速率限制与查询深度保护
通过限制客户端的调用频率和查询嵌套深度，保护您的 Spring Boot GraphQL API 免受滥用和拒绝服务攻击。
速率限制与查询深度保护 是 CoddyKit 上的免费 GraphQL APIs with Spring Boot 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 GraphQL APIs with Spring Boot 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 GraphQL APIs with Spring Boot 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。
Why GraphQL Needs Protection
A single GraphQL endpoint accepts arbitrarily complex queries. A malicious or careless client can request deeply nested data or hammer the server, exhausting resources.
Rate limiting and depth protection defend against these attacks.
The Nested Query Threat
Because GraphQL allows cyclic relationships, a client could ask for an author's books, each book's author, that author's books, and so on. This recursion can balloon into an enormous, expensive query.
query {
author {
books { author { books { author { name } } } }
}
}Limiting Query Depth
graphql-java provides MaxQueryDepthInstrumentation, which rejects any query that nests deeper than a set limit before execution begins.
@Bean
public Instrumentation depthLimit() {
return new MaxQueryDepthInstrumentation(10);
}Limiting Field Count
Beyond depth, a broad query can request thousands of fields. MaxQueryComplexityInstrumentation caps the total complexity score of a query.
@Bean
public Instrumentation complexityLimit() {
return new MaxQueryComplexityInstrumentation(200);
}What Is Rate Limiting?
Rate limiting caps how many requests a client may make in a time window. It prevents abuse and ensures fair resource sharing across clients.
The Token Bucket Idea
A common algorithm is the token bucket: each client has a bucket that refills at a steady rate. Every request consumes a token; if the bucket is empty, the request is rejected.
Rate Limiting with Bucket4j
The bucket4j library implements token buckets in Java. Configure a bucket with a refill rate and capacity.
Bandwidth limit = Bandwidth.simple(100, Duration.ofMinutes(1));
Bucket bucket = Bucket.builder().addLimit(limit).build();Enforcing the Limit
Before processing a request, try to consume a token. If none is available, return an error instead of executing the query.
if (!bucket.tryConsume(1)) {
throw new RateLimitException("Too many requests");
}Per-Client Buckets
Track a separate bucket per client, keyed by API key or authenticated user ID, so one heavy client cannot starve everyone else.
Bucket bucket = buckets.computeIfAbsent(userId, k -> newBucket());Combining Defenses
Layer your protections for full coverage:
- Depth limit stops recursive abuse
- Complexity limit stops broad expensive queries
- Rate limit stops request floods
- Timeouts stop slow runaway operations
Best Practices
Tune limits to your real traffic:
- Start strict, then relax based on monitoring
- Return clear errors so clients can back off
- Apply tighter limits to unauthenticated traffic
- Log rejected queries to spot abuse patterns
Quick Check
Test your API protection knowledge.
Recap
You hardened your GraphQL API:
- Depth and complexity instrumentation block expensive queries
- Rate limiting caps request frequency per client
- Token buckets (bucket4j) implement fair limits
- Layer depth, complexity, rate, and timeout defenses
These guards keep your API available and resilient under abuse.
用 AI 导师学习 GraphQL APIs with Spring Boot — 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
「速率限制与查询深度保护」课时是免费的吗？
是的 — 「速率限制与查询深度保护」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 GraphQL APIs with Spring Boot 课程的其余内容，请升级到 CoddyKit PRO。 GraphQL APIs with Spring Boot 课程共包含 4 节课。
「速率限制与查询深度保护」这节课中我会学到什么？
通过限制客户端的调用频率和查询嵌套深度，保护您的 Spring Boot GraphQL API 免受滥用和拒绝服务攻击。 你通过在浏览器中直接运行的动手代码来练习 GraphQL APIs with Spring Boot，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 GraphQL APIs with Spring Boot 需要有经验吗？
无需任何先前经验。CoddyKit 上的 GraphQL APIs with Spring Boot 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。
「速率限制与查询深度保护」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 GraphQL APIs with Spring Boot 课中编写并运行代码吗？
能。每节 GraphQL APIs with Spring Boot 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。