How to manage cache consistency and avoid cache stampedes during high-concurrency flash sales?

Overview

During flash sale events on e-commerce platforms, thousands of concurrent users query popular item details simultaneously. If a cached key expires at peak traffic, hundreds of requests fall through to the relational database at once—a phenomenon known as the cache stampede (or thundering herd problem).

Strategies for Mitigation

• Cache-Aside with Mutex Locking: Only the first worker that acquires a distributed lock queries the database to repopulate the cache.
• Probabilistic Early Expiration (XFetch): Recomputes the cached value ahead of time based on a probabilistic function.
• Soft Expiration: Serves stale data temporarily while a background worker updates the store.

Mutex Lock Invalidation Pattern in Python

Here is an example using Redis distributed locks to eliminate cache stampedes:

import redis
import time

r = redis.Redis(host='localhost', port=6379)

def get_product_details(product_id):
    cache_key = f"product:{product_id}"
    data = r.get(cache_key)
    if data:
        return data
    
    lock_key = f"lock:{product_id}"
    if r.set(lock_key, "true", nx=True, ex=5):
        try:
            data = fetch_from_database(product_id)
            r.setex(cache_key, 3600, data)
            return data
        finally:
            r.delete(lock_key)
    else:
        time.sleep(0.05)
        return get_product_details(product_id)