Mastering Cache Strategies: Warmup, Avalanche, Penetration, and Breakdown Solutions

1. Cache Warmup

1. Definition
Cache warmup is loading relevant cache data into the cache system before the system goes live, so users can query pre‑warmed data directly without hitting the database.

2. Solutions
1) Manually trigger a cache refresh page during deployment.
2) Load data automatically at application startup when data volume is small.
3) Periodically refresh the cache.

2. Cache Avalanche

Causes
- Redis host failure leading to total crash (hardware issue).
- Massive keys expire simultaneously (software issue).

Prevention & Solutions
- Set keys to never expire or stagger expiration times.
- Use high‑availability Redis clusters (master‑slave + Sentinel, Redis Cluster).
- Enable persistence (AOF/RDB) for quick recovery.
- Combine multiple caches (e.g., Ehcache local + Redis) to mitigate avalanche.
- Apply service degradation (Hystrix or Alibaba Sentinel for rate‑limiting and fallback).

3. Cache Penetration

3.1 What it is

When a request queries a key that is absent in both Redis and MySQL, the request still hits the database repeatedly, causing pressure. This phenomenon is called cache penetration.

3.2 Solutions

Solution 1: Empty object or default value cache

If a key is missing, store a predefined default value (e.g., 0, -1, "defaultNull") in Redis after the first miss. Subsequent requests read the default value from Redis, preventing repeated database hits. This approach does not protect against malicious attacks that query many non‑existent keys.

Solution 2: Guava Bloom Filter

Maintain a Bloom filter containing all legitimate keys. Queries first check the filter; if the key is not present, the request is rejected without hitting Redis or the database. The filter may produce false positives but no false negatives, and keys cannot be removed.

4. Cache Breakdown (Hot Key Miss)

4.1 What it is

When a hot key expires, a surge of requests simultaneously miss the cache and hit the database, overwhelming it.

4.2 Solutions

1) Stagger expiration times or avoid setting expiration for frequently accessed hot keys.
2) Use mutex locking (double‑check locking): the first request that detects a miss acquires a lock, loads data into cache, and releases the lock; other concurrent requests wait and then read from the populated cache.

5. Summary

These cache management techniques—warmup, avalanche prevention, penetration mitigation, and breakdown handling—help maintain system stability and protect backend databases.

Original source: https://blog.51cto.com/u_13236892/8955597