Redis in Java Backend

Direct Answer

Redis is an in-memory data store with structures such as strings, hashes, sets, sorted sets, and streams. A Java backend may use it for caching, sessions, counters, rate limits, or messaging. It is built for low latency, but no fixed response time is guaranteed; network, command complexity, data size, persistence, and deployment all matter.

Code Example

@Service
public class ProductService {
    private final ProductRepository products;
    private final RedisTemplate<String, Product> redis;

    public Product getProduct(long productId) {
        String key = "product:" + productId;
        Product cached = redis.opsForValue().get(key);
        if (cached != null) {
            return cached;
        }

        Product product = products.findById(productId).orElseThrow();
        redis.opsForValue().set(key, product, Duration.ofMinutes(5));
        return product;
    }
}

This is cache-aside, not a latency guarantee. A production design also handles invalidation, concurrent misses, serialization, timeouts, and Redis failure. For a fixed-window rate limit, set the count and expiry atomically—commonly with a Lua script—instead of separate INCR and EXPIRE calls.

Common Mistakes & Pitfalls

• Caching without an invalidation rule: A TTL limits staleness but does not make cached data correct immediately after a write.
• Splitting rate-limit steps: INCR followed later by EXPIRE can leave a key without expiry if the client fails between commands; use an atomic script or documented pattern.
• Assuming Redis is only a disposable cache: Decide from persistence, replication, backup, recovery, and tolerated data loss rather than from a slogan.