How DevOps Is Used in Indian E-commerce Companies Today

import hashlib

FLAGS = {"new_checkout_ui": {"enabled": True, "rollout_pct": 10}}

def is_on(flag: str, user_id: str) -> bool:
    cfg = FLAGS.get(flag, {})
    if not cfg.get("enabled"):
        return False
    bucket = int(hashlib.sha256(f"{flag}:{user_id}".encode()).hexdigest(), 16) % 100
    return bucket < cfg["rollout_pct"]

users = [f"u{i}" for i in range(1000)]
share = sum(is_on("new_checkout_ui", u) for u in users) / len(users)
print(f"users seeing new checkout: {share:.0%}")   # about 10%
# Switch off instantly: FLAGS["new_checkout_ui"]["enabled"] = False

# .github/workflows/storefront.yml
name: storefront
on:
  pull_request:
  push:
    branches: [main]
jobs:
  build-test:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with: { node-version: 20 }
      - run: npm ci
      - run: npm test
      - run: npm run build
      - run: docker build -t storefront:${{ github.sha }} .

// Run: k6 run sale_test.js  (against a test environment only)
import http from 'k6/http';
import { check, sleep } from 'k6';

export const options = {
  stages: [
    { duration: '2m', target: 200 },
    { duration: '5m', target: 1000 },   // simulated sale start
    { duration: '2m', target: 0 },
  ],
  thresholds: { http_req_duration: ['p(95)<800'], http_req_failed: ['rate<0.01'] },
};

export default function () {
  const res = http.get('https://staging.example.com/category/phones');
  check(res, { 'status 200': (r) => r.status === 200 });
  sleep(1);
}

apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: catalog-api
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: catalog-api
  minReplicas: 10        # raised before the sale starts
  maxReplicas: 80
  behavior:
    scaleUp:
      stabilizationWindowSeconds: 0
      policies:
        - type: Percent
          value: 100
          periodSeconds: 30
  metrics:
    - type: Resource
      resource:
        name: cpu
        target: { type: Utilization, averageUtilization: 55 }

groups:
  - name: checkout
    rules:
      - alert: CheckoutHighErrorRate
        expr: |
          sum(rate(http_requests_total{service="checkout",status=~"5.."}[5m]))
          / sum(rate(http_requests_total{service="checkout"}[5m])) > 0.02
        for: 5m
        labels: { severity: page }
        annotations:
          summary: "Checkout 5xx error rate above 2%"
          runbook: "https://wiki.example.com/runbooks/checkout-errors"

def orders_drop_alert(now: int, same_time_last_week: int, threshold: float = 0.5) -> str:
    if same_time_last_week == 0:
        return "no baseline"
    ratio = now / same_time_last_week
    if ratio < threshold:
        return f"ALERT: orders at {ratio:.0%} of last week's level"
    return f"ok ({ratio:.0%} of last week)"

print(orders_drop_alert(now=120, same_time_last_week=400))
print(orders_drop_alert(now=390, same_time_last_week=400))

Project: Mini storefront platform (dummy data)

- catalog-api and checkout-api in containers
- CI pipeline: tests, build, image push
- Feature flag for a new checkout UI (10% rollout)
- Kubernetes deployment with HPA
- k6 load test simulating a sale start
- Prometheus alerts for checkout errors and latency
- Runbook for checkout incidents

Festive sale readiness

[ ] Load test at expected peak plus margin
[ ] Min replicas raised before sale start
[ ] Cache warmed for top categories
[ ] Database connection limits reviewed
[ ] Change freeze agreed
[ ] On-call rota and escalation published
[ ] Rollback tested within the last two weeks
[ ] Dashboards for search, cart, checkout, orders/min

Incident: Checkout errors during evening peak
Impact: elevated payment failures for a short period (test scenario)
Timeline: alert -> on-call ack -> flag switched off -> recovery
Cause: new checkout UI called a slow endpoint
Fix: flag off; endpoint optimised; load test added
Prevention: performance test required before flag rollout