Skip to content
标签
网络
字数
13928 字
阅读时间
68 分钟

一、概述

Netty 是一个异步的、基于事件驱动的网络应用框架,用于快速开发可维护、高性能的网络服务器和客户端

1.1 Netty应用

Netty 在 Java 网络应用框架中的地位就好比:Spring 框架在 JavaEE 开发中的地位

以下的框架都使用了 Netty,因为它们有网络通信需求!

  • Cassandra - nosql 数据库
  • Spark - 大数据分布式计算框架
  • Hadoop - 大数据分布式存储框架
  • RocketMQ - ali 开源的消息队列
  • ElasticSearch - 搜索引擎
  • gRPC - rpc 框架
  • Dubbo - rpc 框架
  • Spring 5.x - flux api 完全抛弃了 tomcat ,使用 netty 作为服务器端
  • Zookeeper - 分布式协调框架

1.2 Netty的优势

  • Netty vs NIO,工作量大,bug 多
    • 需要自己构建协议
    • 解决 TCP 传输问题,如粘包、半包
    • epoll 空轮询导致 CPU 100%
    • 对 API 进行增强,使之更易用,如 FastThreadLocal => ThreadLocal,ByteBuf => ByteBuffer
  • Netty vs 其它网络应用框架
    • Mina 由 apache 维护,将来 3.x 版本可能会有较大重构,破坏 API 向下兼容性,Netty 的开发迭代更迅速,API 更简洁、文档更优秀
    • 久经考验,16年,Netty 版本
      • 2.x 2004
      • 3.x 2008
      • 4.x 2013
      • 5.x 已废弃(没有明显的性能提升,维护成本高)

二、基本使用

2.1 入门demo

依赖

xml
<dependency>
    <groupId>io.netty</groupId>
    <artifactId>netty-all</artifactId>
    <version>4.1.39.Final</version>
</dependency>

服务端

java
new ServerBootstrap()
    // 创建 NioEventLoopGroup,可以简单理解为 线程池 + Selector
    .group(new NioEventLoopGroup()) 
    // 选择服务 Scoket 实现类,其中 NioServerSocketChannel 表示基于 NIO 的服务器端实现
    .channel(NioServerSocketChannel.class) 
    // 添加的处理器都是给 SocketChannel 用的,而不是给 ServerSocketChannel。
    .childHandler(new ChannelInitializer<NioSocketChannel>() { 
        protected void initChannel(NioSocketChannel ch) {
            // SocketChannel 的处理器,解码 ByteBuf => String
            ch.pipeline().addLast(new StringDecoder());
            // SocketChannel 的业务处理器,使用上一个处理器的处理结果
            ch.pipeline().addLast(new SimpleChannelInboundHandler<String>() { 
                @Override
                protected void channelRead0(ChannelHandlerContext ctx, String msg) {
                    System.out.println(msg);
                }
            });
        }
    })
    // ServerSocketChannel 绑定的监听端口
    .bind(8080);

客户端

java
nehttps://img.xplly.eu.org/picventLoopGroup,同 Server
    .group(new NioEventLoopGroup())
    // 选择客户 Socket 实现类,NioSocketChannel 表示基于 NIO 的客户端实现
    .channel(NioSocketChannel.class) 
    // 添加 SocketChannel 的处理器,ChannelInitializer 处理器(仅执行一次),它的作用是待客户端 SocketChannel 建立连接后,执行 initChannel 以便添加更多的处理器
    .handler(new ChannelInitializer<Channel>() {
        @Override
        protected void initChannel(Channel ch) {
            // 消息会经过通道 handler 处理,这里是将 String => ByteBuf 发出
            ch.pipeline().addLast(new StringEncoder()); 
        }
    })
    // 指定要连接的服务器和端口
    .connect("127.0.0.1", 8080) 
    // Netty 中很多方法都是异步的,如 connect,这时需要使用 sync 方法等待 connect 建立连接完毕
    .sync() 
    // 获取 channel 对象,它即为通道抽象,可以进行数据读写操作
    .channel()
    // 写入消息并清空缓冲区
    .writeAndFlush(new Date() + ": hello world!");

相关概念

  • 把 channel 理解为数据的通道
  • 把 msg 理解为流动的数据,最开始输入是 ByteBuf,但经过 pipeline 的加工,会变成其它类型对象,最后输出又变成 ByteBuf
  • 把 handler 理解为数据的处理工序
    • 工序有多道,合在一起就是 pipeline,pipeline 负责发布事件(读、读取完成...)传播给每个 handler, handler 对自己感兴趣的事件进行处理(重写了相应事件处理方法)
    • handler 分 Inbound 和 Outbound 两类
  • 把 eventLoop 理解为处理数据的工人
    • 工人可以管理多个 channel 的 io 操作,并且一旦工人负责了某个 channel,就要负责到底(绑定)
    • 工人既可以执行 io 操作,也可以进行任务处理,每位工人有任务队列,队列里可以堆放多个 channel 的待处理任务,任务分为普通任务、定时任务
    • 工人按照 pipeline 顺序,依次按照 handler 的规划(代码)处理数据,可以为每道工序指定不同的工人

三、组件

3.1 EventLoop

事件循环对象

EventLoop 本质是一个单线程执行器(同时维护了一个 Selector),里面有 run 方法处理 Channel 上源源不断的 io 事件。

它的继承关系比较复杂

  • 一条线是继承自 j.u.c.ScheduledExecutorService 因此包含了线程池中所有的方法
  • 另一条线是继承自 netty 自己的 OrderedEventExecutor,
    • 提供了 boolean inEventLoop(Thread thread) 方法判断一个线程是否属于此 EventLoop
    • 提供了 parent 方法来看看自己属于哪个 EventLoopGroup

事件循环组

EventLoopGroup 是一组 EventLoop,Channel 一般会调用 EventLoopGroup 的 register 方法来绑定其中一个 EventLoop,后续这个 Channel 上的 io 事件都由此 EventLoop 来处理(保证了 io 事件处理时的线程安全)

  • 继承自 netty 自己的 EventExecutorGroup
    • 实现了 Iterable 接口提供遍历 EventLoop 的能力
    • 另有 next 方法获取集合中下一个 EventLoop

可通过 .next方法和遍历group对象获取其中的事件循环对象

优雅关闭

优雅关闭 shutdownGracefully 方法。该方法会首先切换 EventLoopGroup 到关闭状态从而拒绝新的任务的加入,然后在任务队列的任务都处理完成后,停止线程的运行。从而确保整体应用是在正常有序的状态下退出的

服务器添加worker工人处理io事件

服务器端

java
// 非nio工人
DefaultEventLoopGroup normalWorkers = new DefaultEventLoopGroup(2);
new ServerBootstrap()
    .group(new NioEventLoopGroup(1), new NioEventLoopGroup(2))
    .channel(NioServerSocketChannel.class)
    .childHandler(new ChannelInitializer<NioSocketChannel>() {
        @Override
        protected void initChannel(NioSocketChannel ch)  {
            
           // nio处理事件
            ch.pipeline().addLast(new LoggingHandler(LogLevel.DEBUG));
            // 添加非nio处理读取事件,不添加normalWorkers会采用默认定义的nio
            ch.pipeline().addLast(normalWorkers,"myhandler",
              new ChannelInboundHandlerAdapter() {
                @Override
                public void channelRead(ChannelHandlerContext ctx, Object msg) {
                    ByteBuf byteBuf = msg instanceof ByteBuf ? ((ByteBuf) msg) : null;
                    if (byteBuf != null) {
                        byte[] buf = new byte[16];
                        ByteBuf len = byteBuf.readBytes(buf, 0, byteBuf.readableBytes());
                        log.debug(new String(buf));
                    }
                }
            });
        }
    }).bind(8080).sync();

客户端

java
public static void main(String[] args) throws InterruptedException {
    Channel channel = new Bootstrap()
            .group(new NioEventLoopGroup(1))
            .handler(new ChannelInitializer<NioSocketChannel>() {
                @Override
                protected void initChannel(NioSocketChannel ch) throws Exception {
                    System.out.println("init...");
                    ch.pipeline().addLast(new LoggingHandler(LogLevel.DEBUG));
                }
            })
            .channel(NioSocketChannel.class).connect("localhost", 8080)
            .sync()
            .channel();

    channel.writeAndFlush(ByteBufAllocator.DEFAULT.buffer().writeBytes("wangwu".getBytes()));
    Thread.sleep(2000);
    channel.writeAndFlush(ByteBufAllocator.DEFAULT.buffer().writeBytes("wangwu".getBytes()));
handler 执行中如何换人?

关键代码 io.netty.channel.AbstractChannelHandlerContext#invokeChannelRead()

java
static void invokeChannelRead(final AbstractChannelHandlerContext next, Object msg) {
    final Object m = next.pipeline.touch(ObjectUtil.checkNotNull(msg, "msg"), next);
    // 下一个 handler 的事件循环是否与当前的事件循环是同一个线程
    EventExecutor executor = next.executor();
    
    // 是,直接调用
    if (executor.inEventLoop()) {
        next.invokeChannelRead(m);
    } 
    // 不是,将要执行的代码作为任务提交给下一个事件循环处理(换人)
    else {
        executor.execute(new Runnable() {
            @Override
            public void run() {
                next.invokeChannelRead(m);
            }
        });
    }
}
  • 如果两个 handler 绑定的是同一个线程,那么就直接调用
  • 否则,把要调用的代码封装为一个任务对象,由下一个 handler 的线程来调用
NioEventLoop 处理普通任务

NioEventLoop 除了可以处理 io 事件,同样可以向它提交普通任务

java
NioEventLoopGroup nioWorkers = new NioEventLoopGroup(2);

log.debug("server start...");
Thread.sleep(2000);
nioWorkers.execute(()->{
    log.debug("normal task...");
});

输出

22:30:36 [DEBUG] [main] c.i.o.EventLoopTest2 - server start...
22:30:38 [DEBUG] [nioEventLoopGroup-2-1] c.i.o.EventLoopTest2 - normal task...

可以用来执行耗时较长的任务

NioEventLoop 处理定时任务
java
NioEventLoopGroup nioWorkers = new NioEventLoopGroup(2);

log.debug("server start...");
Thread.sleep(2000);
nioWorkers.scheduleAtFixedRate(() -> {
    log.debug("running...");
}, 0, 1, TimeUnit.SECONDS);

输出

22:35:15 [DEBUG] [main] c.i.o.EventLoopTest2 - server start...
22:35:17 [DEBUG] [nioEventLoopGroup-2-1] c.i.o.EventLoopTest2 - running...
22:35:18 [DEBUG] [nioEventLoopGroup-2-1] c.i.o.EventLoopTest2 - running...
22:35:19 [DEBUG] [nioEventLoopGroup-2-1] c.i.o.EventLoopTest2 - running...
22:35:20 [DEBUG] [nioEventLoopGroup-2-1] c.i.o.EventLoopTest2 - running...
...

