Snippet

Tuning Caffeine In-Memory Cache Eviction Within Spring CacheManager

Spring's Cache abstraction integrates with high-performance Caffeine caching to manage memory utilization under high-throughput workloads. Configuring maximumSize activates W-TinyLFU eviction algorithms, bounding memory growth and preventing heap exhaustion from continuous key generation, while expireAfterWrite ensures time-based data freshness.

snippet.java

java

@Configuration
@EnableCaching
public class CacheConfig {
 
    @Bean
    public CacheManager cacheManager() {
        CaffeineCacheManager cacheManager = new CaffeineCacheManager("productMetrics", "fxRates");
        cacheManager.setCaffeine(Caffeine.newBuilder()
            .initialCapacity(100)
            .maximumSize(5_000)
            .expireAfterWrite(15, TimeUnit.MINUTES)
            .recordStats());
        return cacheManager;
    }
}
 
@Service
public class ProductService {
 
    @Cacheable(value = "productMetrics", key = "#sku", unless = "#result == null")
    public ProductMetrics fetchMetrics(String sku) {
        return heavyComputationFromExternalApi(sku);
    }
}

spring

Breakdown

cacheManager.setCaffeine(Caffeine.newBuilder()

Defines the Caffeine builder specifications applied to all caches managed by this CacheManager.

.maximumSize(5_000)

Imposes a hard threshold on cached item count, triggering automatic entry eviction when full.

@Cacheable(value = "productMetrics", key = "#sku", unless = "#result == null")

Interprets cache hit/miss semantics and prevents storing null values in memory via conditional SpEL filtering.

Previous snippet Next snippet

More Intermediate Java exercises →

Tuning Caffeine In-Memory Cache Eviction Within Spring CacheManager

Related