Cloud Migration Guide for IT Managers

For: IT managers and CIOs planning cloud migration
Goal: Successfully migrate to cloud (AWS, Azure, or GCP)
Outcome: 30-50% cost savings, increased agility, reduced risk

Why Migrate to the Cloud?

Business Drivers

Cost Savings:

• 30-50% reduction in infrastructure costs
• Pay-per-use (no upfront CapEx)
• Eliminate data center expenses

Agility & Speed:

• Deploy infrastructure in minutes (vs. weeks)
• Scale up/down on demand
• Faster time-to-market for new products

Innovation:

• Access to AI/ML, analytics, IoT services
• Focus on business value, not infrastructure
• Continuous updates and new features

Reliability:

• 99.99% SLA (vs. 99.5% on-prem typical)
• Built-in redundancy and failover
• Global reach

Cloud Provider Comparison

| Feature | AWS | Azure | GCP | |---------|-----|-------|-----| | Market Share | 32% (#1) | 23% (#2) | 10% (#3) | | Best For | Startups, tech companies | Microsoft shops, enterprises | Data/AI workloads | | Strengths | Broadest service catalog, mature | Microsoft integration, hybrid cloud | Data analytics, Kubernetes, pricing | | Weaknesses | Complex pricing, learning curve | UI complexity | Smaller ecosystem, fewer services | | IaaS | EC2, EBS, VPC | Virtual Machines, Disks | Compute Engine | | PaaS | Elastic Beanstalk, Lambda | App Service, Functions | App Engine, Cloud Run | | Database | RDS, DynamoDB, Aurora | SQL Database, Cosmos DB | Cloud SQL, Spanner | | Storage | S3, EBS, EFS | Blob, Files, Disks | Cloud Storage | | AI/ML | SageMaker | Azure ML | Vertex AI | | Pricing | Pay-as-you-go | Pay-as-you-go | Pay-as-you-go (often 10-20% cheaper) |

Recommendation:

• AWS if: Greenfield, need broadest services
• Azure if: Heavy Microsoft (AD, Office 365, .NET)
• GCP if: Data/analytics focus, Kubernetes
• Multi-cloud if: Avoid vendor lock-in (but adds complexity)

The 7 R's of Cloud Migration

Migration Strategies

1. Rehost ("Lift and Shift")

• What: Move as-is to cloud (VM to VM)
• Pros: Fast (weeks), low risk, minimal code changes
• Cons: Doesn't optimize for cloud, misses cost savings
• Best For: Legacy apps, quick wins, test workloads
• Tools: AWS Application Migration Service, Azure Migrate

2. Replatform ("Lift, Tinker, Shift")

• What: Migrate with minor optimizations
• Example: On-prem SQL Server → AWS RDS (managed database)
• Pros: Some cloud benefits, still fast
• Cons: Not fully cloud-native
• Best For: Databases, middleware

3. Refactor ("Re-architect")

• What: Redesign for cloud-native
• Example: Monolith → Microservices + containers
• Pros: Maximum cloud benefits (cost, scale, features)
• Cons: Expensive, time-consuming (months)
• Best For: Strategic apps, technical debt cleanup

4. Repurchase ("Drop and Shop")

• What: Replace with SaaS
• Example: On-prem Exchange → Microsoft 365
• Pros: No infrastructure to manage
• Cons: Less customization, subscription costs
• Best For: Email, CRM, HR systems

5. Retain ("Revisit")

• What: Keep on-premises for now
• Reasons: Compliance, latency, not ready
• Timeline: Migrate later or never

6. Retire

• What: Decommission unused applications
• Impact: 10-20% of portfolio (discovered during migration)
• Savings: Immediate cost reduction

7. Relocate

• What: Move to cloud without changes (hypervisor-level)
• Example: VMware on-prem → VMware Cloud on AWS
• Best For: VMware environments, hybrid cloud

Cloud Migration Roadmap (6-12 Months)

Phase 1: Assessment & Planning (4-8 weeks)

Step 1: Discover & Inventory (1-2 weeks)

Catalog all applications and infrastructure:

• Servers (physical, virtual)
• Databases
• Storage
• Network dependencies
• Licensing

Discovery Tools:

• AWS: Application Discovery Service
• Azure: Azure Migrate
• Third-party: Device42, ServiceNow, CloudPhysics

