Load Testing & Performance Benchmarking (JMeter & k6) · レッスン

WebSocketとストリーミングのテスト

WebSocketとストリーミングプロトコルを使って、リアルタイムアプリケーションのパフォーマンステストを行う方法を学びます。

「WebSocketとストリーミングのテスト」はCoddyKit上の無料Load Testing & Performance Benchmarking (JMeter & k6)レッスンです。これはレッスン3/4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Load Testing & Performance Benchmarking (JMeter & k6)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。 Load Testing & Performance Benchmarking (JMeter & k6)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。

Real-Time Apps Need Special Tests

Chat applications, live dashboards, and online games are examples of real-time applications. They demand constant, fast updates and immediate interaction.

Traditional HTTP testing, which relies on a request-response model, doesn't fully capture the behavior of these dynamic systems. We need specific tools and methods to test them effectively.

Understanding WebSockets

WebSockets provide a persistent, bi-directional communication channel over a single TCP connection. This means both the client and server can send data to each other at any time.

Unlike HTTP, which opens and closes connections for each request, WebSockets establish an initial 'handshake' and then maintain an open connection for continuous, 'full-duplex' communication. This makes them ideal for real-time interactions.

Unique Challenges of WebSocket Testing

Performance testing WebSockets presents distinct challenges compared to traditional HTTP:

Persistent Connections: Simulating many open, long-lived connections for extended durations.
Bi-directional Flow: Handling messages sent by both client and server asynchronously.
Dynamic Content: Validating constantly updating data streams rather than static responses.

Our testing tools must be capable of managing these continuous interactions.

JMeter for WebSocket Load Testing

JMeter can be extended to test WebSockets using third-party plugins (e.g., the WebSocket Samplers by Maciej Zaleski).

These plugins allow you to:

Open and close WebSocket connections.
Send messages to the server.
Listen for and capture incoming messages.

You'll configure these steps within JMeter's graphical user interface (GUI).

Configuring a JMeter WebSocket Test

To establish a WebSocket connection in JMeter, you typically add a 'WebSocket Open Connection' sampler. Here, you specify the WebSocket URL (ws:// or wss://).

Subsequent 'WebSocket Request' samplers can then be used to send messages. To end the connection, a 'WebSocket Close' sampler is used. You can also add 'Response Assertions' to validate received messages.

k6: Scripting WebSocket Tests with JavaScript

k6 offers native support for WebSockets through its ws module, making it a powerful and flexible choice for real-time testing.

You write your test logic in JavaScript, defining how virtual users interact with the WebSocket server. This provides great flexibility for creating complex, stateful scenarios that accurately simulate user behavior.

Basic k6 WebSocket Connection

Let's see how to establish a simple WebSocket connection with k6. This script connects to a public test WebSocket echo server and then immediately closes the connection.

import ws from 'k6/ws';
import { check } from 'k6';

export default function () {
  const url = 'ws://echo.websocket.events/'; // A public echo server
  const params = { tags: { ws_tag: 'hello' } };

  const res = ws.connect(url, params, function (socket) {
    socket.on('open', () => console.log('Connected!'));
    socket.on('close', () => console.log('Disconnected!'));
    socket.on('error', (e) => console.log('Error:', e.error()));

    socket.close(); // Close immediately after connecting
  });

  check(res, { 'status is 101': (r) => r && r.status === 101 });
}

Sending and Receiving Messages in k6

Now, let's extend our k6 script to send a message and wait for a response from the echo server. We use socket.send() to transmit data and handle incoming messages with socket.on('message').

import ws from 'k6/ws';
import { check } from 'k6';

export default function () {
  const url = 'ws://echo.websocket.events/';

  ws.connect(url, null, function (socket) {
    socket.on('open', () => {
      console.log('Connected, sending message...');
      socket.send('Hello CoddyKit!');
    });

    socket.on('message', (data) => {
      console.log('Received:', data);
      check(data, { 'message is "Hello CoddyKit!"': (d) => d === 'Hello CoddyKit!' });
      socket.close(); // Close after receiving the message
    });

    socket.on('close', () => console.log('Disconnected!'));
    socket.on('error', (e) => console.log('Error:', e.error()));

    // Keep the VU alive until a message is received or timeout
    socket.prune();
  });
}

Ensuring Correctness with k6 Checks

Just like with HTTP requests, you need to validate the content of WebSocket messages to ensure the real-time data flow is correct under load.

k6's check() function is perfect for this. You can assert on the message content, format, or specific fields received from the server.

import ws from 'k6/ws';
import { check } from 'k6';

export default function () {
  const url = 'ws://echo.websocket.events/';
  const expectedMessage = 'Test Message 123';

  ws.connect(url, null, function (socket) {
    socket.on('open', () => {
      socket.send(expectedMessage);
    });

    socket.on('message', (data) => {
      console.log('Received:', data);
      check(data, {
        'received message matches sent': (d) => d === expectedMessage,
        'message length is correct': (d) => d.length === expectedMessage.length,
      });
      socket.close();
    });

    socket.on('close', () => console.log('Disconnected!'));
    socket.on('error', (e) => console.log('Error:', e.error()));

    socket.prune();
  });
}

Testing Other Streaming Protocols

While WebSockets are popular, other streaming protocols exist:

Server-Sent Events (SSE): A uni-directional protocol where the server pushes updates to the client. Easier for simple data feeds.
gRPC Streaming: Uses HTTP/2 for efficient bi-directional streaming, often favored in microservices architectures. Requires specialized gRPC testing tools.

Each protocol has its own testing considerations and may require dedicated tools or libraries.

Quick Check

Consider a real-time chat application where users constantly send and receive messages. Which statement best describes why WebSockets are generally preferred over traditional HTTP for this scenario?

Recap: Mastering Real-Time Tests

We've explored the unique challenges of testing real-time applications using WebSockets and streaming protocols.

You learned that:

WebSockets offer persistent, bi-directional communication.
JMeter (with plugins) and k6 (native support) are key tools for testing them.
k6 provides flexible JavaScript scripting for complex scenarios.
Validation of dynamic messages using checks is crucial.

Understanding these concepts is vital for ensuring the performance of modern, interactive applications.

よくある質問

「WebSocketとストリーミングのテスト」レッスンは無料ですか？

はい。「WebSocketとストリーミングのテスト」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）、Load Testing & Performance Benchmarking (JMeter & k6)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。 Load Testing & Performance Benchmarking (JMeter & k6)コースには全4レッスンが含まれています。

「WebSocketとストリーミングのテスト」で何を学びますか？

WebSocketとストリーミングプロトコルを使って、リアルタイムアプリケーションのパフォーマンステストを行う方法を学びます。ブラウザで直接実行するハンズオンコードでLoad Testing & Performance Benchmarking (JMeter & k6)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。

Load Testing & Performance Benchmarking (JMeter & k6)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Load Testing & Performance Benchmarking (JMeter & k6)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3/4です。

「WebSocketとストリーミングのテスト」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Load Testing & Performance Benchmarking (JMeter & k6)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Load Testing & Performance Benchmarking (JMeter & k6)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

← Load Testing & Performance Benchmarking (JMeter & k6)に戻る