Grpc系列二 Grpc4种服务方法的定义和实现

1. 概述

1.1 服务定义

向其它的RPC服务一样，GPRC的基础是服务的定义。服务定义远程调用方法的名称、传入参数和返回参数。GRPC默认使用 Protobuf描述服务，protobuf的信息见这篇博客Protobuf3 的第一个Java demo

GRPC一共定义4种服务方法：

一元RPC(Unary RPCs )：这是最简单的定义，客户端发送一个请求，服务端返回一个结果
服务器流RPC（Server streaming RPCs）：客户端发送一个请求，服务端返回一个流给客户端，客户从流中读取一系列消息，直到读取所有小心
客户端流RPC(Client streaming RPCs )：客户端通过流向服务端发送一系列消息，然后等待服务端读取完数据并返回处理结果
双向流RPC(Bidirectional streaming RPCs)：客户端和服务端都可以独立向对方发送或接受一系列的消息。客户端和服务端读写的顺序是任意。

以上的服务方法定义在proto文件，如下:

syntax = "proto3";option java_multiple_files = true;
option java_package = "com.hry.spring.grpc.mystream";
option java_outer_classname = "HelloStreamEntity";service HelloStream {// A Unary RPC.rpc simpleRpc(Simple) returns (SimpleFeature) {}// A server-to-client streaming RPC.rpc server2ClientRpc(SimpleList) returns (stream SimpleFeature) {}// A client-to-server streaming RPC.rpc client2ServerRpc(stream Simple) returns (SimpleSummary) {}// A Bidirectional streaming RPC.rpc bindirectionalStreamRpc(stream Simple) returns (stream Simple) {}
}message Simple {int32 num = 1;string name = 2;
}message SimpleList {repeated Simple simpleList = 1;
}message SimpleFeature {string name = 1;Simple location = 2;
}message SimpleSummary {int32 feature_count = 2;
}// 测试类
message SimpleFeatureDatabase {repeated SimpleFeature feature = 1;
}

1.2 同步RPC和异步RPC

GRPC 同时支持同步RPC和异步RPC。
同步RPC调用服务方法只支持流RPC（Server streaming RPCs）和一元RPC(Unary RPCs )。异步RPC调用服务方法支持4种方法。

1.3 生成基础代码 ###

参考的Grpc系列一第一个hello world 例子生成protobuf基础类和HelloStreamGrpc，详细见这里代码

2. 测试代码

这里实现服务端和客户端代码

2.1 服务端

服务端类的实现，通过继承HelloStreamGrpc.HelloStreamImplBase实现，具体服务接口实现见下面：

/*** 服务端类的实现**/private static class HelloStreamService extends HelloStreamGrpc.HelloStreamImplBase {private final List<SimpleFeature> features;public HelloStreamService(List<SimpleFeature> features) {this.features = features;}...

一元RPC(Unary RPCs )服务端实现：

        @Overridepublic void simpleRpc(Simple request, StreamObserver<SimpleFeature> responseObserver) {SimpleFeature rtn = SimpleFeature.newBuilder().setName(request.getName() + "simpleRpc").setLocation(request).build();logger.info("recevier simpleRpc ： {}", request);responseObserver.onNext(rtn);responseObserver.onCompleted();}

服务器流RPC（Server streaming RPCs）服务端实现：

        @Overridepublic void server2ClientRpc(SimpleList request, StreamObserver<SimpleFeature> responseObserver) {logger.info("recevier server2ClientRpc ： {}", request);for (SimpleFeature feature : this.features) {Simple simpleLocation = feature.getLocation();for (Simple o : request.getSimpleListList()) {if (o.getNum() == simpleLocation.getNum()) {// 推送记录responseObserver.onNext(feature);}}}responseObserver.onCompleted();}

