AWS Operations · Cloud Architect · 2025–2026 · Active

Mobilife AWS High-Availability Architecture

Designed a production-grade AWS architecture and operational runbook for Mobilife — scalable, observable, and safe to deploy.

AWS Route 53, ACM, ALB, EC2, Auto Scaling, Launch Templates, ECR, SSM, Secrets Manager, Docker, Prometheus, Grafana, Node Exporter, Blackbox Exporter, CloudWatch, SNS, RDS, S3

mobilife.mn
Mobilife — production platform on AWS
30%
lower change failure rate
Multi-AZ
high availability
Zero-downtime
rollouts

Architecture

Multi-AZ EC2 platform behind a load balancer, with the runbook that operates it

Route 53
DNS + health checks
CloudFront
CDN — edge caching
S3 + ACM
assets · TLS certificates
VPC
Availability Zone AAvailability Zone B
Public subnet
Application Load Balancer
TLS termination · target group routing
Private subnet — application
Auto Scaling group — CPU target tracking
EC2 — Docker
Launch Template bootstrap
EC2 — Docker
Launch Template bootstrap
Private subnet — cache
Redis
in-memory cache — cuts RDS load
Private subnet — data
RDS — MySQL
Multi-AZ
RDS — MySQL
standby
Monitoring & alerting
Prometheus
metrics collection
Grafana
dashboards
Node Exporter
host metrics
Blackbox Exporter
endpoint probes
CloudWatch
ASG lifecycle · RDS CPU
SNS
alert delivery
Config, secrets & delivery
ECR
versioned container images
Secrets Manager
runtime secrets
SSM
parameters · secure access
Recovery & rollback
  • RDS standby promoted automatically when the primary fails
  • Versioned images in ECR — roll back by redeploying the previous tag
  • Launch Templates rebuild any instance from scratch, repeatably
  • ASG health checks replace failed instances without intervention
  • Route 53 and Blackbox probes catch degradation before users do
  • Documented deployment, troubleshooting and rollback runbook

This work formalized a new AWS EC2-based production architecture for Mobilife and turned it into a practical handover document that infrastructure and operations teams can run with confidence.

The setup uses Route 53, ACM, an Application Load Balancer, Target Groups, an Auto Scaling Group, and EC2 Launch Templates — replacing a fragile single-instance model with a scalable, high-availability deployment pattern.

The runbook emphasizes professional operating practice: cache-aware design, right-sizing to reduce cost, secure secret delivery, repeatable Docker deployments, health checks, rollback procedures, and layered monitoring across infrastructure and application.

Highlights

  • Route 53 → ALB → Target Group → Auto Scaling Group → EC2 architecture with TLS termination and API health checks.
  • Launch Template bootstrap: Docker startup, ECR login, Secrets Manager loading, container replacement, repeatable provisioning.
  • Caching strategy to cut database load, improve response times, and support efficient scaling under peak traffic.
  • Cost optimization via Auto Scaling, instance right-sizing, and reducing idle capacity.
  • Observability with Prometheus, Grafana, and Node Exporter; production health checks via Route 53 + Blackbox Exporter.
  • CloudWatch + SNS alerting for ASG lifecycle events, service degradation, and RDS CPU thresholds.
  • Documented deployment, troubleshooting, and rollback steps — reducing change failure rate by 30%.