WebFlux and Reactive Support

Redis Cache Java fully supports Spring WebFlux and reactive programming with Project Reactor.

Overview

As of Spring Framework 6.1+, the @Cacheable annotation automatically adapts to reactive return types like Mono<T> and Flux<T>:

The emitted value (not the Mono/Flux itself) is cached
Cache hits return the cached value wrapped in a Mono/Flux
Caching happens after the reactive stream completes
Errors are not cached

Configuration

When using Redis Cache Java with Spring WebFlux, you must enable async mode to avoid blocking operations on reactive threads:

RedisCacheConfiguration config = RedisCacheConfiguration.defaultConfig()
    .async(true)  // Required for WebFlux
    .entryTtl(Duration.ofMinutes(10));

Without async mode, cache write operations will call .block() which throws IllegalStateException when executed on reactive threads (e.g., parallel-1, parallel-2).

Example

Service with Reactive Caching

@Service
public class PersonService {

    @Cacheable("persons")
    public Mono<Person> findPerson(String id) {
        // Reactive data loading
        return webClient.get()
            .uri("/persons/{id}", id)
            .retrieve()
            .bodyToMono(Person.class);
    }
}

Configuration

@Configuration
@EnableCaching
public class CacheConfig {

    @Bean
    public CacheManager cacheManager(RedisClient redisClient) {
        RedisCacheConfiguration config = RedisCacheConfiguration.defaultConfig()
            .async(true)  // Enable async mode for WebFlux
            .entryTtl(Duration.ofMinutes(10))
            .redisType(RedisType.JSON);

        return RedisCacheManager.builder(redisClient)
            .defaults(config)
            .build();
    }
}

Additional Resources

For complete working examples, see the WebFlux integration tests.

See Also