Application Assessment:

| Application | Category | Users | Criticality | Migration Complexity | Strategy | |-------------|----------|-------|-------------|---------------------|----------| | Email (Exchange) | Communication | 500 | Critical | Low | Repurchase (→ M365) | | ERP (SAP) | Business | 200 | Critical | High | Rehost → Refactor (Phase 2) | | File Server | Storage | 500 | High | Low | Replatform (→ S3/Azure Files) | | Dev/Test | Internal | 50 | Low | Low | Rehost | | Legacy CRM | Business | 10 | Low | N/A | Retire |

Step 2: Define Cloud Strategy (1-2 weeks)

Key Decisions:

1. Cloud Provider(s)

• Single cloud (simpler) vs. multi-cloud (avoid lock-in)
• Primary cloud for majority of workloads

2. Migration Approach

• Big bang (migrate all at once) - Higher risk
• Phased (pilot → non-prod → prod) - Recommended

3. Timeline

• Aggressive: 6 months
• Moderate: 12 months
• Conservative: 18-24 months

4. Budget

• Migration costs (tools, services, consulting)
• Run rate (ongoing cloud costs)

Step 3: Build Business Case (1-2 weeks)

TCO Comparison (5 Year):

| Cost Category | On-Premises | Cloud | Savings | |---------------|-------------|-------|---------| | Hardware | $500K | $0 | $500K | | Data Center | $300K | $0 | $300K | | Software Licenses | $200K | $150K | $50K | | Cloud Services | $0 | $1.5M | -$1.5M | | Personnel | $2M | $1.5M | $500K | | Total 5-Year | $3M | $3.15M | -$150K |

Wait, cloud is MORE expensive?!

Often true for pure cost comparison, BUT:

Add Strategic Benefits:

• Faster time-to-market ($1M value)
• Improved availability 99.5% → 99.99% ($500K avoided downtime)
• Innovation capacity (AI/ML, analytics)
• Scalability for growth
• True ROI: Positive after strategic value

Step 4: Plan Pilot Migration (1-2 weeks)

Select Pilot Workload:

• Non-critical (low risk if fails)
• Representative (similar to other apps)
• Complete stack (app + database + dependencies)
• Examples: Dev/test environment, internal tools

Pilot Goals:

• Validate migration process
• Train team
• Identify issues
• Build confidence

Timeline: 2-4 weeks for pilot

Phase 2: Foundation Setup (2-4 weeks)

Cloud Account Setup

AWS:

• Create AWS Organization
• Set up accounts per environment (dev, test, prod)
• Configure IAM roles and policies
• Enable CloudTrail, Config, GuardDuty

Azure:

• Create management groups
• Set up subscriptions per environment
• Configure Azure AD and RBAC
• Enable Security Center, Monitor

GCP:

• Create organization and folders
• Set up projects per environment
• Configure IAM and service accounts
• Enable Cloud Security Command Center

Network Architecture

Hybrid Connectivity:

Option 1: Site-to-Site VPN

• Speed: Up to 1 Gbps
• Cost: $50-100/month
• Setup: Hours
• Best For: Low/moderate traffic

Option 2: Direct Connect / ExpressRoute / Cloud Interconnect

• Speed: 1-100 Gbps
• Cost: $500-5K+/month
• Setup: 4-12 weeks
• Best For: High traffic, low latency needs

Cloud Network Design:

┌─────────────────────────────────────┐
│         Cloud Account               │
│                                     │
│  ┌──────────────────────────────┐  │
│  │   VPC / Virtual Network      │  │
│  │                              │  │
│  │  ┌────────────┐ ┌─────────┐ │  │
│  │  │Public Subnet│ │Private │ │  │
│  │  │(Web Tier)  │ │Subnet  │ │  │
│  │  │            │ │(App/DB)│ │  │
│  │  └────────────┘ └─────────┘ │  │
│  └──────────────────────────────┘  │
│                ↕                    │
│         VPN/Direct Connect          │
└─────────────────────────────────────┘
               ↕
      On-Premises Network

Security & Compliance

Identity & Access:

• Federation (on-prem AD → cloud IAM)
• Multi-factor authentication (MFA)
• Least privilege access

Data Protection:

• Encryption at rest (all storage)
• Encryption in transit (TLS 1.2+)
• Key management (AWS KMS, Azure Key Vault)

