Advanced Spring Boot 4: Event-Driven Architecture (Kafka) · レッスン

KStreamとKTableによるストリーム処理

不変のイベントストリームにはKStreamを、状態を持ち更新可能なデータビューにはKTableを使用し、フィルタリングやマッピングなどの操作を行う方法を学習します。

「KStreamとKTableによるストリーム処理」はCoddyKit上の無料Advanced Spring Boot 4: Event-Driven Architecture (Kafka)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Advanced Spring Boot 4: Event-Driven Architecture (Kafka)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。 Advanced Spring Boot 4: Event-Driven Architecture (Kafka)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。

KStream & KTable Unveiled

Welcome! In Kafka Streams, KStream and KTable are your primary tools for processing data. They represent different views of your data in motion.

Think of them as two sides of the same coin, each suited for distinct stream processing tasks. Understanding their differences is key to building powerful stream applications.

KStream: Immutable Events

A KStream represents an infinite, immutable sequence of events. Each record in a KStream is a self-contained fact, an independent event that happened at a specific point in time.

It's like a transaction log: once an event is added, it's never changed.
Operations on a KStream produce new KStreams, leaving the original untouched.
It's ideal for processing individual events like clicks, sensor readings, or log entries.

Filtering KStream Events

One common KStream operation is filtering. You can selectively keep records that match certain criteria, creating a new KStream with only the relevant events.

Here's a simple example filtering messages that contain 'hello'.

import org.apache.kafka.common.serialization.Serdes;
import org.apache.kafka.streams.KafkaStreams;
import org.apache.kafka.streams.StreamsBuilder;
import org.apache.kafka.streams.StreamsConfig;
import org.apache.kafka.streams.kstream.KStream;

import java.util.Properties;

public class Main {
    public static void main(String[] args) {
        Properties props = new Properties();
        props.put(StreamsConfig.APPLICATION_ID_CONFIG, "filter-app");
        props.put(StreamsConfig.BOOTSTRAP_SERVERS_CONFIG, "localhost:9092");
        props.put(StreamsConfig.DEFAULT_KEY_SERDE_CLASS_BY_DEFAULT, Serdes.String().getClass());
        props.put(StreamsConfig.DEFAULT_VALUE_SERDE_CLASS_BY_DEFAULT, Serdes.String().getClass());

        StreamsBuilder builder = new StreamsBuilder();
        KStream<String, String> sourceStream = builder.stream("input-topic");

        KStream<String, String> filteredStream = sourceStream.filter(
            (key, value) -> value.contains("hello")
        );

        filteredStream.to("output-topic");

        KafkaStreams streams = new KafkaStreams(builder.build(), props);
        // In a real app, you'd start and manage this lifecycle:
        // streams.start();
        // Runtime.getRuntime().addShutdownHook(new Thread(streams::close));
        System.out.println("KStream filter setup complete. Send 'hello world' to input-topic!");
    }
}

Transforming KStream Values

The mapValues operation transforms the value of each record in a KStream, producing a new KStream with the modified values. The key remains unchanged.

This is useful for cleaning data, changing formats, or enriching information without altering the message's key.

import org.apache.kafka.common.serialization.Serdes;
import org.apache.kafka.streams.KafkaStreams;
import org.apache.kafka.streams.StreamsBuilder;
import org.apache.kafka.streams.StreamsConfig;
import org.apache.kafka.streams.kstream.KStream;

import java.util.Properties;

public class Main {
    public static void main(String[] args) {
        Properties props = new Properties();
        props.put(StreamsConfig.APPLICATION_ID_CONFIG, "mapvalues-app");
        props.put(StreamsConfig.BOOTSTRAP_SERVERS_CONFIG, "localhost:9092");
        props.put(StreamsConfig.DEFAULT_KEY_SERDE_CLASS_BY_DEFAULT, Serdes.String().getClass());
        props.put(StreamsConfig.DEFAULT_VALUE_SERDE_CLASS_BY_DEFAULT, Serdes.String().getClass());

        StreamsBuilder builder = new StreamsBuilder();
        KStream<String, String> sourceStream = builder.stream("input-topic");

        KStream<String, String> uppercasedStream = sourceStream.mapValues(
            value -> value.toUpperCase()
        );

        uppercasedStream.to("output-topic");

        KafkaStreams streams = new KafkaStreams(builder.build(), props);
        System.out.println("KStream mapValues setup complete. Send 'test' to input-topic!");
    }
}

KTable: A Materialized View

A KTable represents a changelog stream, where each record is an update to a specific key. It's essentially a materialized view of a table, reflecting the latest state for each key.

