Project Overview
4-Tower LLC, provider of SpecAdvisor—a leading medical equipment planning and procurement software for healthcare facilities—approached us with a critical challenge: their decade-old application was running on outdated infrastructure that couldn’t scale to meet growing demand. We executed a complete legacy modernization, transforming a fragile monolith into a robust, containerized, serverless cloud architecture.
The Challenge
The Legacy State
- Single EC2 Instance: Entire application running on one t2.large instance
- Outdated PHP: PHP 5.6 (end-of-life, security vulnerabilities)
- Legacy CRM: Old SugarCRM installation with custom modifications
- Monolithic MySQL: Single database instance, no backups, no failover
- No Environments: Development done directly on production
- Manual Deployments: SSH and FTP-based deployments
- Angular 1.x: Outdated frontend framework
- No Monitoring: Zero visibility into system health
- Security Risks: No WAF, hardcoded credentials, no encryption at rest
Business Impact of Legacy Stack
- Frequent downtime affecting healthcare facility operations
- Unable to onboard new customers due to capacity constraints
- Security audit failures preventing enterprise contracts
- Technical debt blocking feature development
- Developer frustration with brittle codebase
Our Solution
Phased Modernization Strategy
We implemented a zero-downtime migration using the strangler fig pattern, progressively modernizing components while maintaining service continuity.
Modern Cloud Architecture
Containerization
- Dockerized PHP application with multi-stage builds
- ECS Fargate for container orchestration
- Blue-green deployment strategy
- Auto-scaling based on CPU and memory metrics
Database Modernization
- Migrated MySQL to RDS Aurora Serverless v2
- Automated daily backups with 30-day retention
- Multi-AZ deployment for high availability
- Read replicas for reporting workloads
- Database encryption at rest and in transit
Frontend Optimization
- Angular application modernization
- Static assets moved to S3
- CloudFront CDN for global distribution
- Gzip and Brotli compression
- Lazy loading and code splitting
Serverless Backend
- AWS Lambda for background tasks
- API Gateway for RESTful endpoints
- EventBridge for scheduled jobs
- SQS for asynchronous processing
- Step Functions for complex workflows
Authentication & Security
- AWS Cognito for user management
- MFA support for admin users
- SSO integration for enterprise customers
- WAF with managed rule sets
- Secrets Manager for credential rotation
Storage & Assets
- S3 for document storage
- S3 Lifecycle policies for archival
- CloudFront signed URLs for secure access
- Medical equipment spec sheets and drawings
- Project documentation and reports
Technology Stack
Application Layer
- Backend: PHP 8.2 with Composer
- Framework: Laravel 10.x (migrated from custom PHP)
- Frontend: Angular 15 with TypeScript
- API: RESTful with OpenAPI documentation
AWS Infrastructure
- Compute: ECS Fargate, Lambda (Node.js runtime)
- Database: RDS Aurora Serverless v2 (MySQL compatible)
- Storage: S3 Standard/Intelligent-Tiering
- CDN: CloudFront with custom SSL
- Authentication: Cognito User Pools
- API: API Gateway HTTP APIs
- Orchestration: Step Functions
- Messaging: SQS, SNS, EventBridge
- Monitoring: CloudWatch, X-Ray
- Security: WAF, Secrets Manager, GuardDuty
DevOps & CI/CD
- Version Control: GitHub with branch protection
- CI/CD: GitHub Actions
- Containers: Docker, ECR
- Infrastructure: Terraform
- Monitoring: New Relic, CloudWatch Dashboards
- Logging: CloudWatch Logs with retention policies
Development Environments
- Development: Isolated ECS cluster, RDS snapshot
- QA/Staging: Production-like environment for testing
- Production: Multi-AZ, auto-scaling, blue-green deployments
Development Process
Phase 1: Assessment & Planning (3 weeks)
- Complete infrastructure audit
- Dependency mapping
- Security vulnerability assessment
- Database schema analysis
- Customer usage pattern analysis
- Modernization roadmap creation
Phase 2: Foundation (6 weeks)
- Set up DEV/QA/PROD environments
- Migrate version control to Git
- Establish CI/CD pipelines
- Dockerize application
- Create RDS test instance
- Implement monitoring baselines
Phase 3: PHP Upgrade (8 weeks)
- PHP 5.6 → 7.4 compatibility fixes
- PHP 7.4 → 8.0 migration
- PHP 8.0 → 8.2 optimization
- Legacy code refactoring
- Deprecation warnings resolution
- Performance testing at each step
Phase 4: Database Migration (6 weeks)
- Schema optimization
- Create RDS Aurora cluster
- Data migration scripts
