API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway) · 课时

隔离舱与限流:提升恢复能力

使用隔离舱隔离和网关的 RequestRateLimiter 过滤器隔离故障并保护下游服务。

第 4 / 4 课13 个步骤

隔离舱与限流:提升恢复能力 是 CoddyKit 上的免费 API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway) 课时。 这是第 4 节课,共 4 节。 你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。 这是 API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway) 学习路径的一部分,你的进度在网页和 CoddyKit 应用中同步。 API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway) 课程共包含 4 节课。

本课时的部分内容尚未翻译,以英文显示。

Containing the Blast Radius

Circuit breakers stop calls to a broken service, but a flood of traffic to one route can still starve others. Bulkheads and rate limiting keep one busy route from sinking the whole gateway.

The Bulkhead Pattern

Named after a ship's watertight compartments, a bulkhead caps how many concurrent calls a route may have. If one service slows down, only its compartment fills, sparing the rest.

Resilience4j Bulkhead Config

Resilience4j offers a bulkhead that limits concurrent calls per instance.

resilience4j:
  bulkhead:
    instances:
      orders:
        maxConcurrentCalls: 20

Why Rate Limiting Differs

A bulkhead caps concurrency; a rate limiter caps requests over time. Together they protect both fast bursts and sustained load.

The RequestRateLimiter Filter

Spring Cloud Gateway ships a RequestRateLimiter filter backed by a Redis token bucket. It rejects excess requests with HTTP 429.

filters:
  - name: RequestRateLimiter
    args:
      redis-rate-limiter.replenishRate: 10
      redis-rate-limiter.burstCapacity: 20

Replenish Rate and Burst

replenishRate is the steady tokens per second; burstCapacity is the most that can be spent in a spike. Burst should be greater than or equal to replenish.

Choosing a KeyResolver

A KeyResolver decides what to limit by, such as user, API key, or IP. Here we limit per user from a header.

@Bean
KeyResolver userKeyResolver() {
    return exchange -> Mono.just(
        exchange.getRequest().getHeaders()
            .getFirst("X-User-Id"));
}

Redis Backing Store

The limiter needs Redis so counts are shared across gateway instances. Add the reactive Redis starter.

<dependency>
  <groupId>org.springframework.boot</groupId>
  <artifactId>spring-boot-starter-data-redis-reactive</artifactId>
</dependency>

Combining Patterns

Stack resilience filters on a route: circuit breaker first, then rate limiter, so traffic is shaped before reaching a possibly fragile backend.

filters:
  - name: CircuitBreaker
    args:
      name: ordersCB
      fallbackUri: forward:/fallback/orders
  - name: RequestRateLimiter
    args:
      redis-rate-limiter.replenishRate: 10
      redis-rate-limiter.burstCapacity: 20

Graceful Rejection

When the limit is exceeded the client gets 429. You can customize the status, but always return a clear, retryable response so clients can back off.

args:
  redis-rate-limiter.replenishRate: 5
  statusCode: TOO_MANY_REQUESTS

Observing the Limits

Watch metrics for 429 counts and bulkhead rejections. Tune the numbers based on real downstream capacity, not guesses.

Quick Check

What is the core difference between a bulkhead and a rate limiter?

Recap

You added two more resilience tools:

  • Bulkheads cap concurrent calls per route
  • RequestRateLimiter uses a Redis token bucket
  • replenishRate and burstCapacity shape traffic
  • A KeyResolver defines the limiting key

Combined with circuit breakers, retries, and fallbacks, your gateway degrades gracefully under stress.

免费开始

用 AI 导师学习 API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway) — 免费

在浏览器中编写并运行真实代码,获得全天候 AI 导师的即时帮助,并在网页或应用中继续学习。

课程
12
课程
48

常见问题解答

「隔离舱与限流:提升恢复能力」课时是免费的吗?

是的 — 「隔离舱与限流:提升恢复能力」的完整文本可在网页上免费阅读。要进行交互式练习(内置代码编辑器和全天候 AI 导师)并解锁 API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway) 课程的其余内容,请升级到 CoddyKit PRO。 API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway) 课程共包含 4 节课。

「隔离舱与限流:提升恢复能力」这节课中我会学到什么?

使用隔离舱隔离和网关的 RequestRateLimiter 过滤器隔离故障并保护下游服务。 你通过在浏览器中直接运行的动手代码来练习 API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway),全天候 AI 导师会在你学习这节课的过程中回答你的问题。

学习 API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway) 需要有经验吗?

无需任何先前经验。CoddyKit 上的 API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway) 课程适合初学者到高级学习者,你可以从这里开始或从头开始,按照自己的节奏学习。 这是第 4 节课,共 4 节。

「隔离舱与限流:提升恢复能力」课时需要多长时间?

大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动,所以你能稳步进步,并在网页和应用中从离开的地方继续。

我能在这节 API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway) 课中编写并运行代码吗?

能。每节 API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway) 课都包含内置代码编辑器,你可以在浏览器中直接编写并运行真实代码,并获得即时 AI 反馈 — 无需本地设置。

此课程中的所有课时

  1. 使用 Resilience4j 实现熔断器
  2. 重试与超时配置
  3. 错误处理与降级方案
  4. 隔离舱与限流:提升恢复能力
← 返回 API Gateway & Reverse Proxy (Nginx + Spring Cloud Gateway)