🚀 safeDeploy demo application

A minimal TypeScript + Express service used to demonstrate SafeDeploy patterns such as health checks, controlled failures, synthetic latency, and Prometheus metrics when running on kubernetes safedeploy environment.

✨ Features

• ✅ GET /health returns a healthy response for readiness/liveness checks.
• 🐢 GET /slow simulates a slow dependency by waiting ~3 seconds.
• 💥 GET /error intentionally returns 500 for failure-path testing.
• ➕ GET /sum increments and returns an in-memory counter.
• 📈 GET /metrics exposes Prometheus metrics (default process metrics + custom HTTP metrics).

🧰 Tech stack

• Node.js
• TypeScript
• Express
• Vitest + Supertest
• prom-client

📋 Prerequisites

• Node.js 20+ (recommended)
• npm

⚡ Getting started

npm install
npm run build
npm start

The app starts on http://0.0.0.0:3000.

🛠️ Development

npm run dev

🧪 Testing

npm run test

☸️ Kubernetes runtime notes

This service is intended to run in a Kubernetes environment as a Pod (for example, as a canary and/or stable workload behind a Service).

When deployed to Kubernetes, the endpoints in this README can be used for:

• health verification (/health)
• failure-path validation (/error)
• synthetic latency checks (/slow)
• metrics scraping (/metrics)

🔁 CI/CD pipeline notes

This repository is designed to fit into a CI/CD workflow where pipelines typically:

1. install dependencies
2. run tests
3. build the TypeScript bundle
4. build and push a container image to GHCR (GitHub Container Registry)
5. publish/deploy via a GitOps flow to Kubernetes

A common command sequence in CI is:

npm ci
npm test -- --run
npm run build

In a GitOps setup, deployment state is tracked in Git, and the cluster continuously reconciles to the desired manifests that reference the image stored in GHCR.

🐌 Manual slow deployment test (current workaround)

ℹ️ SafeDeploy is still under active development, so slow rollout behavior is tested manually for now.

Start a temporary curl pod in Kubernetes:

kubectl run curltest \
  --rm -it \
  --image=curlimages/curl \
  --restart=Never -- sh

Then issue repeated requests to the canary app's slow endpoint:

for i in {1..20}; do
  curl http://demo-app-canary:3000/slow
done

🔎 API quick reference

✅ GET /health

{ "status": "ok - Application reached" }

🐢 GET /slow

{ "status": "slow", "delayMs": 3000 }

💥 GET /error

{ "status": "error", "message": "Intentional error for SafeDeploy demo" }

➕ GET /sum

{ "status": "sum", "sum": 1 }

📈 GET /metrics

Returns Prometheus-formatted metrics, including:

• http_request_duration_seconds
• http_requests_total{status="..."}
• default Node/process metrics