แนวคิดนามธรรมของแคช Spring

Caching stores the result of expensive operations so repeated calls return instantly from memory instead of recomputing or re-querying.

• Lower latency
• Less database load
• Better throughput

Spring provides a cache abstraction — a uniform API over many backends (in-memory, Redis, Caffeine). You annotate methods; Spring handles the storage.

Add @EnableCaching to a configuration class to turn on annotation processing.

A CacheManager owns the named caches. Spring Boot auto-configures one based on what is on the classpath.

With no cache library on the classpath, Spring Boot uses a simple ConcurrentMapCacheManager backed by a map.

Each cached method targets a named cache. Names let you tune and clear caches independently.

On call, Spring builds a key, checks the cache: if present (a hit), it returns the stored value and skips the method body. On a miss, it runs the method and stores the result.

By default the key is derived from the method arguments. You can override it with a SpEL expression.

Add a library to switch backends automatically: Caffeine for fast local caching, Redis for a shared distributed cache.

Caffeine supports size limits and expiry via properties.

Caching suits read-heavy, slow-changing data. Avoid it for rapidly changing values or when stale reads are unacceptable without proper eviction.

You learned the cache abstraction:

• @EnableCaching turns the feature on
• A CacheManager owns named caches
• Default backend is an in-memory ConcurrentMap
• Swap to Caffeine or Redis by adding the library