- Sync mechanisms for parallel run
- Cutover planning and execution
- Rollback procedures testing
Phase 5: Containerization (8 weeks)
- Multi-stage Dockerfile creation
- ECS task definitions
- Load balancer configuration
- Auto-scaling policies
- Blue-green deployment setup
- Container health checks
Phase 6: Frontend Modernization (10 weeks)
- Angular 1.x → Angular 15 migration
- TypeScript integration
- Component architecture redesign
- State management with RxJS
- S3 + CloudFront deployment
- Progressive Web App features
Phase 7: Serverless Integration (6 weeks)
- Lambda functions for background jobs
- API Gateway endpoints
- Event-driven architecture
- Workflow automation with Step Functions
- Email notifications via SES
- Report generation optimization
Phase 8: Security Hardening (4 weeks)
- Cognito user pool creation
- User migration from legacy auth
- WAF rule implementation
- Secrets Manager integration
- Encryption at rest enablement
- Security audit and penetration testing
Phase 9: Monitoring & Optimization (3 weeks)
- CloudWatch dashboard creation
- Alert configuration
- Performance tuning
- Cost optimization
- Documentation
- Team training
Key Features Implemented
Multi-Environment Strategy
- Development: Rapid iteration, developer sandboxes
- QA: Automated testing, staging for client demos
- Production: HA, auto-scaling, blue-green deployments
- DR: Cross-region RDS snapshots, S3 replication
Automated Backup & Recovery
- Automated RDS snapshots every 6 hours
- Point-in-time recovery up to 35 days
- S3 versioning for documents
- Disaster recovery runbook
- Quarterly DR drills
Authentication System
- Cognito-based user management
- Role-based access control (RBAC)
- MFA for administrators
- SSO for enterprise customers
- Password policies enforcement
- Session management
Performance Optimization
- ECS task auto-scaling (2-20 tasks)
- RDS Aurora auto-scaling (0.5-16 ACUs)
- CloudFront edge caching
- Database query optimization
- API response caching with ElastiCache
- Lazy loading for large datasets
Monitoring & Observability
- Real-time CloudWatch dashboards
- Application performance monitoring (New Relic)
- Distributed tracing with X-Ray
- Log aggregation and analysis
- PagerDuty integration for incidents
- Weekly SRE review meetings
Results & Impact
Performance Improvements
- Uptime: From 94% to 99.95% (60% reduction in downtime)
- Response Time: API latency reduced from 2.5s to 180ms (93% faster)
- Page Load: Frontend load time from 8s to 1.2s (85% improvement)
- Scalability: From 100 to 5,000+ concurrent users
- Deployment Time: From 4 hours to 12 minutes
Cost Optimization
- Infrastructure Costs: 35% reduction despite increased capacity
- RDS Serverless: Only pay for actual usage
- ECS Fargate: No idle EC2 instances
- S3 Lifecycle: Automatic archival saving 40% on storage
- Lambda: Pay-per-invocation model
- Reserved Instances: Strategic use for predictable workloads
Business Impact
- New Customers: Onboarded 45 new healthcare facilities in 6 months
- Enterprise Deals: Closed 3 major contracts requiring security compliance
- Feature Velocity: Development speed increased 3x
- Customer Satisfaction: NPS score improved from 42 to 78
- Technical Debt: Reduced by 70%
- Developer Retention: Zero turnover post-modernization
Security & Compliance
- HIPAA Readiness: Architecture aligned with HIPAA requirements
- SOC 2: Passed audit on first attempt
- Zero Breaches: No security incidents post-migration
- Encryption: 100% of data encrypted at rest and in transit
- Access Controls: Fine-grained IAM policies
- Audit Trails: Complete CloudTrail logging
Operational Excellence
- Zero-Downtime Deployments: Blue-green strategy eliminates outages
- Automated Rollbacks: Failed deployments auto-revert in < 2 minutes
- Infrastructure as Code: 100% of infrastructure in Terraform
- Documentation: Comprehensive runbooks and architecture diagrams
- Team Confidence: Developers empowered to deploy daily
Technical Deep Dive
PHP 5 → PHP 8 Migration Strategy
The PHP upgrade was the most challenging aspect, requiring careful planning:
Compatibility Analysis
- Automated scanning with PHPCompatibility
- Manual code review for breaking changes
- Third-party dependency updates
- Custom SugarCRM modifications rewriting
Step-by-Step Approach
- PHP 5.6 → 7.0: Removed deprecated functions
- PHP 7.0 → 7.4: Addressed type juggling changes
- PHP 7.4 → 8.0: Union types and named arguments
- PHP 8.0 → 8.2: JIT compiler optimization
Performance Gains
- 60% faster execution time
- 40% memory reduction
- Better CPU utilization
- Improved garbage collection
