IITM Backend API: Building Scalable Microservices

APIs are the backbone of modern applications. This backend service for IIT Madras demonstrates how to build robust, scalable APIs that can handle thousands of concurrent users.

Architecture

Client → Nginx (Load Balancer)
       → API Gateway (Rate Limiting)
       → Django REST Framework
       → PostgreSQL (Database)
       → Redis (Cache + Sessions)

Authentication & Security

JWT Token Implementation:

from rest_framework_simplejwt.tokens import RefreshToken

def get_tokens_for_user(user):
    refresh = RefreshToken.for_user(user)
    return {
        'refresh': str(refresh),
        'access': str(refresh.access_token),
    }

# In views
class LoginView(APIView):
    def post(self, request):
        username = request.data['username']
        password = request.data['password']
        user = authenticate(username=username, password=password)

        if user:
            tokens = get_tokens_for_user(user)
            return Response(tokens)
        return Response({'error': 'Invalid credentials'}, status=401)

Performance Optimization

Redis Caching Strategy:

from django.core.cache import cache

def get_course_data(course_id):
    cache_key = f'course_{course_id}'
    data = cache.get(cache_key)

    if data is None:
        data = Course.objects.get(id=course_id)
        cache.set(cache_key, data, 3600)  # 1 hour
    return data

Result: 95% cache hit rate, 200ms → 15ms response time.

Rate Limiting

Prevent abuse with throttling:

REST_FRAMEWORK = {
    'DEFAULT_THROTTLE_CLASSES': [
        'rest_framework.throttling.AnonRateThrottle',
        'rest_framework.throttling.UserRateThrottle'
    ],
    'DEFAULT_THROTTLE_RATES': {
        'anon': '100/hour',
        'user': '1000/hour'
    }
}

API Documentation

Auto-generated docs with drf-spectacular:

from drf_spectacular.utils import extend_schema

@extend_schema(
    request=CourseSerializer,
    responses={200: CourseSerializer},
    description="Create a new course"
)
def create_course(request):
    # Implementation
    pass

Swagger UI at /api/docs/ for interactive testing.

Results

• Performance: 500 req/s on single instance
• Uptime: 99.9% over 6 months
• Users: 2000+ concurrent students
• API Calls: 50,000+ daily

Key takeaway: Good APIs are invisible - they just work.