It's like a database table: keys have associated values, and new records for a key overwrite previous ones.
KTable is stateful, maintaining the latest value for each key over time.
It's perfect for aggregating data, maintaining counts, or storing user profiles.

KTable's Stateful Nature

The core idea behind a KTable is that it keeps track of the latest value for each unique key. When a new record with an existing key arrives, the KTable updates its internal state.

This makes KTable ideal for scenarios where you care about the current state of an entity, rather than every single event that led to that state.

KStream to KTable: Counting

You can transform a KStream into a KTable, typically to perform aggregations. A common example is counting occurrences of keys using groupByKey().count().

Each time a message arrives, the count for its key is updated, and the KTable emits the new total.

import org.apache.kafka.common.serialization.Serdes;
import org.apache.kafka.streams.KafkaStreams;
import org.apache.kafka.streams.StreamsBuilder;
import org.apache.kafka.streams.StreamsConfig;
import org.apache.kafka.streams.kstream.KStream;
import org.apache.kafka.streams.kstream.KTable;
import org.apache.kafka.streams.kstream.Materialized;

import java.util.Properties;

public class Main {
    public static void main(String[] args) {
        Properties props = new Properties();
        props.put(StreamsConfig.APPLICATION_ID_CONFIG, "kstream-to-ktable-app");
        props.put(StreamsConfig.BOOTSTRAP_SERVERS_CONFIG, "localhost:9092");
        props.put(StreamsConfig.DEFAULT_KEY_SERDE_CLASS_BY_DEFAULT, Serdes.String().getClass());
        props.put(StreamsConfig.DEFAULT_VALUE_SERDE_CLASS_BY_DEFAULT, Serdes.String().getClass());

        StreamsBuilder builder = new StreamsBuilder();
        KStream<String, String> sourceStream = builder.stream("input-topic");

        KTable<String, Long> wordCounts = sourceStream
            .groupByKey() // Group by the existing key
            .count(Materialized.as("counts-store")); // Count occurrences, store in state

        wordCounts.toStream().to("output-topic"); // Convert back to stream to send out

        KafkaStreams streams = new KafkaStreams(builder.build(), props);
        System.out.println("KStream to KTable count setup. Send 'word' with key 'A' to input-topic!");
    }
}

KTable for Aggregation

KTables are excellent for continuous aggregation. Beyond simple counts, you can use operations like aggregate to maintain sums, averages, or custom aggregates over time.

This allows your application to always have an up-to-date summary of data for specific keys.

When to Use Which?

The choice between KStream and KTable depends on your processing needs:

Use KStream when you need to process individual events, react to every occurrence, or build a pipeline of transformations that don't depend on historical state. Think real-time alerts or event logging.
Use KTable when you need to maintain a current state, aggregate data over time, or join with other data sources based on the latest value. Think user profiles, stock prices, or aggregated metrics.

KStream vs. KTable Check

You've learned about KStream and KTable. Let's test your understanding.

KStream & KTable Recap

Great job! You've explored the core differences and uses of KStream and KTable.

KStream handles individual, immutable events, perfect for event-by-event processing.
KTable maintains a materialized view, tracking the latest state for each key, ideal for aggregations and stateful processing.

These two primitives are the foundation for building powerful and flexible stream processing applications with Kafka Streams.

よくある質問

「KStreamとKTableによるストリーム処理」レッスンは無料ですか？

はい。「KStreamとKTableによるストリーム処理」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）、Advanced Spring Boot 4: Event-Driven Architecture (Kafka)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。 Advanced Spring Boot 4: Event-Driven Architecture (Kafka)コースには全4レッスンが含まれています。

「KStreamとKTableによるストリーム処理」で何を学びますか？

不変のイベントストリームにはKStreamを、状態を持ち更新可能なデータビューにはKTableを使用し、フィルタリングやマッピングなどの操作を行う方法を学習します。ブラウザで直接実行するハンズオンコードでAdvanced Spring Boot 4: Event-Driven Architecture (Kafka)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。

Advanced Spring Boot 4: Event-Driven Architecture (Kafka)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Advanced Spring Boot 4: Event-Driven Architecture (Kafka)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。

「KStreamとKTableによるストリーム処理」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Advanced Spring Boot 4: Event-Driven Architecture (Kafka)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Advanced Spring Boot 4: Event-Driven Architecture (Kafka)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

← Advanced Spring Boot 4: Event-Driven Architecture (Kafka)に戻る