消息去重与一致性哈希交换器插件
探索两个实用的 RabbitMQ 插件：用于分配负载的一致性哈希交换器，以及确保恰好处理一次的去重策略。
消息去重与一致性哈希交换器插件 是 CoddyKit 上的免费 RabbitMQ Messaging & Async Systems 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 RabbitMQ Messaging & Async Systems 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 RabbitMQ Messaging & Async Systems 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。
Beyond Built-in Exchanges
RabbitMQ plugins extend the broker with new exchange types and behaviors. Two useful ones are the consistent hash exchange and message deduplication plugins.
The Load Distribution Problem
Fanout broadcasts to all queues; direct sends to matched queues. Neither evenly spreads a stream across a set of queues by a key.
The consistent hash exchange solves this.
Enabling the Plugin
Enable the plugin on the broker before use.
rabbitmq-plugins enable rabbitmq_consistent_hash_exchangeDeclaring the Exchange
Declare an exchange of type x-consistent-hash.
channel.exchange_declare(
exchange='shards',
exchange_type='x-consistent-hash'
)Binding With Weights
Bind queues with a numeric weight as the binding key. Higher weight means a larger share of the hash ring.
channel.queue_bind(queue='shard-1', exchange='shards', routing_key='10')
channel.queue_bind(queue='shard-2', exchange='shards', routing_key='10')How Hashing Distributes
The exchange hashes each message's routing key and maps it onto the ring, sending it to one bound queue. The same key always lands on the same queue, giving sticky distribution.
Use Cases for Hashing
- Sharding work by customer or entity id
- Keeping per-key ordering across parallel workers
- Even load spread without manual partitioning
The Duplicate Delivery Problem
RabbitMQ guarantees at-least-once delivery, so retries can cause duplicates. Deduplication ensures the same logical message is processed once.
Deduplication Plugin
The community deduplication plugin adds an exchange that drops messages carrying a recently seen deduplication header within a TTL window.
props = pika.BasicProperties(headers={'x-deduplication-header': 'order-42'})Plugin vs Application Dedup
The plugin is convenient but bounded by its cache window. For strong guarantees, also store processed ids in your application database as a backstop.
Operational Notes
- Plugins must be enabled on every node in a cluster
- Verify version compatibility with your broker
- Test failover behavior before relying on it
Quick Check
Test your plugin knowledge.
Recap
The consistent hash exchange distributes a message stream evenly and stickily across queues by hashing the routing key, while deduplication plugins drop repeated messages within a cache window.
Enable plugins on all cluster nodes and back deduplication with application-level checks for strong guarantees.
常见问题解答
「消息去重与一致性哈希交换器插件」课时是免费的吗？
是的 — 「消息去重与一致性哈希交换器插件」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 RabbitMQ Messaging & Async Systems 课程的其余内容，请升级到 CoddyKit PRO。 RabbitMQ Messaging & Async Systems 课程共包含 4 节课。
「消息去重与一致性哈希交换器插件」这节课中我会学到什么？
探索两个实用的 RabbitMQ 插件：用于分配负载的一致性哈希交换器，以及确保恰好处理一次的去重策略。 你通过在浏览器中直接运行的动手代码来练习 RabbitMQ Messaging & Async Systems，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 RabbitMQ Messaging & Async Systems 需要有经验吗？
无需任何先前经验。CoddyKit 上的 RabbitMQ Messaging & Async Systems 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。
「消息去重与一致性哈希交换器插件」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 RabbitMQ Messaging & Async Systems 课中编写并运行代码吗？
能。每节 RabbitMQ Messaging & Async Systems 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。