競合コンシューマーパターン
競合コンシューマーパターンを実装し、1つのキューから複数のコンシューマーでメッセージを処理できるようにします。コンシューマーを追加して処理能力をスケールさせます。
「競合コンシューマーパターン」はCoddyKit上の無料RabbitMQ Messaging & Async Systemsレッスンです。 これはレッスン1/4です。 下記で完全なレッスンを無料で読むことができます。その後、ブラウザ内の組み込みコードエディタと24時間対応のAIチューターでハンズオン演習できます。 これはRabbitMQ Messaging & Async Systems学習パスの一部であり、ウェブとCoddyKitアプリ全体で進捗が同期されます。 RabbitMQ Messaging & Async Systemsコースには全4レッスンが含まれています。
このレッスンの一部はまだ翻訳されておらず、英語で表示されています。
Scaling with Competing Consumers
Welcome! In distributed systems, you often need to process many tasks efficiently. The Competing Consumers pattern is a powerful way to achieve this.
It allows you to scale your message processing capacity by simply adding more workers.
How Competing Consumers Work
Imagine a single queue of tasks. Instead of one worker taking all tasks, multiple workers (consumers) listen to this same queue.
- Each message is delivered to only one of the competing consumers.
- Consumers "compete" to receive the next available message.
- This distributes the workload automatically.
Key Benefits of the Pattern
The Competing Consumers pattern offers several advantages:
- Scalability: Easily increase processing power by adding more consumer instances.
- Reliability: If one consumer fails, others can pick up its share of messages.
- Load Balancing: Messages are spread across available consumers, balancing the workload.
- Decoupling: Producers don't need to know how many consumers there are or where they are.
Producer: Sending Tasks
Let's set up a basic producer that sends messages (tasks) to a queue named task_queue. Each message will be a simple string.
Run this code to send a few messages:
import com.rabbitmq.client.Channel;
import com.rabbitmq.client.Connection;
import com.rabbitmq.client.ConnectionFactory;
public class TaskProducer {
private final static String QUEUE_NAME = "task_queue";
public static void main(String[] argv) throws Exception {
ConnectionFactory factory = new ConnectionFactory();
factory.setHost("localhost"); // Assuming RabbitMQ is local
try (Connection connection = factory.newConnection();
Channel channel = connection.createChannel()) {
// Declare a durable queue
channel.queueDeclare(QUEUE_NAME, true, false, false, null);
for (int i = 0; i < 10; i++) {
String message = "Task " + (i + 1);
channel.basicPublish("", QUEUE_NAME, null, message.getBytes("UTF-8"));
System.out.println(" [x] Sent '" + message + "'");
Thread.sleep(100); // Small delay to visualize
}
}
}
}Consumer 1: Processing Tasks
Now, let's create our first consumer. It will connect to task_queue and start processing messages. Each message will be acknowledged after simulating work.
Run this consumer in a terminal:
import com.rabbitmq.client.Channel;
import com.rabbitmq.client.Connection;
import com.rabbitmq.client.ConnectionFactory;
import com.rabbitmq.client.DeliverCallback;
public class TaskConsumer {
private final static String QUEUE_NAME = "task_queue";
public static void main(String[] argv) throws Exception {
ConnectionFactory factory = new ConnectionFactory();
factory.setHost("localhost");
Connection connection = factory.newConnection();
Channel channel = connection.createChannel();
channel.queueDeclare(QUEUE_NAME, true, false, false, null);
System.out.println(" [*] Consumer 1 waiting for messages.");
// Basic QoS: Prefetch 1 message at a time to ensure fair dispatch
channel.basicQos(1);
DeliverCallback deliverCallback = (consumerTag, delivery) -> {
String message = new String(delivery.getBody(), "UTF-8");
System.out.println(" [C1] Received '" + message + "'");
try {
Thread.sleep(1000); // Simulate work
} finally {
System.out.println(" [C1] Done '" + message + "'");
channel.basicAck(delivery.getEnvelope().getDeliveryTag(), false);
}
};
channel.basicConsume(QUEUE_NAME, false, deliverCallback, consumerTag -> {});
}
}Running Multiple Consumers
To truly see the competing consumers pattern in action, open a new terminal window and run the exact same TaskConsumer code again.
