Netty客户端发送消息并同步获取结果

客户端发送消息并同步获取结果,实际上是违背Netty的设计原则的,可是有时候不得不这么作的话,那么建议进行以下的设计:css

好比咱们的具体用法以下:html

  NettyRequest request = new NettyRequest();
                request.setRequestId(UUID.randomUUID().toString());
                request.setClassName(method.getDeclaringClass().getName());
                request.setMethodName(method.getName());
                request.setParameterTypes(method.getParameterTypes());
                request.setParameterValues(args);

                NettyMessage nettyMessage = new NettyMessage();
                nettyMessage.setType(MessageType.SERVICE_REQ.value());
                nettyMessage.setBody(request);

                if (serviceDiscovery != null) {
                    serverAddress = serviceDiscovery.discover();
                }
                String[] array = serverAddress.split(":");
                String host = array[0];
                int port = Integer.parseInt(array[1]);

                NettyClient client = new NettyClient(host, port);
                NettyMessage nettyResponse = client.send(nettyMessage);
                if (nettyResponse != null) {
                    return JSON.toJSONString(nettyResponse.getBody());
                } else {
                    return null;
                }

先来看看NettyClient的写法 和 send方法的写法:bootstrap

public class NettyClient {

    /**
     * 日志记录
     */
    private static final Logger logger = LoggerFactory.getLogger(NettyClient.class);

    /**
     * 客户端业务处理handler
     */
    private ClientHandler clientHandler = new ClientHandler();

    /**
     * 事件池
     */
    private EventLoopGroup group = new NioEventLoopGroup();

    /**
     * 启动器
     */
    private Bootstrap bootstrap = new Bootstrap();

    /**
     * 客户端通道
     */
    private Channel clientChannel;

    /**
     * 客户端链接
     * @param host
     * @param port
     * @throws InterruptedException
     */
    public NettyClient(String host, int port) throws InterruptedException {
        bootstrap.group(group)
                .channel(NioSocketChannel.class)
                .option(ChannelOption.TCP_NODELAY, true)
                .handler(new ChannelInitializer<SocketChannel>() {
                    @Override
                    protected void initChannel(SocketChannel channel) throws Exception {
                        channel.pipeline().addLast("idleStateHandler", new IdleStateHandler(5, 5, 12));
                        channel.pipeline().addLast("nettyMessageDecoder", new NettyMessageDecoder(1024 * 1024, 4, 4));
                        channel.pipeline().addLast("nettyMessageEncoder", new NettyMessageEncoder());
                        channel.pipeline().addLast("heartBeatHandler", new HeartBeatRequestHandler());
                        channel.pipeline().addLast("clientHandler", clientHandler);
                        channel.pipeline().addLast("loginAuthHandler", new LoginAuthRequestHandler());
                    }
                });

        //发起同步链接操做
        ChannelFuture channelFuture = bootstrap.connect(host, port);

        //注册链接事件
        channelFuture.addListener((ChannelFutureListener)future -> {
            //若是链接成功
            if (future.isSuccess()) {
                logger.info("客户端[" + channelFuture.channel().localAddress().toString() + "]已链接...");
                clientChannel = channelFuture.channel();
            }
            //若是链接失败,尝试从新链接
            else{
                logger.info("客户端[" + channelFuture.channel().localAddress().toString() + "]链接失败,从新链接中...");
                future.channel().close();
                bootstrap.connect(host, port);
            }
        });

        //注册关闭事件
        channelFuture.channel().closeFuture().addListener(cfl -> {
            close();
            logger.info("客户端[" + channelFuture.channel().localAddress().toString() + "]已断开...");
        });
    }

    /**
     * 客户端关闭
     */
    private void close() {
        //关闭客户端套接字
        if(clientChannel!=null){
            clientChannel.close();
        }
        //关闭客户端线程组
        if (group != null) {
            group.shutdownGracefully();
        }
    }

    /**
     * 客户端发送消息
     * @param message
     * @return
     * @throws InterruptedException
     * @throws ExecutionException
     */
    public NettyMessage send(NettyMessage message) throws InterruptedException, ExecutionException {
        ChannelPromise promise = clientHandler.sendMessage(message);
        promise.await(3, TimeUnit.SECONDS);
        return clientHandler.getResponse();
    }
}

能够看出,咱们使用了clientHandler来进行消息发送行为,经过promise阻塞来同步获取返回结果,接下来看看sendMessage的写法:promise

public class ClientHandler extends ChannelInboundHandlerAdapter {

    private static final Logger logger = LoggerFactory.getLogger(ClientHandler.class);

    private ChannelHandlerContext ctx;
    private ChannelPromise promise;
    private NettyMessage response;

    @Override
    public void channelActive(ChannelHandlerContext ctx) throws Exception {
        super.channelActive(ctx);
        this.ctx = ctx;
    }

    @Override
    public void channelRead(ChannelHandlerContext ctx, Object msg) throws Exception {
        NettyMessage message = (NettyMessage) msg;
        if (message != null && message.getType() == MessageType.SERVICE_RESP.value()) {
            response = message;
            promise.setSuccess();
        } else {
            ctx.fireChannelRead(msg);
        }
    }

    public synchronized ChannelPromise sendMessage(Object message) {
        while (ctx == null) {
            try {
                TimeUnit.MILLISECONDS.sleep(1);
                //logger.error("等待ChannelHandlerContext实例化");
            } catch (InterruptedException e) {
                logger.error("等待ChannelHandlerContext实例化过程当中出错",e);
            }
        }
        promise = ctx.newPromise();
        ctx.writeAndFlush(message);
        return promise;
    }

    public NettyMessage getResponse(){
        return response;
    }

}

能够看到,在利用ChannelHanderContext进行发送消息前,咱们先建立了一个promise并返回给send方法,那么send方法此时就会阻塞等待;当咱们收到服务端消息后,promise.setSuccess就会解除send方法的等待行为,这样咱们就能获取结果了。dom

 

此法针对真正须要同步等待获取结果的场景,如非必要,仍是建议利用future来改造。ide

 

benchmark测试代表,此种同步获取结果的行为,表现挺稳定的,可是ops 在 150 左右, 真是性能太差了。高性能场合禁用此法。oop

相关文章
相关标签/搜索