Production Debugging & Incident Response Playbook · レッスン

トレーシングツールの活用（例：OpenTelemetry）

OpenTelemetryなど、効果的な監視に役立つ一般的な分散トレーシングツールや標準を実際に使って学びます。

「トレーシングツールの活用（例：OpenTelemetry）」はCoddyKit上の無料Production Debugging & Incident Response Playbook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Production Debugging & Incident Response Playbook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。 Production Debugging & Incident Response Playbook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。

Tracing Tools Introduction

In distributed systems, a single user request often touches many services. Understanding its journey is vital for debugging.

Manual logging across services becomes a nightmare. This is where specialized tracing tools come in, automating the collection and visualization of these request paths.

Meet OpenTelemetry (OTel)

OpenTelemetry (OTel) is a vendor-neutral, open-source set of APIs, SDKs, and tools.

It's designed to standardize how you collect telemetry data: traces, metrics, and logs.
For tracing, OTel helps you instrument your applications to generate and export trace data to a backend of your choice.

OTel Concepts: Tracers & Spans

At the heart of OTel tracing are Tracers and Spans:

A Tracer is an object that creates Span objects.
A Span represents a single unit of work within a trace. It has a name, start and end times, and attributes.
Spans can be nested, forming parent-child relationships to show the flow of operations.

OTel Concepts: Context Propagation

How do spans know they belong to the same request, even across different services?

This is handled by Context Propagation. OTel ensures unique trace identifiers are passed along with requests, linking spans together to form a complete trace.

It's like passing a special ID card along with a task, so everyone knows it's part of the same project.

Setting Up OTel for Your App

To use OpenTelemetry, you typically need to:

Add the OTel SDK to your project (e.g., via Maven or npm).
Initialize the OTel SDK at application startup.
Configure an Exporter to send your trace data to a collection backend.

This setup allows OTel to automatically or manually instrument your code.

Creating Your First Span

Let's see a basic Java example of creating and ending a span. This code uses a ConsoleSpanExporter to print span details to the console.

import io.opentelemetry.api.OpenTelemetry;
import io.opentelemetry.api.trace.Span;
import io.opentelemetry.api.trace.Tracer;
import io.opentelemetry.sdk.OpenTelemetrySdk;
import io.opentelemetry.sdk.trace.SdkTracerProvider;
import io.opentelemetry.sdk.trace.export.ConsoleSpanExporter;
import io.opentelemetry.sdk.trace.export.SimpleSpanProcessor;

public class SimpleSpanDemo {
    private static final OpenTelemetry openTelemetry;
    private static final Tracer tracer;

    static {
        // Configure OpenTelemetry SDK with a console exporter
        SdkTracerProvider tracerProvider = SdkTracerProvider.builder()
            .addSpanProcessor(SimpleSpanProcessor.create(ConsoleSpanExporter.create()))
            .build();
        openTelemetry = OpenTelemetrySdk.builder()
            .setTracerProvider(tracerProvider)
            .buildAndRegisterGlobal();
        tracer = openTelemetry.getTracer("my-app", "1.0.0");
    }

    public static void main(String[] args) {
        System.out.println("App starting...");
        Span span = tracer.spanBuilder("processRequest").startSpan();
        try {
            System.out.println("Inside 'processRequest' span.");
            Thread.sleep(100); // Simulate work
            span.setAttribute("request.id", "XYZ123");
            span.setAttribute("user.name", "coddy");
        } catch (InterruptedException e) {
            Thread.currentThread().interrupt();
        } finally {
            span.end();
            System.out.println("Span 'processRequest' ended.");
        }
        System.out.println("App finished.");
    }
}

Understanding Span Attributes

In the example, we used span.setAttribute("key", "value").

Attributes are key-value pairs that provide rich context to your spans. They help you understand what happened during that unit of work.

Examples include user IDs, request parameters, database query details, or error messages. These attributes make your traces much more useful for debugging.

Connecting Spans Across Services

When a request leaves one service and enters another, OTel uses injectors and extractors to manage context.

The sending service injects trace context (like trace ID) into the request headers.
The receiving service extracts this context to create new spans that are children of the incoming span.

This seamless passing of context is crucial for building end-to-end traces across your microservices.

Exporting and Visualizing Traces

After your application generates spans, the OTel Exporter sends them to a backend system.

Popular tracing backends like Jaeger, Zipkin, or commercial observability platforms (e.g., DataDog, New Relic) then collect and store this data.

These backends provide powerful UIs to visualize the entire trace, showing the path of a request through all services and its latency at each step.

Quick Check: OTel Tracing

Test your understanding of OpenTelemetry's tracing capabilities.

Recap: OTel Power!

Congratulations! You've learned about OpenTelemetry and its role in distributed tracing.

OTel provides a standard way to instrument your code.
Key concepts include Tracers, Spans, and Context Propagation.
You can add custom Attributes to spans for rich context.
Traces are exported to backends for powerful visualization and analysis.

This knowledge is crucial for gaining deep insights into your distributed systems!

よくある質問

「トレーシングツールの活用（例：OpenTelemetry）」レッスンは無料ですか？

はい。「トレーシングツールの活用（例：OpenTelemetry）」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）、Production Debugging & Incident Response Playbook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。 Production Debugging & Incident Response Playbookコースには全4レッスンが含まれています。

「トレーシングツールの活用（例：OpenTelemetry）」で何を学びますか？

OpenTelemetryなど、効果的な監視に役立つ一般的な分散トレーシングツールや標準を実際に使って学びます。ブラウザで直接実行するハンズオンコードでProduction Debugging & Incident Response Playbook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。

Production Debugging & Incident Response Playbook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Production Debugging & Incident Response Playbook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。

「トレーシングツールの活用（例：OpenTelemetry）」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Production Debugging & Incident Response Playbook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Production Debugging & Incident Response Playbook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

← Production Debugging & Incident Response Playbookに戻る