Containerization Best Practices
# Multi-stage build for optimized images
FROM php:8.2-fpm-alpine AS builder
# Build dependencies and Composer install
RUN apk add --no-cache git zip
COPY composer.json composer.lock ./
RUN composer install --no-dev --optimize-autoloader
FROM php:8.2-fpm-alpine
# Runtime dependencies only
COPY --from=builder /app/vendor ./vendor
# Healthcare-specific extensions
RUN docker-php-ext-install pdo_mysql opcache
# Security hardening
RUN adduser -D -u 1000 appuser
USER appuser
RDS Aurora Serverless Configuration
- Min Capacity: 0.5 ACUs (idle periods)
- Max Capacity: 16 ACUs (peak loads)
- Auto-Pause: After 5 minutes of inactivity
- Scale Points: Smart scaling during low-activity periods
- Connection Pooling: RDS Proxy for Lambda connections
ECS Task Definition
{
"cpu": "512",
"memory": "1024",
"requiresCompatibilities": ["FARGATE"],
"networkMode": "awsvpc",
"containerDefinitions": [{
"name": "app",
"image": "4tower/specadvisor:latest",
"healthCheck": {
"command": ["CMD-SHELL", "curl -f http://localhost/health || exit 1"],
"interval": 30,
"timeout": 5,
"retries": 3
}
}]
}
Migration Challenges & Solutions
Challenge 1: Zero-Downtime Migration
Problem: Healthcare facilities operate 24/7; downtime unacceptable
Solution:
- Blue-green deployment strategy
- Database replication with bidirectional sync
- Feature flags for gradual rollout
- Comprehensive rollback procedures
Challenge 2: Custom SugarCRM Modifications
Problem: Heavily customized CRM with undocumented changes
Solution:
- Code archaeology to document customizations
- Extracted to separate microservice
- API-based integration
- Gradual feature parity in new system
Challenge 3: Large Database Migration
Problem: 500GB database with complex relationships
Solution:
- AWS Database Migration Service (DMS)
- Continuous replication during transition
- Data validation scripts
- Phased cutover by customer segment
Challenge 4: Legacy Angular 1.x Frontend
Problem: Complete rewrite would take too long
Solution:
- Hybrid approach using ngUpgrade
- Component-by-component migration
- Maintained feature parity throughout
- TypeScript for new components
Challenge 5: Hardcoded Secrets
Problem: Database credentials and API keys in code
Solution:
- Secrets Manager for all credentials
- Automated rotation for RDS passwords
- IAM roles instead of access keys
- Audit of all hardcoded values
Client Testimonial
“Greicodex took our decade-old application and transformed it into a modern, scalable platform. The migration was seamless—our customers didn’t experience a single minute of downtime. We can now confidently pursue enterprise contracts knowing our infrastructure meets the highest security standards. The investment has already paid for itself through new customer acquisition.”
— CTO, 4-Tower LLC
Technologies Used
- Backend: PHP 8.2, Laravel 10.x
- Frontend: Angular 15, TypeScript, RxJS
- Containers: Docker, Amazon ECS Fargate, ECR
- Database: Amazon RDS Aurora Serverless v2 (MySQL)
- Serverless: AWS Lambda, API Gateway, Step Functions
- Storage: S3, CloudFront CDN
- Authentication: AWS Cognito
- Monitoring: CloudWatch, New Relic, X-Ray
- Security: WAF, Secrets Manager, GuardDuty
- CI/CD: GitHub Actions, Terraform
- Messaging: SQS, SNS, EventBridge
Lessons Learned
- Incremental Migration Wins: Strangler fig pattern reduced risk vs. big bang
- PHP Version Jumps: Skipping versions is dangerous; step-by-step is safer
- Container Sizing: Right-sizing saved 40% on compute costs
- RDS Serverless: Perfect for variable workloads in healthcare
- Documentation Matters: Legacy code archaeology was time-consuming
- Customer Communication: Weekly updates built trust during migration
- Feature Flags: Essential for risk-free deployments
- Monitoring First: Should have been implemented earlier in the process
Long-Term Benefits
Technical
- Modern tech stack attracting top talent
- Reduced technical debt enabling faster development
- Infrastructure as Code for reproducible environments
- Comprehensive test coverage (now at 78%)
Business
- Competitive advantage in enterprise market
- Reduced support burden from stability improvements
- Faster time-to-market for new features
- Foundation for mobile app development
Operational
- Automated operations reducing manual toil
- Improved developer experience and productivity
- Clear upgrade path for future enhancements
- Disaster recovery confidence
Want to Modernize Your Legacy Application?
We specialize in zero-downtime migrations of mission-critical healthcare applications. Our proven methodologies minimize risk while maximizing business value.