客户端流RPC(Client streaming RPCs )服务端实现：

/*** 接收完所有的请求后，才返回一个对象*/@Overridepublic StreamObserver<Simple> client2ServerRpc(StreamObserver<SimpleSummary> responseObserver) {return new StreamObserver<Simple>() {int feature_count = 0;@Overridepublic void onNext(Simple value) {// 接收请求logger.info("num={}, client2ServerRpc, content={} ", feature_count, value);feature_count++;}@Overridepublic void onError(Throwable t) {logger.error("Simple cancelled, e={}", t);}@Overridepublic void onCompleted() {logger.info("onCompleted");// 接收所有请求后，返回总数SimpleSummary summary = SimpleSummary.newBuilder().setFeatureCount(feature_count).build();responseObserver.onNext(summary);// 结束请求responseObserver.onCompleted();}};}

双向流RPC(Bidirectional streaming RPCs)服务端实现：

        /*** 每接收一个请求，立即返回一个对象*/@Overridepublic StreamObserver<Simple> bindirectionalStreamRpc(StreamObserver<Simple> responseObserver) {return new StreamObserver<Simple>() {@Overridepublic void onNext(Simple value) {logger.info("bindirectionalStreamRpc receive {}", value);for (SimpleFeature feature : features) {Simple simpleLocation = feature.getLocation();if (value.getNum() == simpleLocation.getNum()) {// 接收请求后，马上推送记录Simple rtn = Simple.newBuilder().setName(feature.getName() + "rtn").setNum(feature.getLocation().getNum()).build();responseObserver.onNext(rtn);}}}@Overridepublic void onError(Throwable t) {logger.error("bindirectionalStreamRpc cancelled, e={}", t);}@Overridepublic void onCompleted() {responseObserver.onCompleted();}};}

启动服务端main方法：

    private static final Logger logger = LoggerFactory.getLogger(HelloStreamServer.class);private final int port;private final Server server;public HelloStreamServer(int port) throws IOException {this.port = port;this.server = ServerBuilder.forPort(port).addService(new HelloStreamService(HelloUtil.parseFeatures())).build();}// 启动服务public void start() throws IOException {server.start();logger.info("Server started, listening on " + port);Runtime.getRuntime().addShutdownHook(new Thread() {@Overridepublic void run() {System.err.println("*** shutting down gRPC server since JVM is shutting down");HelloStreamServer.this.stop();System.err.println("*** server shut down");}});}// 启动服务public void stop() {if (server != null) {server.shutdown();}}/*** Await termination on the main thread since the grpc library uses daemon* threads.*/private void blockUntilShutdown() throws InterruptedException {if (server != null) {server.awaitTermination();}}public static void main(String[] args) throws Exception {java.util.logging.Logger.getGlobal().setLevel(java.util.logging.Level.OFF);HelloStreamServer server = new HelloStreamServer(8980);server.start();server.blockUntilShutdown();}

2.2 客户端

客户端实现代码：创建客户端stub类
HelloStreamBlockingStub blockingStub：阻塞客户端，支持简单一元服务和流输出调用服务
HelloStreamStub asyncStub：异步客户端，支持所有类型调用

public class HelloStreamClient {private static final Logger logger = LoggerFactory.getLogger(HelloStreamClient.class);private final ManagedChannel channel;private final HelloStreamBlockingStub blockingStub;private final HelloStreamStub asyncStub;private Random random = new Random();public HelloStreamClient(String host, int port) {ManagedChannelBuilder<?> channelBuilder = ManagedChannelBuilder.forAddress(host, port).usePlaintext(true);channel = channelBuilder.build();// 创建一个阻塞客户端，支持简单一元服务和流输出调用服务blockingStub = HelloStreamGrpc.newBlockingStub(channel);// 创建一个异步客户端，支持所有类型调用asyncStub = HelloStreamGrpc.newStub(channel);}public void shutdown() throws InterruptedException {channel.shutdown().awaitTermination(5, TimeUnit.SECONDS);}......// 创建Simple对象private Simple newSimple(int num) {return Simple.newBuilder().setName("simple" + num).setNum(num).build();}}

一元RPC(Unary RPCs )客户端实现：

    /*** 一元服务调用*/public void simpleRpc(int num) {logger.info("request simpleRpc: num={}", num);Simple simple = Simple.newBuilder().setName("simpleRpc").setNum(num).build();SimpleFeature feature;try {feature = blockingStub.simpleRpc(simple);} catch (StatusRuntimeException e) {logger.info("RPC failed: {0}", e.getStatus());return;}logger.info("simpleRpc end called {}", feature);}