可以用来执行定时任务

3.2 Channel

channel 的主要作用

  • close() 可以用来关闭 channel
  • closeFuture() 用来处理 channel 的关闭
    • sync 方法作用是同步等待 channel 关闭
    • 而 addListener 方法是异步等待 channel 关闭
  • pipeline() 方法添加处理器
  • write() 方法将数据写入
  • writeAndFlush() 方法将数据写入并刷出
ChannelFuture
java
// 返回的是 ChannelFuture 对象,它的作用是利用 channel() 方法来获取 Channel 对象
//  connect 方法是异步的,意味着不等连接建立,方法执行就返回了。因此 channelFuture 对象中不能【立刻】获得到正确的 Channel 对象
ChannelFuture channelFuture = new Bootstrap()
    .group(new NioEventLoopGroup())
    .channel(NioSocketChannel.class)
    .handler(new ChannelInitializer<Channel>() {
        @Override
        protected void initChannel(Channel ch) {
            ch.pipeline().addLast(new StringEncoder());
        }
    })
    .connect("127.0.0.1", 8080);
// 连接未建立,打印 [id: 0x2e1884dd]
System.out.println(channelFuture.channel());
// sync 方法是同步等待连接建立完成
channelFuture.sync(); 
// 异步回调方式 等待连接建立
channelFuture.addListener((ChannelFutureListener) future -> {
    System.out.println(future.channel()); // 2
});

// 连接肯定建立了,打印 [id: 0x2e1884dd, L:/127.0.0.1:57191 - R:/127.0.0.1:8080]
System.out.println(channelFuture.channel()); // 3
CloseFuture
java
@Slf4j
public class CloseFutureClient {
    public static void main(String[] args) throws InterruptedException {
        NioEventLoopGroup group new NioEventLoopGroup();
        ChannelFuture channelFuture = new Bootstrap()
                .group(group)
                .channel(NioSocketChannel.class)
                .handler(new ChannelInitializer<NioSocketChannel>() {
                    @Override // 在连接建立后被调用
                    protected void initChannel(NioSocketChannel ch) throws Exception {
                        ch.pipeline().addLast(new LoggingHandler(LogLevel.DEBUG));
                        ch.pipeline().addLast(new StringEncoder());
                    }
                })
                .connect(new InetSocketAddress("localhost", 8080));
        Channel channel = channelFuture.sync().channel();
        log.debug("{}", channel);
        new Thread(()->{
            Scanner scanner = new Scanner(System.in);
            while (true) {
                String line = scanner.nextLine();
                if ("q".equals(line)) {
                    channel.close(); // close 异步操作 1s 之后
//                    log.debug("处理关闭之后的操作"); // 不能在这里善后
                    break;
                }
                channel.writeAndFlush(line);
            }
        }, "input").start();

        // 获取 CloseFuture 对象, 1) 同步处理关闭, 2) 异步处理关闭
        ChannelFuture closeFuture = channel.closeFuture();
        /*log.debug("waiting close...");
        closeFuture.sync();
        log.debug("处理关闭之后的操作");*/
        closeFuture.addListener(new ChannelFutureListener() {
            @Override
            public void operationComplete(ChannelFuture future) throws Exception {
                log.debug("处理关闭之后的操作");
                group.shutdownGracefully();
            }
        });
    }
}

3.3 Future & Promise

在异步处理时,经常用到这两个接口

首先要说明 netty 中的 Future 与 jdk 中的 Future 同名,但是是两个接口,netty 的 Future 继承自 jdk 的 Future,而 Promise 又对 netty Future 进行了扩展

  • jdk Future 只能同步等待任务结束(或成功、或失败)才能得到结果
  • netty Future 可以同步等待任务结束得到结果,也可以异步方式得到结果,但都是要等任务结束
  • netty Promise 不仅有 netty Future 的功能,而且脱离了任务独立存在,只作为两个线程间传递结果的容器
功能/名称jdk Futurenetty FuturePromise
cancel取消任务--
isCanceled任务是否取消--
isDone任务是否完成,不能区分成功失败--
get获取任务结果,阻塞等待--
getNow-获取任务结果,非阻塞,还未产生结果时返回 null-
await-等待任务结束,如果任务失败,不会抛异常,而是通过 isSuccess 判断-
sync-等待任务结束,如果任务失败,抛出异常-
isSuccess-判断任务是否成功-
cause-获取失败信息,非阻塞,如果没有失败,返回null-
addLinstener-添加回调,异步接收结果-
setSuccess--设置成功结果
setFailure--设置失败结果
例1 同步处理任务成功
java
DefaultEventLoop eventExecutors = new DefaultEventLoop();
DefaultPromise<Integer> promise = new DefaultPromise<>(eventExecutors);

eventExecutors.execute(()->{
    try {
        Thread.sleep(1000);
    } catch (InterruptedException e) {
        e.printStackTrace();
    }
    log.debug("set success, {}",10);
    promise.setSuccess(10);
});

log.debug("start...");
log.debug("{}",promise.getNow()); // 还没有结果
log.debug("{}",promise.get());

输出

11:51:53 [DEBUG] [main] c.i.o.DefaultPromiseTest2 - start...
11:51:53 [DEBUG] [main] c.i.o.DefaultPromiseTest2 - null
11:51:54 [DEBUG] [defaultEventLoop-1-1] c.i.o.DefaultPromiseTest2 - set success, 10
11:51:54 [DEBUG] [main] c.i.o.DefaultPromiseTest2 - 10
例2 异步处理任务成功
java
DefaultEventLoop eventExecutors = new DefaultEventLoop();
DefaultPromise<Integer> promise = new DefaultPromise<>(eventExecutors);

// 设置回调,异步接收结果
promise.addListener(future -> {
    // 这里的 future 就是上面的 promise
    log.debug("{}",future.getNow());
});

// 等待 1000 后设置成功结果
eventExecutors.execute(()->{
    try {
        Thread.sleep(1000);
    } catch (InterruptedException e) {
        e.printStackTrace();
    }
    log.debug("set success, {}",10);
    promise.setSuccess(10);
});

log.debug("start...");

输出

11:49:30 [DEBUG] [main] c.i.o.DefaultPromiseTest2 - start...
11:49:31 [DEBUG] [defaultEventLoop-1-1] c.i.o.DefaultPromiseTest2 - set success, 10
11:49:31 [DEBUG] [defaultEventLoop-1-1] c.i.o.DefaultPromiseTest2 - 10
例3 同步处理任务失败 - sync & get
java
DefaultEventLoop eventExecutors = new DefaultEventLoop();
        DefaultPromise<Integer> promise = new DefaultPromise<>(eventExecutors);

        eventExecutors.execute(() -> {
            try {
                Thread.sleep(1000);
            } catch (InterruptedException e) {
                e.printStackTrace();
            }
            RuntimeException e = new RuntimeException("error...");
            log.debug("set failure, {}", e.toString());
            promise.setFailure(e);
        });

        log.debug("start...");
        log.debug("{}", promise.getNow());
        promise.get(); // sync() 也会出现异常,只是 get 会再用 ExecutionException 包一层异常

输出

12:11:07 [DEBUG] [main] c.i.o.DefaultPromiseTest2 - start...
12:11:07 [DEBUG] [main] c.i.o.DefaultPromiseTest2 - null
12:11:08 [DEBUG] [defaultEventLoop-1-1] c.i.o.DefaultPromiseTest2 - set failure, java.lang.RuntimeException: error...
Exception in thread "main" java.util.concurrent.ExecutionException: java.lang.RuntimeException: error...
	at io.netty.util.concurrent.AbstractFuture.get(AbstractFuture.java:41)
	at com.itcast.oio.DefaultPromiseTest2.main(DefaultPromiseTest2.java:34)
Caused by: java.lang.RuntimeException: error...
	at com.itcast.oio.DefaultPromiseTest2.lambda$main$0(DefaultPromiseTest2.java:27)
	at io.netty.channel.DefaultEventLoop.run(DefaultEventLoop.java:54)
	at io.netty.util.concurrent.SingleThreadEventExecutor$5.run(SingleThreadEventExecutor.java:918)
	at io.netty.util.internal.ThreadExecutorMap$2.run(ThreadExecutorMap.java:74)
	at io.netty.util.concurrent.FastThreadLocalRunnable.run(FastThreadLocalRunnable.java:30)
	at java.lang.Thread.run(Thread.java:745)
例4 同步处理任务失败 - await
java
DefaultEventLoop eventExecutors = new DefaultEventLoop();
DefaultPromise<Integer> promise = new DefaultPromise<>(eventExecutors);

eventExecutors.execute(() -> {
    try {
        Thread.sleep(1000);
    } catch (InterruptedException e) {
        e.printStackTrace();
    }
    RuntimeException e = new RuntimeException("error...");
    log.debug("set failure, {}", e.toString());
    promise.setFailure(e);
});

log.debug("start...");
log.debug("{}", promise.getNow());
promise.await(); // 与 sync 和 get 区别在于,不会抛异常
log.debug("result {}", (promise.isSuccess() ? promise.getNow() : promise.cause()).toString());

输出

12:18:53 [DEBUG] [main] c.i.o.DefaultPromiseTest2 - start...
12:18:53 [DEBUG] [main] c.i.o.DefaultPromiseTest2 - null
12:18:54 [DEBUG] [defaultEventLoop-1-1] c.i.o.DefaultPromiseTest2 - set failure, java.lang.RuntimeException: error...
12:18:54 [DEBUG] [main] c.i.o.DefaultPromiseTest2 - result java.lang.RuntimeException: error...
例5 异步处理任务失败
java
DefaultEventLoop eventExecutors = new DefaultEventLoop();
DefaultPromise<Integer> promise = new DefaultPromise<>(eventExecutors);

promise.addListener(future -> {
    log.debug("result {}", (promise.isSuccess() ? promise.getNow() : promise.cause()).toString());
});

eventExecutors.execute(() -> {
    try {
        Thread.sleep(1000);
    } catch (InterruptedException e) {
        e.printStackTrace();
    }
    RuntimeException e = new RuntimeException("error...");
    log.debug("set failure, {}", e.toString());
    promise.setFailure(e);
});

log.debug("start...");

