM inventory before activating full Migration Center discovery.
Migration Center TCO Analysis: Compares on-premises CapEx/OpEx (hardware, power, cooling, virtualization licensing, labour) against projected GCP 1-year/3-year CUD rates.
Billing Export to BigQuery: Foundational tool for detailed cost attribution, label filtering, and custom Looker Studio cost dashboards.
Budgets & Spend Alerts: Set threshold notifications (email or Pub/Sub). !! Budgets do not automatically cap spend; custom Pub/Sub automation is required to enforce hard spend limits.

Comprehensive Cost Optimization Strategies

Compute Cost Optimization

Active Assist / Recommender: Automatically analyzes VM performance telemetry and identifies underutilized instances. Always right-size instances BEFORE purchasing CUD commitments.
Committed Use Discounts (CUDs):
- Resource-Based CUDs: 1 or 3-year commitments to specific vCPU/RAM in a specific region (up to 55-70% discount). Best for stable baselines.
- Flexible CUDs (“Flex CUDs”): Dollar-per-hour spend commitment applying across machine families, regions, GKE, and Cloud Run (28-46% discount). Best for dynamic/modernizing workloads.
Spot VMs: Use stateless, fault-tolerant batch/container workloads on Spot VMs for up to 60-91% discount over standard rates.
Serverless Scale-to-Zero: Leverage Cloud Run and Cloud Functions so compute billing drops to zero when requests cease.
GKE Autopilot: Pay strictly for Pod resource requests rather than paying for idle worker node VM capacity.

Storage & Operational Cost Optimization

Cloud Storage Lifecycle Rules: Automatically transition objects from Standard → Nearline (30d) → Coldline (90d) → Archive (365d) or trigger auto-deletion.
BigQuery Cost Optimization: Enforce Partition Pruning and Clustering on queries to minimize bytes scanned; avoid SELECT *.
Log Exclusion Filters: Filter out noisy debug logs at the Log Router before ingestion; route compliance audit logs to low-cost Cloud Storage sinks.

Well-Architected Framework (Domain 7 Lens)

Pillar	Domain 7 Application
Operational Excellence	Migration Center discovery/assessment; IaC-based landing zones.
Reliability	RTO/RPO-matched DR architecture (pilot light → active-active); Backup Vault air-gaps.
Security	Backup Vault ransomware protection; pre-migration landing zone security guardrails.
Cost Optimization	Active Assist rightsizing; Flex CUDs; Spot VMs; GCS lifecycle management; log exclusion sinks.
Performance Optimization	Replatforming workloads to high-performance engines (AlloyDB, Spanner).
Sustainability	Retiring unused on-prem workloads discovered via Migration Center.