You'll now have two consumer instances listening to the task_queue. Run the producer code (from Scene 4) once more. Observe how messages are now distributed between both consumer instances, demonstrating how they compete for messages and share the workload!
RabbitMQ's Dispatching Logic
By default, RabbitMQ uses a round-robin dispatching mechanism when multiple consumers are connected to the same queue. This means messages are sent to consumers sequentially.
- Consumer 1 gets message 1.
- Consumer 2 gets message 2.
- Consumer 1 gets message 3, and so on.
The basicQos(1) setting in our consumer code is crucial here. It tells RabbitMQ not to send more than one unacknowledged message to a consumer at a time, ensuring fair dispatch even if consumers process at different speeds.
Common Use Cases
The Competing Consumers pattern is ideal for scenarios like:
- Image processing: Multiple workers resizing images from a queue.
- Email sending: Sending bulk emails without overwhelming a single service.
- Log processing: Analyzing large volumes of logs in parallel.
- Background jobs: Any task that can be processed independently by multiple workers.
Important Considerations
When using competing consumers, keep these in mind:
- Message Ordering: If strict message order is critical, this pattern might not be suitable directly, as messages can be processed out of order by different consumers.
- Idempotency: Consumers should ideally be idempotent. This means processing the same message multiple times should have the same effect as processing it once. This is vital for fault tolerance and retries.
Competing Consumers Quiz
Test your understanding of the Competing Consumers pattern.
Recap: Competing Consumers
Great job! You've learned about the Competing Consumers pattern:
- It enables multiple consumers to process messages from a single queue.
- It's excellent for scaling and load balancing message processing.
- RabbitMQ's default round-robin dispatch and QoS settings facilitate fair distribution.
- Consider idempotency and potential out-of-order processing for specific use cases.
Next, we'll dive deeper into optimizing consumer efficiency with prefetch counts!
よくある質問
「競合コンシューマーパターン」レッスンは無料ですか？
はい。「競合コンシューマーパターン」の完全なテキストはこのウェブで無料で読めます。インタラクティブに演習し（組み込みコードエディタと24時間対応のAIチューター）、RabbitMQ Messaging & Async Systemsコースの残りをアンロックするには、CoddyKit PROにアップグレードしてください。 RabbitMQ Messaging & Async Systemsコースには全4レッスンが含まれています。
「競合コンシューマーパターン」で何を学びますか？
競合コンシューマーパターンを実装し、1つのキューから複数のコンシューマーでメッセージを処理できるようにします。コンシューマーを追加して処理能力をスケールさせます。 ブラウザで直接実行するハンズオンコードでRabbitMQ Messaging & Async Systemsを演習し、24時間対応のAIチューターがレッスンを進める中での質問に答えます。
RabbitMQ Messaging & Async Systemsを始めるのに経験は必要ですか？
事前経験は必要ありません。CoddyKitのRabbitMQ Messaging & Async Systemsは初級者から上級者向けに構成されているため、ここから始めるか最初から始めて、自分のペースで進むことができます。 これはレッスン1/4です。
「競合コンシューマーパターン」レッスンにはどのくらい時間がかかりますか？
ほとんどのCoddyKitレッスンは約5～10分かかります。各レッスンはコンパクトでインタラクティブなので、着実に進歩し、ウェブとアプリ全体で正確に前回の場所から再開できます。
このRabbitMQ Messaging & Async Systemsレッスンでコードを書いて実行できますか？
はい。すべてのRabbitMQ Messaging & Async Systemsレッスンに組み込みコードエディタが含まれているため、ブラウザでリアルコードを書いて実行し、即座のAIフィードバックを取得できます。ローカル設定は不要です。