输出

12:04:57 [DEBUG] [main] c.i.o.DefaultPromiseTest2 - start...
12:04:58 [DEBUG] [defaultEventLoop-1-1] c.i.o.DefaultPromiseTest2 - set failure, java.lang.RuntimeException: error...
12:04:58 [DEBUG] [defaultEventLoop-1-1] c.i.o.DefaultPromiseTest2 - result java.lang.RuntimeException: error...
例6 await 死锁检查
java
DefaultEventLoop eventExecutors = new DefaultEventLoop();
DefaultPromise<Integer> promise = new DefaultPromise<>(eventExecutors);

eventExecutors.submit(()->{
    System.out.println("1");
    try {
        promise.await();
        // 注意不能仅捕获 InterruptedException 异常
        // 否则 死锁检查抛出的 BlockingOperationException 会继续向上传播
        // 而提交的任务会被包装为 PromiseTask,它的 run 方法中会 catch 所有异常然后设置为 Promise 的失败结果而不会抛出
    } catch (Exception e) { 
        e.printStackTrace();
    }
    System.out.println("2");
});
eventExecutors.submit(()->{
    System.out.println("3");
    try {
        promise.await();
    } catch (Exception e) {
        e.printStackTrace();
    }
    System.out.println("4");
});

输出

1
2
3
4
io.netty.util.concurrent.BlockingOperationException: DefaultPromise@47499c2a(incomplete)
	at io.netty.util.concurrent.DefaultPromise.checkDeadLock(DefaultPromise.java:384)
	at io.netty.util.concurrent.DefaultPromise.await(DefaultPromise.java:212)
	at com.itcast.oio.DefaultPromiseTest.lambda$main$0(DefaultPromiseTest.java:27)
	at io.netty.util.concurrent.PromiseTask$RunnableAdapter.call(PromiseTask.java:38)
	at io.netty.util.concurrent.PromiseTask.run(PromiseTask.java:73)
	at io.netty.channel.DefaultEventLoop.run(DefaultEventLoop.java:54)
	at io.netty.util.concurrent.SingleThreadEventExecutor$5.run(SingleThreadEventExecutor.java:918)
	at io.netty.util.internal.ThreadExecutorMap$2.run(ThreadExecutorMap.java:74)
	at io.netty.util.concurrent.FastThreadLocalRunnable.run(FastThreadLocalRunnable.java:30)
	at java.lang.Thread.run(Thread.java:745)
io.netty.util.concurrent.BlockingOperationException: DefaultPromise@47499c2a(incomplete)
	at io.netty.util.concurrent.DefaultPromise.checkDeadLock(DefaultPromise.java:384)
	at io.netty.util.concurrent.DefaultPromise.await(DefaultPromise.java:212)
	at com.itcast.oio.DefaultPromiseTest.lambda$main$1(DefaultPromiseTest.java:36)
	at io.netty.util.concurrent.PromiseTask$RunnableAdapter.call(PromiseTask.java:38)
	at io.netty.util.concurrent.PromiseTask.run(PromiseTask.java:73)
	at io.netty.channel.DefaultEventLoop.run(DefaultEventLoop.java:54)
	at io.netty.util.concurrent.SingleThreadEventExecutor$5.run(SingleThreadEventExecutor.java:918)
	at io.netty.util.internal.ThreadExecutorMap$2.run(ThreadExecutorMap.java:74)
	at io.netty.util.concurrent.FastThreadLocalRunnable.run(FastThreadLocalRunnable.java:30)
	at java.lang.Thread.run(Thread.java:745)

3.4 Handler & Pipeline

ChannelHandler 用来处理 Channel 上的各种事件,分为入站、出站两种。所有 ChannelHandler 被连成一串,就是 Pipeline

  • 入站处理器通常是 ChannelInboundHandlerAdapter 的子类,主要用来读取客户端数据,写回结果
  • 出站处理器通常是 ChannelOutboundHandlerAdapter 的子类,主要对写回结果进行加工

打个比喻,每个 Channel 是一个产品的加工车间,Pipeline 是车间中的流水线,ChannelHandler 就是流水线上的各道工序,而后面要讲的 ByteBuf 是原材料,经过很多工序的加工:先经过一道道入站工序,再经过一道道出站工序最终变成产品

先搞清楚顺序,服务端

java
new ServerBootstrap()
    .group(new NioEventLoopGroup())
    .channel(NioServerSocketChannel.class)
    .childHandler(new ChannelInitializer<NioSocketChannel>() {
        protected void initChannel(NioSocketChannel ch) {
            ch.pipeline().addLast(new ChannelInboundHandlerAdapter(){
                @Override
                public void channelRead(ChannelHandlerContext ctx, Object msg) {
                    System.out.println(1);
                    ctx.fireChannelRead(msg); // 1
                }
            });
            ch.pipeline().addLast(new ChannelInboundHandlerAdapter(){
                @Override
                public void channelRead(ChannelHandlerContext ctx, Object msg) {
                    System.out.println(2);
                    ctx.fireChannelRead(msg); // 2
                }
            });
            ch.pipeline().addLast(new ChannelInboundHandlerAdapter(){
                @Override
                public void channelRead(ChannelHandlerContext ctx, Object msg) {
                    System.out.println(3);
                    ctx.channel().write(msg); // 3
                }
            });
            ch.pipeline().addLast(new ChannelOutboundHandlerAdapter(){
                @Override
                public void write(ChannelHandlerContext ctx, Object msg, 
                                  ChannelPromise promise) {
                    System.out.println(4);
                    ctx.write(msg, promise); // 4
                }
            });
            ch.pipeline().addLast(new ChannelOutboundHandlerAdapter(){
                @Override
                public void write(ChannelHandlerContext ctx, Object msg, 
                                  ChannelPromise promise) {
                    System.out.println(5);
                    ctx.write(msg, promise); // 5
                }
            });
            ch.pipeline().addLast(new ChannelOutboundHandlerAdapter(){
                @Override
                public void write(ChannelHandlerContext ctx, Object msg, 
                                  ChannelPromise promise) {
                    System.out.println(6);
                    ctx.write(msg, promise); // 6
                }
            });
        }
    })
    .bind(8080);

客户端

java
new Bootstrap()
    .group(new NioEventLoopGroup())
    .channel(NioSocketChannel.class)
    .handler(new ChannelInitializer<Channel>() {
        @Override
        protected void initChannel(Channel ch) {
            ch.pipeline().addLast(new StringEncoder());
        }
    })
    .connect("127.0.0.1", 8080)
    .addListener((ChannelFutureListener) future -> {
        future.channel().writeAndFlush("hello,world");
    });

服务器端打印:

1
2
3
6
5
4

可以看到,ChannelInboundHandlerAdapter 是按照 addLast 的顺序执行的,而 ChannelOutboundHandlerAdapter 是按照 addLast 的逆序执行的。ChannelPipeline 的实现是一个 ChannelHandlerContext(包装了 ChannelHandler) 组成的双向链表

  • 入站处理器中,ctx.fireChannelRead(msg) 是 调用下一个入站处理器
    • 如果注释掉 1 处代码,则仅会打印 1
    • 如果注释掉 2 处代码,则仅会打印 1 2
  • 3 处的 ctx.channel().write(msg) 会 从尾部开始触发 后续出站处理器的执行
    • 如果注释掉 3 处代码,则仅会打印 1 2 3
  • 类似的,出站处理器中,ctx.write(msg, promise) 的调用也会 触发上一个出站处理器
    • 如果注释掉 6 处代码,则仅会打印 1 2 3 6
  • ctx.channel().write(msg) vs ctx.write(msg)
    • 都是触发出站处理器的执行
    • ctx.channel().write(msg) 从尾部开始查找出站处理器
    • ctx.write(msg) 是从当前节点找上一个出站处理器
    • 3 处的 ctx.channel().write(msg) 如果改为 ctx.write(msg) 仅会打印 1 2 3,因为节点3 之前没有其它出站处理器了
    • 6 处的 ctx.write(msg, promise) 如果改为 ctx.channel().write(msg) 会打印 1 2 3 6 6 6... 因为 ctx.channel().write() 是从尾部开始查找,结果又是节点6 自己

图1 - 服务端 pipeline 触发的原始流程,图中数字代表了处理步骤的先后次序

3.5 ByteBuf

是对字节数据的封装

初始化方法
java
// 创建了一个默认的 ByteBuf(池化基于直接内存的 ByteBuf),初始容量是 10
ByteBuf buffer = ByteBufAllocator.DEFAULT.buffer(10);
log(buffer);
private static void log(ByteBuf buffer) {
    int length = buffer.readableBytes();
    int rows = length / 16 + (length % 15 == 0 ? 0 : 1) + 4;
    StringBuilder buf = new StringBuilder(rows * 80 * 2)
        .append("read index:").append(buffer.readerIndex())
        .append(" write index:").append(buffer.writerIndex())
        .append(" capacity:").append(buffer.capacity())
        .append(NEWLINE);
    appendPrettyHexDump(buf, buffer);
    System.out.println(buf.toString());
}


// 直接内存创建和销毁的代价昂贵,但读写性能高(少一次内存复制),适合配合池化功能一起用
// 直接内存对 GC 压力小,因为这部分内存不受 JVM 垃圾回收的管理,但也要注意及时主动释放
// 堆内存
ByteBuf buffer = ByteBufAllocator.DEFAULT.heapBuffer(10);
// 直接内存
ByteBuf buffer = ByteBufAllocator.DEFAULT.directBuffer(10);
池化与非池化

池化的最大意义在于可以重用 ByteBuf,优点有

  • 没有池化,则每次都得创建新的 ByteBuf 实例,这个操作对直接内存代价昂贵,就算是堆内存,也会增加 GC 压力
  • 有了池化,则可以重用池中 ByteBuf 实例,并且采用了与 jemalloc 类似的内存分配算法提升分配效率
  • 高并发时,池化功能更节约内存,减少内存溢出的可能

池化功能是否开启,可以通过下面的系统环境变量来设置

java
-Dio.netty.allocator.type={unpooled|pooled}
  • 4.1 以后,非 Android 平台默认启用池化实现,Android 平台启用非池化实现
  • 4.1 之前,池化功能还不成熟,默认是非池化实现
组成

ByteBuf 由四部分组成

最开始读写指针都在 0 位置

写入

方法列表,省略一些不重要的方法

方法签名含义备注
writeBoolean(boolean value)写入 boolean 值用一字节 01|00 代表 true|false
writeByte(int value)写入 byte 值
writeShort(int value)写入 short 值
writeInt(int value)写入 int 值Big Endian,即 0x250,写入后 00 00 02 50
writeIntLE(int value)写入 int 值Little Endian,即 0x250,写入后 50 02 00 00
writeLong(long value)写入 long 值
writeChar(int value)写入 char 值
writeFloat(float value)写入 float 值
writeDouble(double value)写入 double 值
writeBytes(ByteBuf src)写入 netty 的 ByteBuf
writeBytes(byte[] src)写入 byte[]
writeBytes(ByteBuffer src)写入 nio 的 ByteBuffer
int writeCharSequence(CharSequence sequence, Charset charset)写入字符串

