Smoke Testing

Smoke testing is a critical validation step that confirms an application's core functionalities are working after a build, deployment, or infrastructure change. It's quick, targeted, and aims to catch major issues early that would make the application "unusable."

What is a Smoke Test?

A smoke test focuses on:

• Verifying availability of core endpoints
• Ensuring the authentication mechanism works
• Receiving successful responses from backend services
• Connecting to the database or queues (if critical to the app)

It does not replace unit, integration, or end-to-end tests. Instead, it acts as a "gatekeeper" before deeper validation.

Best Practices

1. Runs first after every deployment or restart
2. Executes in under 1 minute — lightweight, fast validation
3. No dependency on specific data — works without test fixtures
4. Clear failure logs — each failure must be clearly visible with status and endpoint
5. Reusable and configurable — e.g., via environment variables

Tools and Techniques

Tool	Description
curl, httpie	Simple HTTP checks from CLI
Python + requests	Custom smoke logic in Python scripts
Azure DevOps, GitHub Actions	Integrate smoke stage into CI/CD pipeline
Retries	Use exponential backoff for services with startup delays

---

Examples of Smoke Test Implementations

Bash script using curl

#!/bin/bash
set -e
BASE_URL="${1:-https://myapp.example.com}"
AUTH_HEADER="Authorization: Bearer ${SMOKE_TEST_TOKEN:-test-token}"

check() {
  local endpoint=$1
  echo -n "Checking $endpoint... "
  status=$(curl -s -o /dev/null -w "%{http_code}" -H "$AUTH_HEADER" "$BASE_URL$endpoint")
  if [ "$status" -eq 200 ]; then
    echo "✅ OK"
  else
    echo "❌ Failed with status $status"
    exit 1
  fi
}

check "/"
check "/api/health"
check "/api/profile"
echo "🎉 Smoke Test Passed"

Python script using requests

import requests, sys, os

BASE_URL = sys.argv[1] if len(sys.argv) > 1 else "https://myapp.example.com"
TOKEN = os.environ.get("SMOKE_TEST_TOKEN", "test-token")

endpoints = ["/", "/api/health", "/api/profile"]
headers = {"Authorization": f"Bearer {TOKEN}"}

for endpoint in endpoints:
    print(f"Checking {endpoint}...", end=" ")
    try:
        r = requests.get(f"{BASE_URL}{endpoint}", headers=headers)
        if r.status_code == 200:
            print("✅ OK")
        else:
            print(f"❌ Failed ({r.status_code})")
            sys.exit(1)
    except Exception as e:
        print(f"❌ Error: {e}")
        sys.exit(1)

print("🎉 All smoke tests passed.")

---

Optional Features

Retry Logic

Add retry mechanisms (e.g., until in Bash, time.sleep() & for attempt in range(n) in Python) to allow the app time to start properly.

Basic Auth / Token auth

Use headers:

-H "Authorization: Bearer $TOKEN"

or:

-u "user:password"

Dry Run Mode

Add a --dry-run flag to print endpoints without making requests. Useful for environment debugging.

Pipeline Integration (e.g., Azure DevOps)

- script: |
    chmod +x smoke_test.sh
    ./smoke_test.sh https://myapp.example.com
  env:
    SMOKE_TEST_TOKEN: $(mySmokeToken)
  displayName: "Run Smoke Test"

or:

- script: |
    python3 smoke_test.py https://myapp.example.com
  env:
    SMOKE_TEST_TOKEN: $(mySmokeToken)
  displayName: "Run Python Smoke Test"

---

What a Smoke Test Is Not

• It doesn’t perform logic validation (business rules)
• It doesn’t check performance or concurrency
• It doesn’t require specific data (fixtures)
• It doesn’t replace E2E or integration tests

Action Flow

• If it passes, proceed to the next stage.
• If it fails, rollback or investigate

Avoid hardcoding sensitive information in scripts — use environment variables or secret injection mechanisms (e.g., Azure Key Vault references).

Smoke testing should be lightweight enough to always run, but critical enough to stop everything when it fails.

---

We'd love your feedback

Was this helpful?

Yes

No