Cloud Project Case

Cloud Migration, Cost Optimization, Monitoring & CI/CD Automation for a Client

AWS Cloud GitLab CI/CD Grafana Cost Optimization Direct Connect
Cloud Cost Trend
-45%
Monthly Spend
$22K to $12K
CI/CD Productivity
+300%
Deploys/Week
5 to 20
<5ms
Latency (SGP-PH)
100%
Observability
99.9%
Uptime SLA
<15min
Pipeline Duration

Project

Summary

A client engaged Recursive Loop Inc. for a comprehensive cloud transformation initiative. The client was running a SaaS platform with remaining on-premise servers that needed migration, manual deployments, no monitoring, and growing scalability demands.

Over a 12-month engagement, we migrated remaining on-premise servers to AWS, implemented GitLab CI/CD automation, deployed Grafana and OpenSearch observability, established AWS Direct Connect for low-latency connectivity, and executed a systematic cost optimization strategy that achieved 45% reduction in cloud spend.

Business

Challenges

Manual Deployments & Long Release Cycles

All deployments performed manually via SSH - inconsistent, error-prone, 2-4 hours per release with no staging validation or automated rollback in place.

No Infrastructure Visibility

Zero centralized monitoring. Teams discovered outages from client complaints. No alerting, no dashboards, no log aggregation across any environment.

High Latency Connectivity

AWS Singapore to Philippines experiencing 150ms+ RTT over public internet - creating financial transaction risk and unacceptable latency for time-sensitive workloads.

Over-Provisioned Cloud Spend

Monthly AWS spend at $22,000+ with idle instances, zero Reserved Instance coverage, and no cost anomaly alerts - 45% above optimal spend.

Our

Solutions

Cloud Migration

Migrated remaining on-premise servers to AWS using MGN & DMS with zero data loss and Multi-AZ architecture for high availability.

Learn more

GitLab CI/CD

Replaced manual SSH deployments with fully automated pipelines: build, SAST, Docker, ECR push, and Blue-Green deployment to ECS Fargate.

Learn more

Grafana & OpenSearch

Deployed Prometheus + Grafana for real-time metrics and Amazon OpenSearch for centralized log ingestion with SNS alerting.

Learn more

Cost Optimization

Systematic right-sizing, Graviton migration, On/Off scheduling, and Savings Plans reducing cloud spend by 45%.

Learn more

Technologies

Used

AWS ECS Fargate
Amazon RDS
Direct Connect
GitLab CI/CD
Grafana
Prometheus
OpenSearch
Docker
Terraform CloudFront WAF Route 53 S3 ALB EC2 Auto-Scaling CloudWatch

Customer

Experience

3.8ms
Direct Connect RTT
Down from 150ms+ over public internet
15-20
Deploys Per Week
Up from 3-5 manual deployments
18+
Grafana Dashboards
Full infrastructure visibility
$10K+
Monthly Savings
45% reduction in cloud spend

I would like to take a moment to express my appreciation for the outstanding service we have received with Recursive Loop in IT infrastructure management and maintenance. The team has consistently demonstrated a high level of expertise and dedication to ensuring our systems run smoothly and efficiently.

From proactive monitoring to swift resolution of issues, their attention to detail has significantly minimized downtimes and disruptions. I am particularly impressed by how they communicate with us throughout the process, keeping us informed every step of the way.

Their professionalism and responsiveness have not only instilled confidence in our IT processes but have also allowed us to focus on our core business activities without worrying about IT problems. Overall, their commitment to delivering great service has truly made a difference in our operations, and we are grateful for their support. Thank you for making our IT environment a seamless and reliable part of our business!

Henry Osorio
Henry Osorio
Chief Technology Officer

Ready for Your Cloud Journey?

CONNECT WITH US!

Address

Level 6, Ayala Triangle Gardens Tower 2

Paseo de Roxas, Makati City,

Metro Manila, Philippines, 1226

Contact

63.961.676.6713

Email

service@recursiveloop.ph

Cloud Migration

Migrated remaining on-premise servers to AWS cloud with zero data loss and minimal downtime.

Migration Approach

  • Phased migration of remaining on-premise servers using AWS Application Migration Service (MGN)
  • Database migration via AWS Database Migration Service (DMS)
  • MD5 checksum validation for 100% data integrity
  • Blue-Green deployment strategy for zero-downtime cutover

Architecture

  • Primary Region: AWS Singapore (ap-southeast-1)
  • Disaster Recovery: AWS Hong Kong (ap-east-1)
  • Multi-AZ deployment for high availability
  • Cross-region RDS replication for DR
  • Route 53 health-check failover routing

Services Deployed

ECS Fargate EC2 Auto-Scaling RDS PostgreSQL S3 CloudFront ALB

GitLab CI/CD Automation

Replaced all manual SSH deployments with fully automated GitLab CI/CD pipelines, reducing deployment time from hours to minutes.

Pipeline Stages

  • Build: Application compilation and artifact generation
  • Test: Unit tests and code quality checks
  • SAST: Security scanning with Trivy
  • Docker Build: Multi-arch images (amd64/arm64)
  • ECR Push: Container registry deployment
  • Deploy: Blue-Green deployment to ECS Fargate

Key Improvements

  • Deployment frequency: 3-5/week to 15-20/week
  • Pipeline duration: 3-4 hours to under 15 minutes
  • Automated rollback on failed smoke tests
  • Environment promotion: DEV → UAT → PROD

Technologies

GitLab CI Docker AWS ECR ECS Fargate Trivy SAST Terraform

Grafana & OpenSearch Observability

Comprehensive observability stack providing real-time infrastructure and application metrics with centralized log aggregation.

Metrics Stack

  • Grafana 10.x OSS for visualization
  • Prometheus with 15-second scrape intervals
  • Node Exporter on all EC2 instances
  • AlertManager with SNS integration
  • 18+ custom dashboards deployed

Log Aggregation

  • Amazon OpenSearch for centralized logs
  • CloudWatch Logs subscription filters
  • Ingestion from EC2, ECS, RDS, ALB, Lambda
  • Full-text search and log dashboards
  • SNS alert integration for anomalies

Coverage

  • 100% of services reporting metrics
  • Alerts fire within 2 minutes of threshold breach
  • Logs searchable within 30 seconds of ingestion
Grafana Prometheus OpenSearch AlertManager CloudWatch

Cost Optimization

Systematic 1-year phased approach to cloud cost optimization, achieving 45%+ total annual cost reduction.

Q1: On/Off Scheduling

  • Identified non-production workloads running 24/7
  • Implemented AWS Instance Scheduler
  • Non-prod runs 8hrs/day on weekdays only
  • Result: ~30% overall cost reduction

Q2: Rightsizing

  • AWS Compute Optimizer analysis
  • Downsized overprovisioned instances
  • Eliminated unused Elastic IPs, EBS volumes, NAT Gateways
  • Result: Additional 20% savings

Q3: Graviton Migration

  • Migrated ECS Fargate to arm64 task definitions
  • Multi-arch Docker images (linux/amd64 + linux/arm64)
  • RDS migrated to db.r6g (Graviton)
  • Result: ~40% compute cost reduction

Q4: Commitment Plans

  • 1-year Compute Savings Plans at 70% baseline
  • Reserved Instances for RDS (1-year partial upfront)
  • AWS Budgets + anomaly detection configured
  • Result: Additional 25-30% savings