注意

  • 这些方法的未指明返回值的,其返回值都是 ByteBuf,意味着可以链式调用
  • 网络传输,默认习惯是 Big Endian

先写入 4 个字节

java
buffer.writeBytes(new byte[]{1, 2, 3, 4});
log(buffer);

结果是

read index:0 write index:4 capacity:10
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 01 02 03 04                                     |....            |
+--------+-------------------------------------------------+----------------+

再写入一个 int 整数,也是 4 个字节

java
buffer.writeInt(5);
log(buffer);

结果是

read index:0 write index:8 capacity:10
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 01 02 03 04 00 00 00 05                         |........        |
+--------+-------------------------------------------------+----------------+

还有一类方法是 set 开头的一系列方法,也可以写入数据,但不会改变写指针位置

扩容

再写入一个 int 整数时,容量不够了(初始容量是 10),这时会引发扩容

java
buffer.writeInt(6);
log(buffer);

扩容规则是

  • 如何写入后数据大小未超过 512,则选择下一个 16 的整数倍,例如写入后大小为 12 ,则扩容后 capacity 是 16
  • 如果写入后数据大小超过 512,则选择下一个 2^n,例如写入后大小为 513,则扩容后 capacity 是 2^10=1024(2^9=512 已经不够了)
  • 扩容不能超过 max capacity 会报错

结果是

read index:0 write index:12 capacity:16
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 01 02 03 04 00 00 00 05 00 00 00 06             |............    |
+--------+-------------------------------------------------+----------------+
读取

例如读了 4 次,每次一个字节

java
System.out.println(buffer.readByte());
System.out.println(buffer.readByte());
System.out.println(buffer.readByte());
System.out.println(buffer.readByte());
log(buffer);

读过的内容,就属于废弃部分了,再读只能读那些尚未读取的部分

1
2
3
4
read index:4 write index:12 capacity:16
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 00 00 00 05 00 00 00 06                         |........        |
+--------+-------------------------------------------------+----------------+

如果需要重复读取 int 整数 5,怎么办?

可以在 read 前先做个标记 mark

java
buffer.markReaderIndex();
System.out.println(buffer.readInt());
log(buffer);

结果

5
read index:8 write index:12 capacity:16
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 00 00 00 06                                     |....            |
+--------+-------------------------------------------------+----------------+

这时要重复读取的话,重置到标记位置 reset

java
buffer.resetReaderIndex();
log(buffer);

这时

read index:4 write index:12 capacity:16
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 00 00 00 05 00 00 00 06                         |........        |
+--------+-------------------------------------------------+----------------+

还有种办法是采用 get 开头的一系列方法,这些方法不会改变 read index

retain & release

由于 Netty 中有堆外内存的 ByteBuf 实现,堆外内存最好是手动来释放,而不是等 GC 垃圾回收。

  • UnpooledHeapByteBuf 使用的是 JVM 内存,只需等 GC 回收内存即可
  • UnpooledDirectByteBuf 使用的就是直接内存了,需要特殊的方法来回收内存
  • PooledByteBuf 和它的子类使用了池化机制,需要更复杂的规则来回收内存

回收内存的源码实现,请关注下面方法的不同实现

protected abstract void deallocate()

Netty 这里采用了引用计数法来控制回收内存,每个 ByteBuf 都实现了 ReferenceCounted 接口

  • 每个 ByteBuf 对象的初始计数为 1
  • 调用 release 方法计数减 1,如果计数为 0,ByteBuf 内存被回收
  • 调用 retain 方法计数加 1,表示调用者没用完之前,其它 handler 即使调用了 release 也不会造成回收
  • 当计数为 0 时,底层内存会被回收,这时即使 ByteBuf 对象还在,其各个方法均无法正常使用

谁来负责 release 呢?

不是我们想象的(一般情况下)

java
ByteBuf buf = ...
try {
    ...
} finally {
    buf.release();
}

请思考,因为 pipeline 的存在,一般需要将 ByteBuf 传递给下一个 ChannelHandler,如果在 finally 中 release 了,就失去了传递性(当然,如果在这个 ChannelHandler 内这个 ByteBuf 已完成了它的使命,那么便无须再传递)

基本规则是,谁是最后使用者,谁负责 release,详细分析如下

  • 起点,对于 NIO 实现来讲,在 io.netty.channel.nio.AbstractNioByteChannel.NioByteUnsafe#read 方法中首次创建 ByteBuf 放入 pipeline(line 163 pipeline.fireChannelRead(byteBuf))
  • 入站 ByteBuf 处理原则
    • 对原始 ByteBuf 不做处理,调用 ctx.fireChannelRead(msg) 向后传递,这时无须 release
    • 将原始 ByteBuf 转换为其它类型的 Java 对象,这时 ByteBuf 就没用了,必须 release
    • 如果不调用 ctx.fireChannelRead(msg) 向后传递,那么也必须 release
    • 注意各种异常,如果 ByteBuf 没有成功传递到下一个 ChannelHandler,必须 release
    • 假设消息一直向后传,那么 TailContext 会负责释放未处理消息(原始的 ByteBuf)
  • 出站 ByteBuf 处理原则
    • 出站消息最终都会转为 ByteBuf 输出,一直向前传,由 HeadContext flush 后 release
  • 异常处理原则
    • 有时候不清楚 ByteBuf 被引用了多少次,但又必须彻底释放,可以循环调用 release 直到返回 true

TailContext 释放未处理消息逻辑

java
// io.netty.channel.DefaultChannelPipeline#onUnhandledInboundMessage(java.lang.Object)
protected void onUnhandledInboundMessage(Object msg) {
    try {
        logger.debug(
            "Discarded inbound message {} that reached at the tail of the pipeline. " +
            "Please check your pipeline configuration.", msg);
    } finally {
        ReferenceCountUtil.release(msg);
    }
}

具体代码

java
// io.netty.util.ReferenceCountUtil#release(java.lang.Object)
public static boolean release(Object msg) {
    if (msg instanceof ReferenceCounted) {
        return ((ReferenceCounted) msg).release();
    }
    return false;
}
slice

【零拷贝】的体现之一,对原始 ByteBuf 进行切片成多个 ByteBuf,切片后的 ByteBuf 并没有发生内存复制,还是使用原始 ByteBuf 的内存,切片后的 ByteBuf 维护独立的 read,write 指针

例,原始 ByteBuf 进行一些初始操作

java
ByteBuf origin = ByteBufAllocator.DEFAULT.buffer(10);
origin.writeBytes(new byte[]{1, 2, 3, 4});
origin.readByte();
System.out.println(ByteBufUtil.prettyHexDump(origin));

输出

         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 02 03 04                                        |...             |
+--------+-------------------------------------------------+----------------+

这时调用 slice 进行切片,无参 slice 是从原始 ByteBuf 的 read index 到 write index 之间的内容进行切片,切片后的 max capacity 被固定为这个区间的大小,因此不能追加 write

java
ByteBuf slice = origin.slice();
System.out.println(ByteBufUtil.prettyHexDump(slice));
// slice.writeByte(5); 如果执行,会报 IndexOutOfBoundsException 异常

输出

         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 02 03 04                                        |...             |
+--------+-------------------------------------------------+----------------+

如果原始 ByteBuf 再次读操作(又读了一个字节)

java
origin.readByte();
System.out.println(ByteBufUtil.prettyHexDump(origin));

输出

         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 03 04                                           |..              |
+--------+-------------------------------------------------+----------------+

这时的 slice 不受影响,因为它有独立的读写指针

java
System.out.println(ByteBufUtil.prettyHexDump(slice));

输出

         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 02 03 04                                        |...             |
+--------+-------------------------------------------------+----------------+

如果 slice 的内容发生了更改

java
slice.setByte(2, 5);
System.out.println(ByteBufUtil.prettyHexDump(slice));

输出

         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 02 03 05                                        |...             |
+--------+-------------------------------------------------+----------------+

这时,原始 ByteBuf 也会受影响,因为底层都是同一块内存

System.out.println(ByteBufUtil.prettyHexDump(origin));

输出

         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 03 05                                           |..              |
+--------+-------------------------------------------------+----------------+
duplicate

【零拷贝】的体现之一,就好比截取了原始 ByteBuf 所有内容,并且没有 max capacity 的限制,也是与原始 ByteBuf 使用同一块底层内存,只是读写指针是独立的

copy

会将底层内存数据进行深拷贝,因此无论读写,都与原始 ByteBuf 无关

CompositeByteBuf

【零拷贝】的体现之一,可以将多个 ByteBuf 合并为一个逻辑上的 ByteBuf,避免拷贝

有两个 ByteBuf 如下

java
ByteBuf buf1 = ByteBufAllocator.DEFAULT.buffer(5);
buf1.writeBytes(new byte[]{1, 2, 3, 4, 5});
ByteBuf buf2 = ByteBufAllocator.DEFAULT.buffer(5);
buf2.writeBytes(new byte[]{6, 7, 8, 9, 10});
System.out.println(ByteBufUtil.prettyHexDump(buf1));
System.out.println(ByteBufUtil.prettyHexDump(buf2));

输出

         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 01 02 03 04 05                                  |.....           |
+--------+-------------------------------------------------+----------------+
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 06 07 08 09 0a                                  |.....           |
+--------+-------------------------------------------------+----------------+

现在需要一个新的 ByteBuf,内容来自于刚才的 buf1 和 buf2,如何实现?

