Cozy Todo - End-to-End DevOps Automation with AWS & Kubernetes

A production-grade Todo application showcasing complete DevOps automation — including CI/CD pipelines, monitoring, logging, and GitOps-based deployments using AWS EKS and Kubernetes.

Table of Contents

• Project Links
• Technologies Used
• AWS Networking, Security, Resource and Service Architecture
  • Networking Architecture
  • Security Architecture
  • Resource and Service Architecture
• Grafana Monitoring Dashboard
• Kibana Logging with ELK Stack
• Argo CD – App of Apps Pattern Deployment
• Alert Manager & Slack
• GitHub Actions Workflow Architecture

Project Links

Applications & Dockerfiles

• Frontend Repository
• Backend Repository

GitHub Actions Workflows

• Frontend CI Workflow
• Backend CI Workflow

Infrastructure as Code (Terraform)

• IaC Repository

Configuration Management (Ansible)

• Ansible Repository

Kubernetes Manifests (Helm Charts & GitOps)

• Manifests Repository

---

Technologies Used

• Cloud Provider: AWS (EKS, EC2, S3, IAM, VPC, RDS, ECR, Secrets Manager, ELB, Route 53)
• Containerization & Orchestration: Docker, Kubernetes, Helm
• Monitoring & Logging: Prometheus, Grafana, Alertmanager, ELK (Elasticsearch, Logstash, Kibana)
• CI/CD & GitOps: GitHub Actions, Argo CD, Trivy
• IaC & Configuration Management: Terraform, Ansible
• Notifications: Slack Webhooks

---

AWS Networking, Security, Resource and Service Architecture

Networking Architecture

Security Architecture

Resource and Service Architecture

---

Grafana Monitoring Dashboard

This dashboard monitors backend performance and system metrics in real time.

Dashboard Overview

Key Metrics

• Process Uptime: E.g., 18.8 minutes
• Total Requests: E.g., 5020
• Error Rate: E.g., 38.7%
• Average Request Duration: E.g., 11.3 ms
• Requests in Progress: Ongoing requests count

Requests - API Throughput

Line chart showing requests per endpoint (/api/todos, /api/todo/2, etc.)

Errors - Requests by Status Code

Donut chart of response status codes (200, 404, etc.)

Duration - API Latency Percentiles

p50, p75, p90, p95, p99 latency metrics

System Metrics

• CPU Usage: e.g., 35.4%
• Memory Usage: Resident & virtual memory
• Open File Descriptors: Tracks system limits

---

Kibana Logging with ELK Stack

Kibana is used for log aggregation and querying with the ELK stack on AWS EKS.

Key Use Cases

• Live Log Streaming during deployments
• Custom Queries, e.g.:
  • status:500
  • correlation_id: "1f4c08aa-f7b8-429c-a115-a1cb469dbdac"
  • message: message

Screenshots

Figure 1: Filter logs with HTTP status code

Figure 2: Filter logs with correlation ID

---

Argo CD – App of Apps Pattern Deployment

Overview

Uses Argo CD with the App of Apps pattern to manage deployments on AWS EKS.

Parent App: apps-dev

• Project: default
• Path: envs/dev
• Repo: todo-manifests
• Purpose: Manage and deploy multiple child apps

Child Applications

1. todo-cozy

   • Path: charts/application
   • Fullstack app (frontend/backend)
   • Status: ✅ Healthy / Synced

2. logging

   • Path: charts/logging
   • Fluent Bit, Elasticsearch, Kibana
   • Status: 🔄 Progressing

3. monitoring

   • Path: charts/monitoring
   • Grafana & alert rules
   • Status: ✅ Healthy / Synced

Benefits

• Centralized environment management
• Scalable and extensible
• GitOps compatible

Screenshot

---

Alert Manager & Slack

Alertmanager is configured to route critical alerts to a Slack channel for real-time incident response.

Configuration Overview

• Alert Source: Prometheus rules defined in the monitoring Helm chart
• Receiver: Slack Webhook URL

Custom Prometheus Alert Rules

These rules proactively monitor the backend API and are configured under the monitoring chart.

AppDown

• Fires when the app is unreachable for 1 minute.
• Severity: critical

HighCPUUsage

• Triggers when CPU usage exceeds 60% for 2 minutes.
• Severity: warning

HighErrorRate

• Alerts if 60% or more requests result in 4xx/5xx errors in the past 5 minutes.
• Severity: critical

Slack Notification Examples

High CPU usage > 60% alert

Alert happened

Alert resolved

---

GitHub Actions Workflow Architecture

Overview

This diagram illustrates the CI/CD pipeline using GitHub Actions, Docker, AWS ECR, Trivy, and Argo CD.

Diagram

CI/CD Steps

1. Developer pushes code to the Frontend or Backend repositories.
2. GitHub Actions gets triggered:
   • 🔔 Send Slack notification on CI start
   • Run Unit Tests and Linting
   • Tag Git version (semantic versioning)
   • Build Docker image
   • Push image to AWS ECR
   • Run Trivy vulnerability scan on local system and Docker image
   • Print vulnerability reports
   • 🔔 Send Slack notification on CI results
3. Argo CD Image Updater watches the ECR image registry:
   • Detects new image tags
   • Commits updated tag to the GitOps manifest repo
   • Argo CD syncs and deploys to AWS EKS