⚡ 4-BO‘LIM: Performance Optimization (Backend, DB, API, Pandas) – savollar va javoblar

1. Django/DRF’da sekin ishlayotgan API’ni qanday optimizatsiya qilganing bor?

1. N+1 query muammosi bormi? → select_related, prefetch_related bilan hal qilinadi.
2. Serializer darajasida – unnecessary nested yoki depth=2 muammosi.
3. QueryCountMiddleware bilan query’lar sonini kuzataman.
4. Pagination – har doim ishlatish kerak.
5. Indexlar – filtering ustunlariga db_index=True yoki Meta: index_together.

2. Django’da select_related va prefetch_related farqi nima?

• select_related – OneToOne / ForeignKey bo‘yicha SQL JOIN qiladi
• prefetch_related – ManyToMany / Reverse FK uchun alohida query yuboradi

Book.objects.select_related("author")       # JOIN bilan
Author.objects.prefetch_related("books")    # alohida query bilan

3. SQLAlchemy’da slow query’ni optimallashtirish?

• joinedload, subqueryload, lazy='selectin' orqali N+1 ni kamaytirish
• Query profiling uchun: echo=True, sqlalchemy.engine.logger
• Index yaratish (Index, unique, nullable=False)

session.query(Post).options(joinedload(Post.comments))

4. DRF API performance: caching qayerda kerak?

1. Low-level: @lru_cache, @cache_page, cache.set()
2. High-level: Redis bilan fragment caching
3. Throttling: DRF’da per-user, per-IP throttles
4. Conditional GET: ETag / Last-Modified bilan ishlash

5. Redis bilan qanday caching qilganing bor?

# Django caching
from django.core.cache import cache
cache.set("key", "value", timeout=60*5)
cache.get("key")

1. View-level cache – @cache_page
2. Low-level – frequently accessed query’lar natijasini saqlash
3. Rate-limiting – Redis + ip:count pattern

6. Pandas’da performance profiling qanday qilinadi?

import pandas as pd
import numpy as np

%timeit df["value"] = df["a"] + df["b"]  # timing
df.memory_usage(deep=True)

1. astype() bilan turlarni siqish
2. category turlar
3. apply() dan qochish → vectorized ops
4. query(), eval() – C backend’da tezroq bajariladi

7. Async bilan performance oshirish misollari?

FastAPI misolida:

@app.get("/users")
async def get_users():
    async with async_session() as session:
        result = await session.execute(select(User))
        return result.scalars().all()

• Async SQLAlchemy yoki Tortoise ORM
• IO-bound task’lar (API, Redis, DB) uchun async kerak

8. DRF’da ListAPIView optimizatsiyasi?

• pagination_class
• filterset_class, search_fields, ordering_fields → DRF’da avtomatik qidiruv
• only() bilan faqat kerakli ustunlarni olish

queryset = MyModel.objects.only("id", "name")

9. PostgreSQL performans: qanday tahlil qilasan?

1. EXPLAIN ANALYZE bilan query execution plan
2. Index qilingan ustunlar: btree, gin, gist
3. Connection pool: pgbouncer yoki conn_max_age sozlamasi

EXPLAIN ANALYZE SELECT * FROM users WHERE email='a@example.com';

10. API ni 1 million request/second ga tayyor qilishda nima qilasan?

1. Horizontal scaling – Kubernetes + HPA
2. Redis caching + async
3. Gunicorn tuning – workers, threads, keep-alive
4. CDN / Edge caching – Cloudflare
5. Database connection pooling
6. Stateless services – scale-friendly