方法1:

java
ByteBuf buf3 = ByteBufAllocator.DEFAULT
    .buffer(buf1.readableBytes()+buf2.readableBytes());
buf3.writeBytes(buf1);
buf3.writeBytes(buf2);
System.out.println(ByteBufUtil.prettyHexDump(buf3));

结果

         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 01 02 03 04 05 06 07 08 09 0a                   |..........      |
+--------+-------------------------------------------------+----------------+

这种方法好不好?回答是不太好,因为进行了数据的内存复制操作

方法2:

java
CompositeByteBuf buf3 = ByteBufAllocator.DEFAULT.compositeBuffer();
// true 表示增加新的 ByteBuf 自动递增 write index, 否则 write index 会始终为 0
buf3.addComponents(true, buf1, buf2);

结果是一样的

         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 01 02 03 04 05 06 07 08 09 0a                   |..........      |
+--------+-------------------------------------------------+----------------+

CompositeByteBuf 是一个组合的 ByteBuf,它内部维护了一个 Component 数组,每个 Component 管理一个 ByteBuf,记录了这个 ByteBuf 相对于整体偏移量等信息,代表着整体中某一段的数据。

  • 优点,对外是一个虚拟视图,组合这些 ByteBuf 不会产生内存复制
  • 缺点,复杂了很多,多次操作会带来性能的损耗
Unpooled

Unpooled 是一个工具类,类如其名,提供了非池化的 ByteBuf 创建、组合、复制等操作

这里仅介绍其跟【零拷贝】相关的 wrappedBuffer 方法,可以用来包装 ByteBuf

java
ByteBuf buf1 = ByteBufAllocator.DEFAULT.buffer(5);
buf1.writeBytes(new byte[]{1, 2, 3, 4, 5});
ByteBuf buf2 = ByteBufAllocator.DEFAULT.buffer(5);
buf2.writeBytes(new byte[]{6, 7, 8, 9, 10});

// 当包装 ByteBuf 个数超过一个时, 底层使用了 CompositeByteBuf
ByteBuf buf3 = Unpooled.wrappedBuffer(buf1, buf2);
System.out.println(ByteBufUtil.prettyHexDump(buf3));

输出

         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 01 02 03 04 05 06 07 08 09 0a                   |..........      |
+--------+-------------------------------------------------+----------------+

也可以用来包装普通字节数组,底层也不会有拷贝操作

java
ByteBuf buf4 = Unpooled.wrappedBuffer(new byte[]{1, 2, 3}, new byte[]{4, 5, 6});
System.out.println(buf4.getClass());
System.out.println(ByteBufUtil.prettyHexDump(buf4));

输出

class io.netty.buffer.CompositeByteBuf
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 01 02 03 04 05 06                               |......          |
+--------+-------------------------------------------------+----------------+

ByteBuf 优势

  • 池化 - 可以重用池中 ByteBuf 实例,更节约内存,减少内存溢出的可能
  • 读写指针分离,不需要像 ByteBuffer 一样切换读写模式
  • 可以自动扩容
  • 支持链式调用,使用更流畅
  • 很多地方体现零拷贝,例如 slice、duplicate、CompositeByteBuf

四、核心知识

4.1 粘包与半包

粘包

  • 现象,发送 abc def,接收 abcdef
  • 原因
    • 应用层:接收方 ByteBuf 设置太大(Netty 默认 1024)
    • 滑动窗口:假设发送方 256 bytes 表示一个完整报文,但由于接收方处理不及时且窗口大小足够大,这 256 bytes 字节就会缓冲在接收方的滑动窗口中,当滑动窗口中缓冲了多个报文就会粘包
    • Nagle 算法:会造成粘包

半包

  • 现象,发送 abcdef,接收 abc def
  • 原因
    • 应用层:接收方 ByteBuf 小于实际发送数据量
    • 滑动窗口:假设接收方的窗口只剩了 128 bytes,发送方的报文大小是 256 bytes,这时放不下了,只能先发送前 128 bytes,等待 ack 后才能发送剩余部分,这就造成了半包
    • MSS 限制:当发送的数据超过 MSS 限制后,会将数据切分发送,就会造成半包

本质是因为 TCP 是流式协议,消息无边界

解决方案

  1. 短链接,发一个包建立一次连接,这样连接建立到连接断开之间就是消息的边界,缺点效率太低
  2. 每一条消息采用固定长度,缺点浪费空间
  3. 每一条消息采用分隔符,例如 \n,缺点需要转义
  4. 每一条消息分为 head 和 body,head 中包含 body 的长度

方法1,短链接

以解决粘包为例

java
public class HelloWorldClient {
    static final Logger log = LoggerFactory.getLogger(HelloWorldClient.class);

    public static void main(String[] args) {
        // 分 10 次发送
        for (int i = 0; i < 10; i++) {
            send();
        }
    }

    private static void send() {
        NioEventLoopGroup worker = new NioEventLoopGroup();
        try {
            Bootstrap bootstrap = new Bootstrap();
            bootstrap.channel(NioSocketChannel.class);
            bootstrap.group(worker);
            bootstrap.handler(new ChannelInitializer<SocketChannel>() {
                @Override
                protected void initChannel(SocketChannel ch) throws Exception {
                    log.debug("conneted...");
                    ch.pipeline().addLast(new LoggingHandler(LogLevel.DEBUG));
                    ch.pipeline().addLast(new ChannelInboundHandlerAdapter() {
                        @Override
                        public void channelActive(ChannelHandlerContext ctx) throws Exception {
                            log.debug("sending...");
                            ByteBuf buffer = ctx.alloc().buffer();
                            buffer.writeBytes(new byte[]{0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15});
                            ctx.writeAndFlush(buffer);
                            // 发完即关
                            ctx.close();
                        }
                    });
                }
            });
            ChannelFuture channelFuture = bootstrap.connect("localhost", 8080).sync();
            channelFuture.channel().closeFuture().sync();

        } catch (InterruptedException e) {
            log.error("client error", e);
        } finally {
            worker.shutdownGracefully();
        }
    }
}

输出,略

半包用这种办法还是不好解决,因为接收方的缓冲区大小是有限的

方法2,固定长度

让所有数据包长度固定(假设长度为 8 字节),服务器端加入

java
ch.pipeline().addLast(new FixedLengthFrameDecoder(8));

客户端测试代码,注意, 采用这种方法后,客户端什么时候 flush 都可以

java
public class HelloWorldClient {
    static final Logger log = LoggerFactory.getLogger(HelloWorldClient.class);

    public static void main(String[] args) {
        NioEventLoopGroup worker = new NioEventLoopGroup();
        try {
            Bootstrap bootstrap = new Bootstrap();
            bootstrap.channel(NioSocketChannel.class);
            bootstrap.group(worker);
            bootstrap.handler(new ChannelInitializer<SocketChannel>() {
                @Override
                protected void initChannel(SocketChannel ch) throws Exception {
                    log.debug("connetted...");
                    ch.pipeline().addLast(new LoggingHandler(LogLevel.DEBUG));
                    ch.pipeline().addLast(new ChannelInboundHandlerAdapter() {
                        @Override
                        public void channelActive(ChannelHandlerContext ctx) throws Exception {
                            log.debug("sending...");
                            // 发送内容随机的数据包
                            Random r = new Random();
                            char c = 'a';
                            ByteBuf buffer = ctx.alloc().buffer();
                            for (int i = 0; i < 10; i++) {
                                byte[] bytes = new byte[8];
                                for (int j = 0; j < r.nextInt(8); j++) {
                                    bytes[j] = (byte) c;
                                }
                                c++;
                                buffer.writeBytes(bytes);
                            }
                            ctx.writeAndFlush(buffer);
                        }
                    });
                }
            });
            ChannelFuture channelFuture = bootstrap.connect("192.168.0.103", 9090).sync();
            channelFuture.channel().closeFuture().sync();

        } catch (InterruptedException e) {
            log.error("client error", e);
        } finally {
            worker.shutdownGracefully();
        }
    }
}

客户端输出


         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 61 61 61 61 00 00 00 00 62 00 00 00 00 00 00 00 |aaaa....b.......|
|00000010| 63 63 00 00 00 00 00 00 64 00 00 00 00 00 00 00 |cc......d.......|
|00000020| 00 00 00 00 00 00 00 00 66 66 66 66 00 00 00 00 |........ffff....|
|00000030| 67 67 67 00 00 00 00 00 68 00 00 00 00 00 00 00 |ggg.....h.......|
|00000040| 69 69 69 69 69 00 00 00 6a 6a 6a 6a 00 00 00 00 |iiiii...jjjj....|
+--------+-------------------------------------------------+----------------+

服务端输出


         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 61 61 61 61 00 00 00 00                         |aaaa....        |
+--------+-------------------------------------------------+----------------+
12:07:00 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0xd739f137, L:/192.168.0.103:9090 - R:/192.168.0.103:53155] READ: 8B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 62 00 00 00 00 00 00 00                         |b.......        |
+--------+-------------------------------------------------+----------------+
12:07:00 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0xd739f137, L:/192.168.0.103:9090 - R:/192.168.0.103:53155] READ: 8B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 63 63 00 00 00 00 00 00                         |cc......        |
+--------+-------------------------------------------------+----------------+
12:07:00 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0xd739f137, L:/192.168.0.103:9090 - R:/192.168.0.103:53155] READ: 8B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 64 00 00 00 00 00 00 00                         |d.......        |
+--------+-------------------------------------------------+----------------+
12:07:00 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0xd739f137, L:/192.168.0.103:9090 - R:/192.168.0.103:53155] READ: 8B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 00 00 00 00 00 00 00 00                         |........        |
+--------+-------------------------------------------------+----------------+
12:07:00 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0xd739f137, L:/192.168.0.103:9090 - R:/192.168.0.103:53155] READ: 8B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 66 66 66 66 00 00 00 00                         |ffff....        |
+--------+-------------------------------------------------+----------------+
12:07:00 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0xd739f137, L:/192.168.0.103:9090 - R:/192.168.0.103:53155] READ: 8B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 67 67 67 00 00 00 00 00                         |ggg.....        |
+--------+-------------------------------------------------+----------------+
12:07:00 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0xd739f137, L:/192.168.0.103:9090 - R:/192.168.0.103:53155] READ: 8B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 68 00 00 00 00 00 00 00                         |h.......        |
+--------+-------------------------------------------------+----------------+
12:07:00 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0xd739f137, L:/192.168.0.103:9090 - R:/192.168.0.103:53155] READ: 8B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 69 69 69 69 69 00 00 00                         |iiiii...        |
+--------+-------------------------------------------------+----------------+
12:07:00 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0xd739f137, L:/192.168.0.103:9090 - R:/192.168.0.103:53155] READ: 8B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 6a 6a 6a 6a 00 00 00 00                         |jjjj....        |
+--------+-------------------------------------------------+----------------+

缺点是,数据包的大小不好把握

  • 长度定的太大,浪费
  • 长度定的太小,对某些数据包又显得不够

方法3,固定分隔符

服务端加入,默认以 \n 或 \r\n 作为分隔符,如果超出指定长度仍未出现分隔符,则抛出异常

java
ch.pipeline().addLast(new LineBasedFrameDecoder(1024));

