Building a Two-Level Cache Framework to Solve Spring-Cache Problems

1. Introduction

The article discusses the limitations of Spring-Cache and presents a two-level cache framework to overcome them, focusing on local and remote caching strategies.

2. Framework Architecture

The framework uses annotations for configuration, with a CacheManagerContainer managing cache implementations. It supports local (Caffeine) and remote (Redis) caches, and includes features like automatic refresh and distributed locks.

3. Key Features

The framework offers TTL, refresh time, cache type selection, and SpEL-based key generation. It also includes a two-level cache approach to reduce remote cache load and improve performance.

auto:

cache:

local:

type: caffeine

remote:

type: redis

host: localhost

port: 6379

timeout: 2000

pool:

min-idle: 2

max-idle: 10

max-active: 20

max-wait: 200

key-serializer: org.springframework.data.redis.serializer.StringRedisSerializer

value-serializer: org.springframework.data.redis.serializer.GenericJackson2JsonRedisSerializer

namespace: testCache

allow-null-values: true

4. Conclusion

The framework enhances caching efficiency and addresses Spring-Cache limitations, with future plans for additional features like distributed locks and monitoring.