服务器流RPC（Server streaming RPCs）客户端实现：

    /*** 阻塞服务器流*/public void server2ClientRpc(int num1, int num2) {logger.info("request server2ClientRpc num1={}, num2={}", num1, num2);Simple simple = Simple.newBuilder().setName("simple" + num1).setNum(num1).build();Simple simple2 = Simple.newBuilder().setName("simple" + num2).setNum(num2).build();SimpleList simpleList = SimpleList.newBuilder().addSimpleList(simple).addSimpleList(simple2).build();Iterator<SimpleFeature> simpleFeatureIter = blockingStub.server2ClientRpc(simpleList);for (int i = 1; simpleFeatureIter.hasNext(); i++) {SimpleFeature feature = simpleFeatureIter.next();logger.info("Result {} : {}", i, feature);}}

客户端流RPC(Client streaming RPCs )客户端实现：

    /*** 异步客户端流* */public void client2ServerRpc(int count) throws InterruptedException {logger.info("request client2ServerRpc {}", count);final CountDownLatch finishLatch = new CountDownLatch(1);StreamObserver<SimpleSummary> responseObserver = new StreamObserver<SimpleSummary>() {@Overridepublic void onNext(SimpleSummary value) {// 返回SimpleSummarylogger.info("client2ServerRpc onNext : {}", value);}@Overridepublic void onError(Throwable t) {logger.error("client2ServerRpc error : {}", t);finishLatch.countDown();}@Overridepublic void onCompleted() {logger.error("client2ServerRpc finish");finishLatch.countDown();}};StreamObserver<Simple> requestObserver = asyncStub.client2ServerRpc(responseObserver);try {for (int i = 0; i < count; i++) {logger.info("simple : {}", i);Simple simple = Simple.newBuilder().setName("client2ServerRpc" + i).setNum(i).build();requestObserver.onNext(simple);Thread.sleep(random.nextInt(200) + 50);}} catch (RuntimeException e) {// Cancel RPCrequestObserver.onError(e);throw e;}// 结束请求requestObserver.onCompleted();// Receiving happens asynchronouslyif (!finishLatch.await(1, TimeUnit.MINUTES)) {logger.error("client2ServerRpc can not finish within 1 minutes");}}

双向流RPC(Bidirectional streaming RPCs)客户端实现：

    /*** 双向流* * @throws InterruptedException*/public void bindirectionalStreamRpc() throws InterruptedException {logger.info("request bindirectionalStreamRpc");final CountDownLatch finishLatch = new CountDownLatch(1);StreamObserver<Simple> requestObserver = asyncStub.bindirectionalStreamRpc(new StreamObserver<Simple>() {@Overridepublic void onNext(Simple value) {logger.info("bindirectionalStreamRpc receive message : {}", value);}@Overridepublic void onError(Throwable t) {logger.error("bindirectionalStreamRpc Failed: {0}", Status.fromThrowable(t));finishLatch.countDown();}@Overridepublic void onCompleted() {logger.info("Finished bindirectionalStreamRpc");finishLatch.countDown();}});try {Simple[] requests = { newSimple(1), newSimple(2), newSimple(3), newSimple(4) };for (Simple request : requests) {logger.info("Sending message {}", request);requestObserver.onNext(request);}} catch (RuntimeException e) {// Cancel RPCrequestObserver.onError(e);throw e;}requestObserver.onCompleted();if (!finishLatch.await(1, TimeUnit.MINUTES)) {logger.error("routeChat can not finish within 1 minutes");}}

启动客户端

    public static void main(String[] args) throws InterruptedException {HelloStreamClient client = new HelloStreamClient("localhost", 8980);try {// simple rpcclient.simpleRpc(1);// server2ClientRpcclient.server2ClientRpc(1, 2);// client2ServerRpcclient.client2ServerRpc(2);// bindirectionalStreamRpcclient.bindirectionalStreamRpc();} finally {client.shutdown();}}

3. 代码

详细代码见这里Github