客户端在每条消息之后,加入 \n 分隔符

java
public class HelloWorldClient {
    static final Logger log = LoggerFactory.getLogger(HelloWorldClient.class);

    public static void main(String[] args) {
        NioEventLoopGroup worker = new NioEventLoopGroup();
        try {
            Bootstrap bootstrap = new Bootstrap();
            bootstrap.channel(NioSocketChannel.class);
            bootstrap.group(worker);
            bootstrap.handler(new ChannelInitializer<SocketChannel>() {
                @Override
                protected void initChannel(SocketChannel ch) throws Exception {
                    log.debug("connetted...");
                    ch.pipeline().addLast(new LoggingHandler(LogLevel.DEBUG));
                    ch.pipeline().addLast(new ChannelInboundHandlerAdapter() {
                        @Override
                        public void channelActive(ChannelHandlerContext ctx) throws Exception {
                            log.debug("sending...");
                            Random r = new Random();
                            char c = 'a';
                            ByteBuf buffer = ctx.alloc().buffer();
                            for (int i = 0; i < 10; i++) {
                                for (int j = 1; j <= r.nextInt(16)+1; j++) {
                                    buffer.writeByte((byte) c);
                                }
                                buffer.writeByte(10);
                                c++;
                            }
                            ctx.writeAndFlush(buffer);
                        }
                    });
                }
            });
            ChannelFuture channelFuture = bootstrap.connect("192.168.0.103", 9090).sync();
            channelFuture.channel().closeFuture().sync();

        } catch (InterruptedException e) {
            log.error("client error", e);
        } finally {
            worker.shutdownGracefully();
        }
    }
}

客户端输出


         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 61 0a 62 62 62 0a 63 63 63 0a 64 64 0a 65 65 65 |a.bbb.ccc.dd.eee|
|00000010| 65 65 65 65 65 65 65 0a 66 66 0a 67 67 67 67 67 |eeeeeee.ff.ggggg|
|00000020| 67 67 0a 68 68 68 68 0a 69 69 69 69 69 69 69 0a |gg.hhhh.iiiiiii.|
|00000030| 6a 6a 6a 6a 6a 6a 6a 6a 6a 6a 6a 0a             |jjjjjjjjjjj.    |
+--------+-------------------------------------------------+----------------+

服务端输出


         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 61                                              |a               |
+--------+-------------------------------------------------+----------------+
14:08:18 [DEBUG] [nioEventLoopGroup-3-5] i.n.h.l.LoggingHandler - [id: 0xa4b3be43, L:/192.168.0.103:9090 - R:/192.168.0.103:63641] READ: 3B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 62 62 62                                        |bbb             |
+--------+-------------------------------------------------+----------------+
14:08:18 [DEBUG] [nioEventLoopGroup-3-5] i.n.h.l.LoggingHandler - [id: 0xa4b3be43, L:/192.168.0.103:9090 - R:/192.168.0.103:63641] READ: 3B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 63 63 63                                        |ccc             |
+--------+-------------------------------------------------+----------------+
14:08:18 [DEBUG] [nioEventLoopGroup-3-5] i.n.h.l.LoggingHandler - [id: 0xa4b3be43, L:/192.168.0.103:9090 - R:/192.168.0.103:63641] READ: 2B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 64 64                                           |dd              |
+--------+-------------------------------------------------+----------------+
14:08:18 [DEBUG] [nioEventLoopGroup-3-5] i.n.h.l.LoggingHandler - [id: 0xa4b3be43, L:/192.168.0.103:9090 - R:/192.168.0.103:63641] READ: 10B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 65 65 65 65 65 65 65 65 65 65                   |eeeeeeeeee      |
+--------+-------------------------------------------------+----------------+
14:08:18 [DEBUG] [nioEventLoopGroup-3-5] i.n.h.l.LoggingHandler - [id: 0xa4b3be43, L:/192.168.0.103:9090 - R:/192.168.0.103:63641] READ: 2B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 66 66                                           |ff              |
+--------+-------------------------------------------------+----------------+
14:08:18 [DEBUG] [nioEventLoopGroup-3-5] i.n.h.l.LoggingHandler - [id: 0xa4b3be43, L:/192.168.0.103:9090 - R:/192.168.0.103:63641] READ: 7B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 67 67 67 67 67 67 67                            |ggggggg         |
+--------+-------------------------------------------------+----------------+
14:08:18 [DEBUG] [nioEventLoopGroup-3-5] i.n.h.l.LoggingHandler - [id: 0xa4b3be43, L:/192.168.0.103:9090 - R:/192.168.0.103:63641] READ: 4B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 68 68 68 68                                     |hhhh            |
+--------+-------------------------------------------------+----------------+
14:08:18 [DEBUG] [nioEventLoopGroup-3-5] i.n.h.l.LoggingHandler - [id: 0xa4b3be43, L:/192.168.0.103:9090 - R:/192.168.0.103:63641] READ: 7B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 69 69 69 69 69 69 69                            |iiiiiii         |
+--------+-------------------------------------------------+----------------+
14:08:18 [DEBUG] [nioEventLoopGroup-3-5] i.n.h.l.LoggingHandler - [id: 0xa4b3be43, L:/192.168.0.103:9090 - R:/192.168.0.103:63641] READ: 11B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 6a 6a 6a 6a 6a 6a 6a 6a 6a 6a 6a                |jjjjjjjjjjj     |
+--------+-------------------------------------------------+----------------+
14:08:18 [DEBUG] [nioEventLoopGroup-3-5] i.n.h.l.LoggingHandler - [id: 0xa4b3be43, L:/192.168.0.103:9090 - R:/192.168.0.103:63641] READ COMPLETE

缺点,处理字符数据比较合适,但如果内容本身包含了分隔符(字节数据常常会有此情况),那么就会解析错误

方法4,预设长度

在发送消息前,先约定用定长字节表示接下来数据的长度

java
// 最大长度,长度偏移,长度占用字节,长度调整,剥离字节数
ch.pipeline().addLast(new LengthFieldBasedFrameDecoder(1024, 0, 1, 0, 1));

客户端代码

java
public class HelloWorldClient {
    static final Logger log = LoggerFactory.getLogger(HelloWorldClient.class);

    public static void main(String[] args) {
        NioEventLoopGroup worker = new NioEventLoopGroup();
        try {
            Bootstrap bootstrap = new Bootstrap();
            bootstrap.channel(NioSocketChannel.class);
            bootstrap.group(worker);
            bootstrap.handler(new ChannelInitializer<SocketChannel>() {
                @Override
                protected void initChannel(SocketChannel ch) throws Exception {
                    log.debug("connetted...");
                    ch.pipeline().addLast(new LoggingHandler(LogLevel.DEBUG));
                    ch.pipeline().addLast(new ChannelInboundHandlerAdapter() {
                        @Override
                        public void channelActive(ChannelHandlerContext ctx) throws Exception {
                            log.debug("sending...");
                            Random r = new Random();
                            char c = 'a';
                            ByteBuf buffer = ctx.alloc().buffer();
                            for (int i = 0; i < 10; i++) {
                                byte length = (byte) (r.nextInt(16) + 1);
                                // 先写入长度
                                buffer.writeByte(length);
                                // 再
                                for (int j = 1; j <= length; j++) {
                                    buffer.writeByte((byte) c);
                                }
                                c++;
                            }
                            ctx.writeAndFlush(buffer);
                        }
                    });
                }
            });
            ChannelFuture channelFuture = bootstrap.connect("192.168.0.103", 9090).sync();
            channelFuture.channel().closeFuture().sync();

        } catch (InterruptedException e) {
            log.error("client error", e);
        } finally {
            worker.shutdownGracefully();
        }
    }
}

客户端输出


         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 09 61 61 61 61 61 61 61 61 61 09 62 62 62 62 62 |.aaaaaaaaa.bbbbb|
|00000010| 62 62 62 62 06 63 63 63 63 63 63 08 64 64 64 64 |bbbb.cccccc.dddd|
|00000020| 64 64 64 64 0f 65 65 65 65 65 65 65 65 65 65 65 |dddd.eeeeeeeeeee|
|00000030| 65 65 65 65 0d 66 66 66 66 66 66 66 66 66 66 66 |eeee.fffffffffff|
|00000040| 66 66 02 67 67 02 68 68 0e 69 69 69 69 69 69 69 |ff.gg.hh.iiiiiii|
|00000050| 69 69 69 69 69 69 69 09 6a 6a 6a 6a 6a 6a 6a 6a |iiiiiii.jjjjjjjj|
|00000060| 6a                                              |j               |
+--------+-------------------------------------------------+----------------+
14:37:10 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0xf0f347b8, L:/192.168.0.103:49979 - R:/192.168.0.103:9090] FLUSH

服务端输出


         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 61 61 61 61 61 61 61 61 61                      |aaaaaaaaa       |
+--------+-------------------------------------------------+----------------+
14:37:10 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x744f2b47, L:/192.168.0.103:9090 - R:/192.168.0.103:49979] READ: 9B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 62 62 62 62 62 62 62 62 62                      |bbbbbbbbb       |
+--------+-------------------------------------------------+----------------+
14:37:10 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x744f2b47, L:/192.168.0.103:9090 - R:/192.168.0.103:49979] READ: 6B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 63 63 63 63 63 63                               |cccccc          |
+--------+-------------------------------------------------+----------------+
14:37:10 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x744f2b47, L:/192.168.0.103:9090 - R:/192.168.0.103:49979] READ: 8B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 64 64 64 64 64 64 64 64                         |dddddddd        |
+--------+-------------------------------------------------+----------------+
14:37:10 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x744f2b47, L:/192.168.0.103:9090 - R:/192.168.0.103:49979] READ: 15B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 65 65 65 65 65 65 65 65 65 65 65 65 65 65 65    |eeeeeeeeeeeeeee |
+--------+-------------------------------------------------+----------------+
14:37:10 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x744f2b47, L:/192.168.0.103:9090 - R:/192.168.0.103:49979] READ: 13B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 66 66 66 66 66 66 66 66 66 66 66 66 66          |fffffffffffff   |
+--------+-------------------------------------------------+----------------+
14:37:10 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x744f2b47, L:/192.168.0.103:9090 - R:/192.168.0.103:49979] READ: 2B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 67 67                                           |gg              |
+--------+-------------------------------------------------+----------------+
14:37:10 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x744f2b47, L:/192.168.0.103:9090 - R:/192.168.0.103:49979] READ: 2B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 68 68                                           |hh              |
+--------+-------------------------------------------------+----------------+
14:37:10 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x744f2b47, L:/192.168.0.103:9090 - R:/192.168.0.103:49979] READ: 14B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 69 69 69 69 69 69 69 69 69 69 69 69 69 69       |iiiiiiiiiiiiii  |
+--------+-------------------------------------------------+----------------+
14:37:10 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x744f2b47, L:/192.168.0.103:9090 - R:/192.168.0.103:49979] READ: 9B
         +-------------------------------------------------+
         |  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f |
+--------+-------------------------------------------------+----------------+
|00000000| 6a 6a 6a 6a 6a 6a 6a 6a 6a                      |jjjjjjjjj       |
+--------+-------------------------------------------------+----------------+
14:37:10 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x744f2b47, L:/192.168.0.103:9090 - R:/192.168.0.103:49979] READ COMPLETE