Compliance:

• Enable audit logging
• Configure compliance monitoring
• Data residency (choose correct region)
• Shared responsibility model (understand what you vs. cloud provider secures)

Phase 3: Migrate Workloads (12-40 weeks)

Migration Wave Planning

Wave 1: Pilot (Week 1-4)

• Dev/test environments
• 1-2 low-risk applications
• Goal: Validate process

Wave 2: Quick Wins (Week 5-12)

• File servers → S3/Azure Files
• Backup/archive → Cloud storage
• Goal: Build momentum, immediate value

Wave 3: Non-Production (Week 13-24)

• Development environments
• Test environments
• Staging
• Goal: Finalize production migration plan

Wave 4: Production (Week 25-40)

• Business applications (by priority)
• Databases
• Goal: Complete migration

Migration Execution Checklist

Pre-Migration:

• ✅ Application dependencies mapped
• ✅ Network connectivity tested
• ✅ Security baselines configured
• ✅ Backup of source system
• ✅ Rollback plan documented
• ✅ Stakeholders notified (downtime window)

During Migration:

• ✅ Database replication setup
• ✅ Application server migrated
• ✅ Data synchronized
• ✅ DNS cutover (when ready)
• ✅ Smoke testing
• ✅ Monitor for issues

Post-Migration:

• ✅ Full testing (functionality, performance)
• ✅ User acceptance testing (UAT)
• ✅ Performance monitoring (24-48 hours)
• ✅ Decommission source (after 30 days stable)
• ✅ Lessons learned documented

Phase 4: Optimize & Modernize (Ongoing)

Cost Optimization

1. Right-Size Instances

• Problem: Over-provisioned VMs (8 vCPU when 2 vCPU sufficient)
• Solution: Analyze utilization, downsize
• Savings: 30-50%

2. Reserved Instances / Savings Plans

• Problem: Pay-as-you-go is expensive for steady workloads
• Solution: 1-3 year commitments
• Savings: 30-70%

3. Auto-Scaling

• Problem: Fixed capacity even when idle
• Solution: Scale up/down based on demand
• Savings: 20-40%

4. Spot Instances (AWS) / Low-Priority VMs (Azure)

• Problem: Paying full price for non-critical workloads
• Solution: Use spare capacity at 60-90% discount
• Best For: Batch jobs, dev/test, fault-tolerant apps

5. Storage Tiering

• Problem: All data on expensive hot storage
• Solution: Lifecycle policies (hot → warm → cold/archive)
• Savings: 50-90% on infrequently accessed data

6. Eliminate Waste

• Problem: Unused resources (forgotten VMs, old snapshots)
• Solution: Regular audits, automated cleanup
• Savings: 10-30%

Cost Optimization Tools:

• AWS Cost Explorer, Trusted Advisor
• Azure Cost Management
• GCP Recommender
• Third-party: CloudHealth, Spot.io, ProsperOps

Performance Optimization

Database:

• Migrate to managed services (RDS, Azure SQL)
• Use read replicas for read-heavy workloads
• Implement caching (Redis, Memcached)

Application:

• Use CDN for static content (CloudFront, Azure CDN)
• Enable auto-scaling
• Optimize code and queries

Network:

• Use region close to users
• Implement load balancing
• Optimize data transfer

Modernization Roadmap

Year 1: Lift-and-shift (rehost)
Year 2: Optimize and replatform
Year 3: Refactor to cloud-native (containers, serverless)

Cloud-Native Architecture:

• Microservices (vs. monolith)
• Containers (Docker, Kubernetes)
• Serverless (Lambda, Azure Functions)
• Managed services (reduce operational burden)

Common Migration Challenges

Challenge #1: Unexpected Costs

Problem: Cloud bill 2-3× higher than estimated

Causes:

• Data transfer costs (egress fees)
• Over-provisioned instances
• Lack of cost monitoring
• Dev/test environments left running 24/7

Prevention:

• Budget alerts (notify when 80% of budget)
• Tagging strategy (track costs by app/department)
• Regular cost reviews (weekly first 3 months)
• Shut down non-prod overnight/weekends

Challenge #2: Performance Issues

Problem: Application slower in cloud than on-prem

Causes:

• Network latency (distance to data center)
• Under-provisioned instances
• I/O bottlenecks (slow storage)
• Inefficient database queries

