双方向ストリーミング
クライアントとサーバーの双方が同時に一連のメッセージを送信できる、双方向ストリーミングを使いこなします。
「双方向ストリーミング」はCoddyKit上の無料gRPC & High Performance APIs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gRPC & High Performance APIs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。 gRPC & High Performance APIs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。
Bidirectional Streaming: Two-Way Talk
What if both sides in a conversation could speak and listen at the same time? That's bidirectional streaming in gRPC! It allows both the client and server to send a sequence of messages to each other, concurrently.
Why Use Bidi Streaming?
Bidirectional streaming is perfect for applications needing real-time, interactive communication. Think of it like a phone call where both parties can talk and hear simultaneously.
- Chat applications: Users send and receive messages instantly.
- Live updates: Servers push data to clients as events happen.
- Gaming: Synchronized game state updates between players and server.
Defining a Bidi Stream Service
To enable bidirectional streaming, you use the stream keyword for both the request and response types in your .proto file.
This tells gRPC that the method will handle a continuous flow of messages in both directions.
syntax = "proto3";
package chat;
message ChatMessage {
string user = 1;
string message = 2;
}
service ChatService {
rpc Chat(stream ChatMessage) returns (stream ChatMessage);
}Server: Receiving Client Messages
On the server, your service method will receive a StreamObserver for the incoming client messages. You'll implement its methods to process each message.
onNext(msg): Called for each message from the client.onError(t): Called if the client stream encounters an error.onCompleted(): Called when the client finishes sending messages.
Server: Sending Back Responses
The server also gets a StreamObserver (usually named responseObserver) to send messages back to the client. This allows the server to push multiple responses.
responseObserver.onNext(resp): Sends a message to the client.responseObserver.onError(t): Signals an error to the client.responseObserver.onCompleted(): Closes the server's outgoing stream.
Client: Starting the Conversation
On the client, you'll use an asynchronous (non-blocking) stub to call the streaming method. This call immediately returns a StreamObserver.
This returned observer is what the client uses to send messages to the server.
Client: Two-Way Communication
The client needs two main parts for bidirectional streaming:
- An observer for outgoing messages: This is used to
onNext()messages to the server. - An observer for incoming messages: This is passed to the gRPC call to handle responses from the server.
Both streams operate independently and concurrently.
Full Server Bidi Stream Example
Here's a simplified gRPC server that echoes messages received from the client. Remember to compile your .proto file and include gRPC dependencies.
Try running this example:
import io.grpc.Server;
import io.grpc.ServerBuilder;
import io.grpc.stub.StreamObserver;
// Assuming ChatServiceGrpc and ChatMessage are generated
// from the proto definition.
import chat.ChatServiceGrpc;
import chat.ChatMessage;
public class ChatServer {
public static void main(String[] args) throws Exception {
Server server = ServerBuilder.forPort(50051)
.addService(new ChatServiceImpl())
.build();
server.start();
System.out.println("Server started on port 50051");
server.awaitTermination();
}
static class ChatServiceImpl extends ChatServiceGrpc.ChatServiceImplBase {
@Override
public StreamObserver<ChatMessage> chat(
final StreamObserver<ChatMessage> responseObserver) {
return new StreamObserver<ChatMessage>() {
@Override
public void onNext(ChatMessage request) {
// Received a message from the client
System.out.println("Server received: " + request.getMessage());
// Echo it back to the client
ChatMessage response = ChatMessage.newBuilder()
.setUser("Server")
.setMessage("Echo: " + request.getMessage())
.build();
responseObserver.onNext(response);
}
@Override
public void onError(Throwable t) {
System.err.println("Server error: " + t.getMessage());
}
@Override
public void onCompleted() {
System.out.println("Client stream completed.");
responseObserver.onCompleted(); // Close server stream
}
};
}
}
}Full Client Bidi Stream Example
This client sends a few messages and listens for responses. Run the server first, then this client. The client will send messages and print the server's echoes.
Try running this example:
import io.grpc.ManagedChannel;
import io.grpc.ManagedChannelBuilder;
import io.grpc.stub.StreamObserver;
// Assuming ChatServiceGrpc and ChatMessage are generated
// from the proto definition.
import chat.ChatServiceGrpc;
import chat.ChatMessage;
import java.util.concurrent.CountDownLatch;
import java.util.concurrent.TimeUnit;
public class ChatClient {
public static void main(String[] args) throws Exception {
ManagedChannel channel = ManagedChannelBuilder.forAddress("localhost", 50051)
.usePlaintext() // For demonstration, use TLS in production
.build();
ChatServiceGrpc.ChatStub asyncStub = ChatServiceGrpc.newStub(channel);
CountDownLatch finishLatch = new CountDownLatch(1);
StreamObserver<ChatMessage> requestObserver = asyncStub.chat(
new StreamObserver<ChatMessage>() {
@Override
public void onNext(ChatMessage response) {
System.out.println("Client received: " + response.getMessage());
}
@Override
public void onError(Throwable t) {
System.err.println("Client error: " + t.getMessage());
finishLatch.countDown();
}
@Override
public void onCompleted() {
System.out.println("Server stream completed.");
finishLatch.countDown();
}
});
try {
// Client sends messages
for (int i = 0; i < 3; i++) {
ChatMessage request = ChatMessage.newBuilder()
.setUser("Client")
.setMessage("Hello " + i)
.build();
requestObserver.onNext(request);
Thread.sleep(500); // Simulate delay
}
} catch (RuntimeException | InterruptedException e) {
requestObserver.onError(e);
throw e;
} finally {
requestObserver.onCompleted(); // Client finishes sending
}
if (!finishLatch.await(1, TimeUnit.MINUTES)) {
System.err.println("Client timed out waiting for server response.");
}
channel.shutdownNow().awaitTermination(5, TimeUnit.SECONDS);
}
}Bidirectional Stream Check
You've learned about bidirectional streaming. Which statement best describes how both client and server communicate in a gRPC bidirectional stream?
Bidirectional Streaming Recap
Great job! You've mastered bidirectional streaming in gRPC.
- It enables both client and server to send sequences of messages.
- Ideal for real-time, interactive applications like chat.
- Defined using the
streamkeyword for both request and response in Protobuf. - Requires separate logic on both client and server to manage incoming and outgoing message streams.
This powerful pattern opens up many possibilities for highly responsive distributed systems.
よくある質問
「双方向ストリーミング」レッスンは無料ですか？
はい。「双方向ストリーミング」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）、gRPC & High Performance APIs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。 gRPC & High Performance APIsコースには全4レッスンが含まれています。
「双方向ストリーミング」で何を学びますか？
クライアントとサーバーの双方が同時に一連のメッセージを送信できる、双方向ストリーミングを使いこなします。 ブラウザで直接実行するハンズオンコードでgRPC & High Performance APIs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。
gRPC & High Performance APIs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gRPC & High Performance APIs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。
「双方向ストリーミング」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gRPC & High Performance APIs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gRPC & High Performance APIs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。