Asynchronous Cache Refresh Mechanism for Interface Optimization Using Redis and RabbitMQ

1. Background and Results

With the rapid development of the Internet and communication devices, users demand faster page and app loading speeds. When third‑party APIs are unstable, performance suffers. In a used‑car business with many external API calls, we devised an asynchronous cache refresh solution using Redis and RabbitMQ, greatly improving system performance, stability, and meeting tp99 latency targets.

2. Solution Iteration

We first accessed APIs directly without caching. As traffic grew, we added a local cache, which introduced inconsistency across servers. Next we moved to a distributed Redis cache, solving consistency but raising data‑freshness concerns for short‑lived data. Finally, we combined Redis with RabbitMQ to create an asynchronous cache refresh mechanism that balances performance and real‑time data needs.

3. Asynchronous Cache Refresh Technical Solution

The architecture consists of user clients, API services, Redis, RabbitMQ, cacheable APIs, data sources, and logging. We built a dedicated microservice for cache handling, provisioned Redis and RabbitMQ capacity, defined a unified cache key rule, and created a separate consumer service to update Redis asynchronously.

Execution flow:

When a client requests an API, the service first checks Redis. If a valid cache exists, it returns the data and, if the cache age exceeds a threshold, pushes a refresh message to RabbitMQ.

If the cache miss occurs, the service calls the downstream API, stores the result in Redis with a timestamp.

A RabbitMQ consumer continuously processes refresh messages, re‑fetches the source API, and overwrites the Redis entry, ensuring data freshness without blocking the client request.

4. Data Support

Performance tests on high‑traffic interfaces show that the asynchronous refresh approach with a 10‑minute Redis TTL achieves similar latency to a 4‑hour static cache, while preserving near‑real‑time data, outperforming a short‑TTL static cache.

5. Summary

The asynchronous cache refresh mechanism dramatically improves interface performance, reduces reliance on unstable third‑party APIs, and satisfies strict tp99 latency requirements while maintaining data freshness, especially for APIs that can only tolerate short cache lifetimes.