RabbitMQ Messaging & Async Systems · レッスン

メッセージフローのデバッグ

Exchangeやキューを通過するメッセージフローをデバッグするためのツールと技法を活用します。メッセージを追跡して経路を把握し、配信上の問題箇所を特定します。

「メッセージフローのデバッグ」はCoddyKit上の無料RabbitMQ Messaging & Async Systemsレッスンです。これはレッスン2/4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。

Why Debug Message Flow?

When building systems with message queues like RabbitMQ, messages don't always go where you expect. They might get lost, not delivered, or pile up in queues.

Understanding message flow debugging is crucial. It helps you trace a message's journey from producer to consumer, pinpointing exactly where issues occur.

Your Debugging Dashboard: Mgmt Plugin

The RabbitMQ Management Plugin is your primary tool for debugging message flow. It offers a web-based UI to inspect your broker's state.

Overview: High-level stats.
Connections/Channels: See active client connections.
Exchanges: View exchange types, bindings.
Queues: Inspect message counts, consumers, and even get/publish messages.

Inspect & Inject Messages in UI

The management plugin lets you directly interact with your message flow:

Publish Messages: On an exchange's page, use the 'Publish message' panel to send test messages. This helps verify routing key and binding logic.
Get Messages: On a queue's page, use the 'Get messages' panel to pull messages from the queue. This confirms if messages are arriving and what their content/properties are.

Message's ID Card: Properties

Every message in RabbitMQ carries important information. When debugging, pay close attention to:

Routing Key: The key used by exchanges to route the message.
Headers: Custom key-value pairs that can be used for routing (Headers Exchange) or metadata.
Delivery Mode: Indicates if the message is persistent.

These properties determine how a message is handled and routed.

Routing Key Mismatches

A common issue is a message not reaching its intended queue due to an incorrect routing key or a missing binding.

For example, a Direct exchange expecting routing key 'errors' will drop messages sent with 'info' if no queue is bound to 'info'. Always verify producer's routing key matches queue bindings.

Producer-Side Insights with Logging

Adding logging to your producer application is vital. It confirms if your application successfully tried to send a message and what routing key it used.

Try running this Java example. Observe the console output.

import com.rabbitmq.client.Channel;
import com.rabbitmq.client.Connection;
import com.rabbitmq.client.ConnectionFactory;

public class DebugProducer {
    private final static String QUEUE_NAME = "debug_queue_log";

    public static void main(String[] argv) throws Exception {
        ConnectionFactory factory = new ConnectionFactory();
        factory.setHost("localhost"); // Assumes local RabbitMQ

        try (Connection connection = factory.newConnection();
             Channel channel = connection.createChannel()) {

            channel.queueDeclare(QUEUE_NAME, false, false, false, null);

            String message = "Hello, debug world!";
            String routingKey = QUEUE_NAME; // Using queue name as routing key
            
            System.out.println(" [Producer] Sending message to queue: " + QUEUE_NAME);
            System.out.println(" [Producer] With routing key: '" + routingKey + "'");
            
            channel.basicPublish("", routingKey, null, message.getBytes("UTF-8"));
            System.out.println(" [Producer] Sent message: '" + message + "'");
        } catch (Exception e) {
            System.err.println(" [Producer] Failed to send: " + e.getMessage());
        }
    }
}

Consumer-Side Diagnostics

Logging in your consumer confirms if messages are being received and processed. This helps distinguish between messages not reaching the queue and messages not being picked up by consumers.

Run this consumer, then run the producer from the previous scene.

import com.rabbitmq.client.Channel;
import com.rabbitmq.client.Connection;
import com.rabbitmq.client.ConnectionFactory;
import com.rabbitmq.client.DeliverCallback;

public class DebugConsumer {
    private final static String QUEUE_NAME = "debug_queue_log";

    public static void main(String[] argv) throws Exception {
        ConnectionFactory factory = new ConnectionFactory();
        factory.setHost("localhost"); // Assumes local RabbitMQ

        Connection connection = factory.newConnection();
        Channel channel = connection.createChannel();

        channel.queueDeclare(QUEUE_NAME, false, false, false, null);
        System.out.println(" [Consumer] Waiting for messages. To exit press CTRL+C");

        DeliverCallback deliverCallback = (consumerTag, delivery) -> {
            String message = new String(delivery.getBody(), "UTF-8");
            System.out.println(" [Consumer] Received message: '" + message + "'");
            // Simulate processing
            try {
                Thread.sleep(500); // Simulate work
            } catch (InterruptedException _e) {
                Thread.currentThread().interrupt();
            }
            channel.basicAck(delivery.getEnvelope().getDeliveryTag(), false); // Manual ack
            System.out.println(" [Consumer] Processed & acknowledged: '" + message + "'");
        };

        channel.basicConsume(QUEUE_NAME, false, deliverCallback, consumerTag -> {});
    }
}

Queue Backlogs & Bottlenecks

If messages are accumulating in a queue (visible in the management UI), it indicates a bottleneck. Possible causes include:

No Consumers: No application is listening to the queue.
Slow Consumers: Consumers can't process messages as fast as they arrive.
Consumer Failure: Consumers crashed or stopped without acknowledging messages.
Prefetch Count: Consumers are receiving too many messages at once, leading to slow processing.

Messages to the DLQ?

If messages seem to disappear from their expected queue, check your Dead Letter Queues (DLQs). Messages are dead-lettered when:

They are rejected (basic.reject or basic.nack) and not requeued.
They expire due to TTL (Time-To-Live).
The queue length limit is exceeded.

Monitoring DLQs helps catch undeliverable messages.

Your Debugging Checklist

When a message flow issue arises, follow these steps:

1. Producer Logs: Did the producer confirm sending the message?
2. Exchange Bindings: Is the queue correctly bound to the exchange with the right routing key?
3. Management UI (Queue): Are messages accumulating in the queue? Use 'Get messages'.
4. Consumer Logs: Is the consumer receiving and acknowledging messages?
5. DLQs: Check if messages ended up in a Dead Letter Queue.
6. Test with UI: Use 'Publish message' in the management UI to isolate routing issues.

Tracing the Path

A producer sends messages to a 'logs' direct exchange with a routing key of 'error'. A queue named 'error_logs' is bound to the 'logs' exchange with the routing key 'warning'.

If messages with the 'error' routing key are not appearing in the 'error_logs' queue, what is the MOST likely immediate cause?

Recap: Master Your Message Flow

In this lesson, you've gained essential skills for debugging message flow in RabbitMQ.

You learned to leverage the Management Plugin for inspection and testing.
You saw the importance of logging in both producers and consumers.
You can now identify common issues like routing key mismatches and queue backlogs.
You understand the role of Dead Letter Queues in catching undeliverable messages.

These techniques empower you to diagnose and resolve message delivery problems effectively!

よくある質問

「メッセージフローのデバッグ」レッスンは無料ですか？

はい。「メッセージフローのデバッグ」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。

「メッセージフローのデバッグ」で何を学びますか？

Exchangeやキューを通過するメッセージフローをデバッグするためのツールと技法を活用します。メッセージを追跡して経路を把握し、配信上の問題箇所を特定します。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2/4です。

「メッセージフローのデバッグ」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。

このコースのすべてのレッスン

← RabbitMQ Messaging & Async Systemsに戻る