4.2 协议设计与解析

TCP/IP 中消息传输基于流的方式,没有边界。

协议的目的就是划定消息的边界,制定通信双方要共同遵守的通信规则

设计协议其实就是给网络传输的信息加上“标点符号”。但通过分隔符来断句不是很好,因为分隔符本身如果用于传输,那么必须加以区分。因此,下面一种协议较为常用

定长字节表示内容长度 + 实际内容

例如,假设一个中文字符长度为 3,按照上述协议的规则,发送信息方式如下,就不会被接收方弄错意思了

0f下雨天留客06天留09我不留

redis 协议举例

java
NioEventLoopGroup worker = new NioEventLoopGroup();
byte[] LINE = {13, 10};
try {
    Bootstrap bootstrap = new Bootstrap();
    bootstrap.channel(NioSocketChannel.class);
    bootstrap.group(worker);
    bootstrap.handler(new ChannelInitializer<SocketChannel>() {
        @Override
        protected void initChannel(SocketChannel ch) {
            ch.pipeline().addLast(new LoggingHandler());
            ch.pipeline().addLast(new ChannelInboundHandlerAdapter() {
                // 会在连接 channel 建立成功后,会触发 active 事件
                @Override
                public void channelActive(ChannelHandlerContext ctx) {
                    set(ctx);
                    get(ctx);
                }
                private void get(ChannelHandlerContext ctx) {
                    ByteBuf buf = ctx.alloc().buffer();
                    buf.writeBytes("*2".getBytes());
                    buf.writeBytes(LINE);
                    buf.writeBytes("$3".getBytes());
                    buf.writeBytes(LINE);
                    buf.writeBytes("get".getBytes());
                    buf.writeBytes(LINE);
                    buf.writeBytes("$3".getBytes());
                    buf.writeBytes(LINE);
                    buf.writeBytes("aaa".getBytes());
                    buf.writeBytes(LINE);
                    ctx.writeAndFlush(buf);
                }
                private void set(ChannelHandlerContext ctx) {
                    ByteBuf buf = ctx.alloc().buffer();
                    buf.writeBytes("*3".getBytes());
                    buf.writeBytes(LINE);
                    buf.writeBytes("$3".getBytes());
                    buf.writeBytes(LINE);
                    buf.writeBytes("set".getBytes());
                    buf.writeBytes(LINE);
                    buf.writeBytes("$3".getBytes());
                    buf.writeBytes(LINE);
                    buf.writeBytes("aaa".getBytes());
                    buf.writeBytes(LINE);
                    buf.writeBytes("$3".getBytes());
                    buf.writeBytes(LINE);
                    buf.writeBytes("bbb".getBytes());
                    buf.writeBytes(LINE);
                    ctx.writeAndFlush(buf);
                }

                @Override
                public void channelRead(ChannelHandlerContext ctx, Object msg) throws Exception {
                    ByteBuf buf = (ByteBuf) msg;
                    System.out.println(buf.toString(Charset.defaultCharset()));
                }
            });
        }
    });
    ChannelFuture channelFuture = bootstrap.connect("localhost", 6379).sync();
    channelFuture.channel().closeFuture().sync();
} catch (InterruptedException e) {
    log.error("client error", e);
} finally {
    worker.shutdownGracefully();
}

http 协议举例

java
NioEventLoopGroup boss = new NioEventLoopGroup();
NioEventLoopGroup worker = new NioEventLoopGroup();
try {
    ServerBootstrap serverBootstrap = new ServerBootstrap();
    serverBootstrap.channel(NioServerSocketChannel.class);
    serverBootstrap.group(boss, worker);
    serverBootstrap.childHandler(new ChannelInitializer<SocketChannel>() {
        @Override
        protected void initChannel(SocketChannel ch) throws Exception {
            ch.pipeline().addLast(new LoggingHandler(LogLevel.DEBUG));
            ch.pipeline().addLast(new HttpServerCodec());
            ch.pipeline().addLast(new SimpleChannelInboundHandler<HttpRequest>() {
                @Override
                protected void channelRead0(ChannelHandlerContext ctx, HttpRequest msg) throws Exception {
                    // 获取请求
                    log.debug(msg.uri());

                    // 返回响应
                    DefaultFullHttpResponse response =
                            new DefaultFullHttpResponse(msg.protocolVersion(), HttpResponseStatus.OK);

                    byte[] bytes = "<h1>Hello, world!</h1>".getBytes();

                    response.headers().setInt(CONTENT_LENGTH, bytes.length);
                    response.content().writeBytes(bytes);

                    // 写回响应
                    ctx.writeAndFlush(response);
                }
            });
            /*ch.pipeline().addLast(new ChannelInboundHandlerAdapter() {
                @Override
                public void channelRead(ChannelHandlerContext ctx, Object msg) throws Exception {
                    log.debug("{}", msg.getClass());

                    if (msg instanceof HttpRequest) { // 请求行,请求头

                    } else if (msg instanceof HttpContent) { //请求体

                    }
                }
            });*/
        }
    });
    ChannelFuture channelFuture = serverBootstrap.bind(8080).sync();
    channelFuture.channel().closeFuture().sync();
} catch (InterruptedException e) {
    log.error("server error", e);
} finally {
    boss.shutdownGracefully();
    worker.shutdownGracefully();
}

自定义协议要素

  • 魔数,用来在第一时间判定是否是无效数据包
  • 版本号,可以支持协议的升级
  • 序列化算法,消息正文到底采用哪种序列化反序列化方式,可以由此扩展,例如:json、protobuf、hessian、jdk
  • 指令类型,是登录、注册、单聊、群聊... 跟业务相关
  • 请求序号,为了双工通信,提供异步能力
  • 正文长度
  • 消息正文
编解码器

根据上面的要素,设计一个登录请求消息和登录响应消息,并使用 Netty 完成收发

java
@Slf4j
public class MessageCodec extends ByteToMessageCodec<Message> {

    @Override
    protected void encode(ChannelHandlerContext ctx, Message msg, ByteBuf out) throws Exception {
        // 1. 4 字节的魔数
        out.writeBytes(new byte[]{1, 2, 3, 4});
        // 2. 1 字节的版本,
        out.writeByte(1);
        // 3. 1 字节的序列化方式 jdk 0 , json 1
        out.writeByte(0);
        // 4. 1 字节的指令类型
        out.writeByte(msg.getMessageType());
        // 5. 4 个字节
        out.writeInt(msg.getSequenceId());
        // 无意义,对齐填充
        out.writeByte(0xff);
        // 6. 获取内容的字节数组
        ByteArrayOutputStream bos = new ByteArrayOutputStream();
        ObjectOutputStream oos = new ObjectOutputStream(bos);
        oos.writeObject(msg);
        byte[] bytes = bos.toByteArray();
        // 7. 长度
        out.writeInt(bytes.length);
        // 8. 写入内容
        out.writeBytes(bytes);
    }

    @Override
    protected void decode(ChannelHandlerContext ctx, ByteBuf in, List<Object> out) throws Exception {
        int magicNum = in.readInt();
        byte version = in.readByte();
        byte serializerType = in.readByte();
        byte messageType = in.readByte();
        int sequenceId = in.readInt();
        in.readByte();
        int length = in.readInt();
        byte[] bytes = new byte[length];
        in.readBytes(bytes, 0, length);
        ObjectInputStream ois = new ObjectInputStream(new ByteArrayInputStream(bytes));
        Message message = (Message) ois.readObject();
        log.debug("{}, {}, {}, {}, {}, {}", magicNum, version, serializerType, messageType, sequenceId, length);
        log.debug("{}", message);
        out.add(message);
    }
}
测试
java
EmbeddedChannel channel = new EmbeddedChannel(
    new LoggingHandler(),
    new LengthFieldBasedFrameDecoder(
        1024, 12, 4, 0, 0),
    new MessageCodec()
);
// encode
LoginRequestMessage message = new LoginRequestMessage("zhangsan", "123", "张三");
//        channel.writeOutbound(message);
// decode
ByteBuf buf = ByteBufAllocator.DEFAULT.buffer();
new MessageCodec().encode(null, message, buf);

ByteBuf s1 = buf.slice(0, 100);
ByteBuf s2 = buf.slice(100, buf.readableBytes() - 100);
s1.retain(); // 引用计数 2
channel.writeInbound(s1); // release 1
channel.writeInbound(s2);
解读

💡 什么时候可以加 @Sharable

  • 当 handler 不保存状态时,就可以安全地在多线程下被共享
  • 但要注意对于编解码器类,不能继承 ByteToMessageCodec 或 CombinedChannelDuplexHandler 父类,他们的构造方法对 @Sharable 有限制
  • 如果能确保编解码器不会保存状态,可以继承 MessageToMessageCodec 父类
java
@Slf4j
@ChannelHandler.Sharable
/**
 * 必须和 LengthFieldBasedFrameDecoder 一起使用,确保接到的 ByteBuf 消息是完整的
 */
public class MessageCodecSharable extends MessageToMessageCodec<ByteBuf, Message> {
    @Override
    protected void encode(ChannelHandlerContext ctx, Message msg, List<Object> outList) throws Exception {
        ByteBuf out = ctx.alloc().buffer();
        // 1. 4 字节的魔数
        out.writeBytes(new byte[]{1, 2, 3, 4});
        // 2. 1 字节的版本,
        out.writeByte(1);
        // 3. 1 字节的序列化方式 jdk 0 , json 1
        out.writeByte(0);
        // 4. 1 字节的指令类型
        out.writeByte(msg.getMessageType());
        // 5. 4 个字节
        out.writeInt(msg.getSequenceId());
        // 无意义,对齐填充
        out.writeByte(0xff);
        // 6. 获取内容的字节数组
        ByteArrayOutputStream bos = new ByteArrayOutputStream();
        ObjectOutputStream oos = new ObjectOutputStream(bos);
        oos.writeObject(msg);
        byte[] bytes = bos.toByteArray();
        // 7. 长度
        out.writeInt(bytes.length);
        // 8. 写入内容
        out.writeBytes(bytes);
        outList.add(out);
    }

    @Override
    protected void decode(ChannelHandlerContext ctx, ByteBuf in, List<Object> out) throws Exception {
        int magicNum = in.readInt();
        byte version = in.readByte();
        byte serializerType = in.readByte();
        byte messageType = in.readByte();
        int sequenceId = in.readInt();
        in.readByte();
        int length = in.readInt();
        byte[] bytes = new byte[length];
        in.readBytes(bytes, 0, length);
        ObjectInputStream ois = new ObjectInputStream(new ByteArrayInputStream(bytes));
        Message message = (Message) ois.readObject();
        log.debug("{}, {}, {}, {}, {}, {}", magicNum, version, serializerType, messageType, sequenceId, length);
        log.debug("{}", message);
        out.add(message);
    }
}

