🌩️ Ghaymah SRE Exam — Amir Kasseb

Site Reliability Engineering technical exam submission on the Ghaymah Cloud platform.

Candidate: Amir Mohamed Kasseb

Profile: qabilah.com/profile/amir-m-kasseb

---

🚀 Live Deployments

Service	URL	Description
🖥️ API App	ghaymah-app-19cc5d65231a.hosted.ghaymah.systems	Go REST API — /health, /metrics endpoints
📊 SRE Dashboard	ghaymah-dashboard-v2-85e116a29519.hosted.ghaymah.systems	Live monitoring dashboard for the API
🔭 Mithal Monitor	ghaymah-mithal-monitor-9e02b47e8af8.hosted.ghaymah.systems	Uptime & latency monitor for mithal.space

---

📁 Repository Structure

ghaymah-exam-amirkasseb-sre/
│
├── q1-deploy-monitor/        # Q1: Container deployment & SRE monitoring
│   ├── app/                  # Go REST API (main.go, Dockerfile)
│   ├── dashboard/            # SRE monitoring dashboard (HTML/CSS/JS + Nginx)
│   ├── health-check.sh       # Bash health monitoring daemon (30s interval)
│   ├── monitor.log           # Health check output log
│   └── README.md
│
├── q2-postmortem/            # Q2: Incident postmortem report
│   └── postmortem-report.md  # OOMKilled outage analysis & recommendations
│
├── q3-cicd/                  # Q3: CI/CD Pipeline
│   ├── workflow.yml           # GitHub Actions workflow definition
│   └── readme.md             # Pipeline design explanation (Arabic)
│
├── q4-scalability/           # Q4: Scalability & load balancing
│   ├── calculations.md       # Container count calculations & cold-start strategy
│   └── architecture.png      # Scalability architecture diagram
│
├── q5-mithal-monitor/        # Q5: mithal.space full-stack monitor
│   ├── monitor.py            # Python monitoring script (uptime, latency, SSL, DNS)
│   ├── index.html            # Interactive monitoring dashboard
│   ├── data.jsonl / data.csv # Collected monitoring data
│   ├── Dockerfile            # Container definition
│   └── README.md
│
├── common-mortakaz/          # Mortakaz integration proposals
│   ├── integration-1.md      # Qumra × Ghaymah integration proposal
│   └── integration-2.md      # Second integration proposal
│
└── common-qabilah/           # Candidate profile
    └── qabilah-profile.txt   # Link to professional profile

---

📋 Questions Summary

Q1 — Container Deployment & SRE Monitoring

Deployed a Go REST API to Ghaymah Cloud with:

• Multi-stage Dockerfile (Alpine Linux, non-root user, port 8080)
• 3 endpoints: GET / · GET /health · GET /metrics
• Bash monitoring daemon (health-check.sh) — probes every 30 seconds, logs HTTP status & response time
• Live SRE dashboard — real-time latency chart, uptime timeline, request count, auto-refresh every 15s

# Verify live endpoints
curl https://gheyma-app-2b88268529f9.hosted.ghaymah.systems/health
curl https://gheyma-app-2b88268529f9.hosted.ghaymah.systems/metrics

📂 q1-deploy-monitor/ · 📊 Live Dashboard

---

Q2 — Incident Postmortem

Authored a full postmortem for a 45-minute OOMKilled outage:

• Root cause: Container memory limit set below actual working-set (in-memory cache was never accounted for)
• Timeline: Pod crash loop from 10:53 → resolved at 11:28 after limit increase
• Recommendations: Memory-based HPA, 75–80% threshold alerts, restart-count circuit breaker, load testing after stateful changes
• Ghaymah-specific: Detection flow using Ghaymah Usage Dashboard & application logs

📂 q2-postmortem/postmortem-report.md

---

Q3 — CI/CD Pipeline

Designed a GitHub Actions CI/CD pipeline deploying to Ghaymah Cloud:

develop → Build & Test → Deploy to Staging
main    → Build & Test → Manual Approval → Deploy to Production

• Build: Docker image built & health-tested on every push
• Staging: Auto-deployed from develop branch
• Production: Requires manual approval via GitHub Environments before deploy
• Ghaymah CLI: Used to authenticate (gy auth login) and deploy (gy resource app launch)
• Secrets: GHAYMAH_EMAIL and GHAYMAH_PW stored in GitHub Secrets

📂 q3-cicd/workflow.yml

---

Q4 — Scalability & Load Balancing

Calculated infrastructure requirements for 15,000 req/s with a 30% safety margin:

Effective load = 15,000 × 1.3 = 19,500 req/s
Containers needed = 19,500 ÷ 500 = 39 containers
→ ~77% capacity utilization per container

Cold-start strategies covered:

Strategy	Description
Predictive Scaling	Spin up containers before threshold is reached
Warm Pool	2–3 standby containers always ready
Golden Images	Pre-baked optimized images to cut init time
Gradual Traffic Ramp-up	Route load gradually to new containers
Readiness Probe	Don't add to load balancer until healthy

Also covered Ghaymah Block Storage for stateful workloads (persistence, snapshots, low-latency I/O).

📂 q4-scalability/calculations.md

---

Q5 — Mithal.space Monitor & Dashboard

Built a full-stack monitoring application for mithal.space:

• Python monitoring script (monitor.py) collecting: Uptime, Latency, SSL certificate expiry, DNS resolution time, Search latency
• Interactive dashboard (index.html) with real-time charts (Chart.js)
• Data persistence in data.jsonl and data.csv
• Dockerized: Single-container deployment with volume mounts for data persistence

# Run with Docker Compose
docker compose up -d
# → Available at http://localhost:8080

# Or directly
pip install -r requirements.txt
python monitor.py

📂 q5-mithal-monitor/ · 🔭 Live Monitor

---

Common — Mortakaz Integration Proposals

Two integration proposals connecting Arabic SaaS products with Ghaymah Cloud:

• Integration 1: Qumra × Ghaymah — Store/website builder using Ghaymah as cloud infrastructure layer, monitored via mithal.space
• Integration 2: integration-2.md — Second integration proposal

---

🛠️ Tech Stack

Layer	Technology
API	Go (Golang), Alpine Linux
Containerization	Docker (multi-stage builds)
CI/CD	GitHub Actions + Ghaymah CLI
Monitoring Script	Bash, Python (requests, ssl, dns)
Dashboard	HTML5, Vanilla CSS, Vanilla JavaScript, Chart.js
Cloud Platform	Ghaymah Cloud

---

✅ Verification

# API health check
curl -i https://gheyma-app-2b88268529f9.hosted.ghaymah.systems/health
# Expected: HTTP/1.1 200 OK  {"status":"healthy"}

# API metrics
curl -i https://gheyma-app-2b88268529f9.hosted.ghaymah.systems/metrics
# Expected: HTTP/1.1 200 OK  {"requests":<count>}

---

Made with ☁️ on Ghaymah Cloud