What architectural patterns, strategies, or algorithms would you implement to prevent a Cache Stampede and ensure the database remains stable when the cache exp

Imagine you are running a high-traffic web platform. To protect your primary database, you cache a heavy, slow-performing query (e.g., a massive SELECT with multiple JOIN operations) in Redis with a Time-To-Live (TTL) of 10 minutes. Your application experiences a constant load of 10,000 requests per second for this specific data. For 10 minutes, everything runs smoothly because Redis handles the traffic easily. However, the exact second the TTL expires, the cache becomes invalid (Cache Miss). Because the traffic is so dense, thousands of concurrent application threads realize the cache is empty at the exact same moment. They all bypass the cache and simultaneously fire the heavy query at your database to rebuild the cache. The database CPU instantly spikes to 100%, connections pool up, timeouts trigger, and the entire system cascades into a failure.