Prevention:

• Performance testing BEFORE migration
• Right-size instances (not too small)
• Use premium storage for databases
• Optimize application code

Challenge #3: Security Concerns

Problem: Misconfigured security, data breaches

Causes:

• Public S3 buckets (default in AWS)
• Weak IAM policies
• No network segmentation
• Unencrypted data

Prevention:

• Security baselines (CIS benchmarks)
• Enable cloud security tools (GuardDuty, Security Center)
• Regular security audits
• Principle of least privilege

Challenge #4: Skills Gap

Problem: Team lacks cloud expertise

Solutions:

• Training: AWS/Azure/GCP certifications
• Hiring: Cloud architects, engineers
• Partners: Engage cloud consultants for migration
• Managed Services: Outsource operations to MSP

Migration Tools & Services

Cloud-Native Tools

AWS:

• Application Migration Service (rehost)
• Database Migration Service (databases)
• DataSync (file servers)
• Snow Family (large data transfer)

Azure:

• Azure Migrate (assessment + migration)
• Database Migration Service
• Data Box (large data transfer)

GCP:

• Migrate for Compute Engine
• Database Migration Service
• Transfer Appliance

Third-Party Tools

| Tool | Purpose | Best For | |------|---------|----------| | CloudEndure | Rehost migrations | Large-scale lift-and-shift | | Carbonite | Disaster recovery + migration | Continuous replication | | Zerto | DR + migration | VMware environments | | Velostrata | Migration without downtime | Production workloads |

Professional Services

Cloud Provider Services:

• AWS Professional Services
• Azure Consulting Services
• Google Cloud Professional Services
• Cost: $200-400/hour

Third-Party Consultants:

• Accenture, Deloitte, Capgemini (large engagements)
• Onica (ClearScale), CloudReach, 2nd Watch (mid-market)
• Cost: $150-300/hour

When to Hire:

• Large migration (>100 servers)
• Lack internal expertise
• Complex architecture
• Tight timeline

Cloud Governance & FinOps

Cloud Center of Excellence (CCoE)

Purpose: Govern cloud usage, share best practices

Team:

• Cloud Architect (technical leadership)
• FinOps Manager (cost optimization)
• Security Lead (cloud security)
• DevOps Engineers (automation, tooling)

Responsibilities:

• Define cloud standards
• Approve cloud accounts and projects
• Cost governance
• Training and enablement
• Vendor management

FinOps (Financial Operations)

FinOps = Bringing financial accountability to cloud

FinOps Lifecycle:

1. Inform

• Visibility into cloud spending
• Allocation and chargeback
• Forecasting

2. Optimize

• Right-sizing
• Reserved instances
• Eliminate waste

3. Operate

• Real-time cost awareness
• Continuous optimization
• Cultural change

FinOps Tools:

• CloudHealth
• Cloudability
• Spot.io
• Native cloud tools

Key Takeaways

✅ Start with pilot - Don't migrate everything at once
✅ Plan for 12-18 months - Cloud migration is a journey
✅ Budget 20% over estimate - Unexpected costs will occur
✅ Optimize continuously - Migration ≠ done, it's beginning
✅ Train your team - Cloud skills are critical
✅ Monitor costs weekly - Cloud bills can spiral quickly
✅ Security from day one - Misconfigurations = breaches

Resources

Templates:

• IT Budget Planning Template - Cloud cost planning
• IT Asset Inventory - Track cloud resources

Related Guides:

• IT Budget Planning
• IT Security Roadmap
• IT Vendor Management

Cloud Training:

• AWS Certified Solutions Architect
• Azure Administrator Associate
• Google Cloud Professional Cloud Architect

FinOps:

• FinOps Foundation: finops.org

Conclusion

Cloud migration is inevitable for most organizations. Done right, it delivers 30-50% cost savings, massive agility gains, and competitive advantage.

Start your cloud journey:

1. Assess current state (applications, dependencies, costs)
2. Define strategy (pilot approach)
3. Build business case (TCO + strategic value)
4. Migrate pilot (2-4 weeks)
5. Scale migration (waves over 12-18 months)
6. Optimize continuously (FinOps culture)

In 12-18 months, you'll wonder how you ever ran IT without the cloud.

Migrating to cloud or already there? Share your experiences! 💬☁️