4.3 空闲检测

连接假死

原因

  • 网络设备出现故障,例如网卡,机房等,底层的 TCP 连接已经断开了,但应用程序没有感知到,仍然占用着资源。
  • 公网网络不稳定,出现丢包。如果连续出现丢包,这时现象就是客户端数据发不出去,服务端也一直收不到数据,就这么一直耗着
  • 应用程序线程阻塞,无法进行数据读写

问题

  • 假死的连接占用的资源不能自动释放
  • 向假死的连接发送数据,得到的反馈是发送超时

服务器端解决

  • 怎么判断客户端连接是否假死呢?如果能收到客户端数据,说明没有假死。因此策略就可以定为,每隔一段时间就检查这段时间内是否接收到客户端数据,没有就可以判定为连接假死

服务端判断连接假死

java
// 用来判断是不是 读空闲时间过长,或 写空闲时间过长
// 5s 内如果没有收到 channel 的数据,会触发一个 IdleState#READER_IDLE 事件
ch.pipeline().addLast(new IdleStateHandler(5, 0, 0));
// ChannelDuplexHandler 可以同时作为入站和出站处理器
ch.pipeline().addLast(new ChannelDuplexHandler() {
    // 用来触发特殊事件
    @Override
    public void userEventTriggered(ChannelHandlerContext ctx, Object evt) throws Exception{
        IdleStateEvent event = (IdleStateEvent) evt;
        // 触发了读空闲事件
        if (event.state() == IdleState.READER_IDLE) {
            log.debug("已经 5s 没有读到数据了");
            ctx.channel().close();
        }
    }
});

客户端定时心跳

客户端可以定时向服务器端发送数据,只要这个时间间隔小于服务器定义的空闲检测的时间间隔,那么就能防止前面提到的误判,客户端可以定义如下心跳处理器

java
// 用来判断是不是 读空闲时间过长,或 写空闲时间过长
// 3s 内如果没有向服务器写数据,会触发一个 IdleState#WRITER_IDLE 事件
ch.pipeline().addLast(new IdleStateHandler(0, 3, 0));
// ChannelDuplexHandler 可以同时作为入站和出站处理器
ch.pipeline().addLast(new ChannelDuplexHandler() {
    // 用来触发特殊事件
    @Override
    public void userEventTriggered(ChannelHandlerContext ctx, Object evt) throws Exception{
        IdleStateEvent event = (IdleStateEvent) evt;
        // 触发了写空闲事件
        if (event.state() == IdleState.WRITER_IDLE) {
            //                                log.debug("3s 没有写数据了,发送一个心跳包");
            ctx.writeAndFlush(new PingMessage());
        }
    }
});

五、优化

5.1 拓展序列化算法

序列化,反序列化主要用在消息正文的转换上

  • 序列化时,需要将 Java 对象变为要传输的数据(可以是 byte[],或 json 等,最终都需要变成 byte[])
  • 反序列化时,需要将传入的正文数据还原成 Java 对象,便于处理

目前的代码仅支持 Java 自带的序列化,反序列化机制,核心代码如下

java
// 反序列化
byte[] body = new byte[bodyLength];
byteByf.readBytes(body);
ObjectInputStream in = new ObjectInputStream(new ByteArrayInputStream(body));
Message message = (Message) in.readObject();
message.setSequenceId(sequenceId);

// 序列化
ByteArrayOutputStream out = new ByteArrayOutputStream();
new ObjectOutputStream(out).writeObject(message);
byte[] bytes = out.toByteArray();

为了支持更多序列化算法,抽象一个 Serializer 接口

java
public interface Serializer {

    // 反序列化方法
    <T> T deserialize(Class<T> clazz, byte[] bytes);

    // 序列化方法
    <T> byte[] serialize(T object);

}

提供两个实现,我这里直接将实现加入了枚举类 Serializer.Algorithm 中

java
enum SerializerAlgorithm implements Serializer {
	// Java 实现
    Java {
        @Override
        public <T> T deserialize(Class<T> clazz, byte[] bytes) {
            try {
                ObjectInputStream in = 
                    new ObjectInputStream(new ByteArrayInputStream(bytes));
                Object object = in.readObject();
                return (T) object;
            } catch (IOException | ClassNotFoundException e) {
                throw new RuntimeException("SerializerAlgorithm.Java 反序列化错误", e);
            }
        }

        @Override
        public <T> byte[] serialize(T object) {
            try {
                ByteArrayOutputStream out = new ByteArrayOutputStream();
                new ObjectOutputStream(out).writeObject(object);
                return out.toByteArray();
            } catch (IOException e) {
                throw new RuntimeException("SerializerAlgorithm.Java 序列化错误", e);
            }
        }
    }, 
    // Json 实现(引入了 Gson 依赖)
    Json {
        @Override
        public <T> T deserialize(Class<T> clazz, byte[] bytes) {
            return new Gson().fromJson(new String(bytes, StandardCharsets.UTF_8), clazz);
        }

        @Override
        public <T> byte[] serialize(T object) {
            return new Gson().toJson(object).getBytes(StandardCharsets.UTF_8);
        }
    };

    // 需要从协议的字节中得到是哪种序列化算法
    public static SerializerAlgorithm getByInt(int type) {
        SerializerAlgorithm[] array = SerializerAlgorithm.values();
        if (type < 0 || type > array.length - 1) {
            throw new IllegalArgumentException("超过 SerializerAlgorithm 范围");
        }
        return array[type];
    }
}

增加配置类和配置文件

java
public abstract class Config {
    static Properties properties;
    static {
        try (InputStream in = Config.class.getResourceAsStream("/application.properties")) {
            properties = new Properties();
            properties.load(in);
        } catch (IOException e) {
            throw new ExceptionInInitializerError(e);
        }
    }
    public static int getServerPort() {
        String value = properties.getProperty("server.port");
        if(value == null) {
            return 8080;
        } else {
            return Integer.parseInt(value);
        }
    }
    public static Serializer.Algorithm getSerializerAlgorithm() {
        String value = properties.getProperty("serializer.algorithm");
        if(value == null) {
            return Serializer.Algorithm.Java;
        } else {
            return Serializer.Algorithm.valueOf(value);
        }
    }
}

配置文件

properties
serializer.algorithm=Json

修改编解码器

java
/**
 * 必须和 LengthFieldBasedFrameDecoder 一起使用,确保接到的 ByteBuf 消息是完整的
 */
public class MessageCodecSharable extends MessageToMessageCodec<ByteBuf, Message> {
    @Override
    public void encode(ChannelHandlerContext ctx, Message msg, List<Object> outList) throws Exception {
        ByteBuf out = ctx.alloc().buffer();
        // 1. 4 字节的魔数
        out.writeBytes(new byte[]{1, 2, 3, 4});
        // 2. 1 字节的版本,
        out.writeByte(1);
        // 3. 1 字节的序列化方式 jdk 0 , json 1
        out.writeByte(Config.getSerializerAlgorithm().ordinal());
        // 4. 1 字节的指令类型
        out.writeByte(msg.getMessageType());
        // 5. 4 个字节
        out.writeInt(msg.getSequenceId());
        // 无意义,对齐填充
        out.writeByte(0xff);
        // 6. 获取内容的字节数组
        byte[] bytes = Config.getSerializerAlgorithm().serialize(msg);
        // 7. 长度
        out.writeInt(bytes.length);
        // 8. 写入内容
        out.writeBytes(bytes);
        outList.add(out);
    }

    @Override
    protected void decode(ChannelHandlerContext ctx, ByteBuf in, List<Object> out) throws Exception {
        int magicNum = in.readInt();
        byte version = in.readByte();
        byte serializerAlgorithm = in.readByte(); // 0 或 1
        byte messageType = in.readByte(); // 0,1,2...
        int sequenceId = in.readInt();
        in.readByte();
        int length = in.readInt();
        byte[] bytes = new byte[length];
        in.readBytes(bytes, 0, length);

        // 找到反序列化算法
        Serializer.Algorithm algorithm = Serializer.Algorithm.values()[serializerAlgorithm];
        // 确定具体消息类型
        Class<? extends Message> messageClass = Message.getMessageClass(messageType);
        Message message = algorithm.deserialize(messageClass, bytes);
//        log.debug("{}, {}, {}, {}, {}, {}", magicNum, version, serializerType, messageType, sequenceId, length);
//        log.debug("{}", message);
        out.add(message);
    }
}

其中确定具体消息类型,可以根据 消息类型字节 获取到对应的 消息 class

java
@Data
public abstract class Message implements Serializable {

    /**
     * 根据消息类型字节,获得对应的消息 class
     * @param messageType 消息类型字节
     * @return 消息 class
     */
    public static Class<? extends Message> getMessageClass(int messageType) {
        return messageClasses.get(messageType);
    }

    private int sequenceId;

    private int messageType;

    public abstract int getMessageType();

    public static final int LoginRequestMessage = 0;
    public static final int LoginResponseMessage = 1;
    public static final int ChatRequestMessage = 2;
    public static final int ChatResponseMessage = 3;
    public static final int GroupCreateRequestMessage = 4;
    public static final int GroupCreateResponseMessage = 5;
    public static final int GroupJoinRequestMessage = 6;
    public static final int GroupJoinResponseMessage = 7;
    public static final int GroupQuitRequestMessage = 8;
    public static final int GroupQuitResponseMessage = 9;
    public static final int GroupChatRequestMessage = 10;
    public static final int GroupChatResponseMessage = 11;
    public static final int GroupMembersRequestMessage = 12;
    public static final int GroupMembersResponseMessage = 13;
    public static final int PingMessage = 14;
    public static final int PongMessage = 15;
    private static final Map<Integer, Class<? extends Message>> messageClasses = new HashMap<>();

    static {
        messageClasses.put(LoginRequestMessage, LoginRequestMessage.class);
        messageClasses.put(LoginResponseMessage, LoginResponseMessage.class);
        messageClasses.put(ChatRequestMessage, ChatRequestMessage.class);
        messageClasses.put(ChatResponseMessage, ChatResponseMessage.class);
        messageClasses.put(GroupCreateRequestMessage, GroupCreateRequestMessage.class);
        messageClasses.put(GroupCreateResponseMessage, GroupCreateResponseMessage.class);
        messageClasses.put(GroupJoinRequestMessage, GroupJoinRequestMessage.class);
        messageClasses.put(GroupJoinResponseMessage, GroupJoinResponseMessage.class);
        messageClasses.put(GroupQuitRequestMessage, GroupQuitRequestMessage.class);
        messageClasses.put(GroupQuitResponseMessage, GroupQuitResponseMessage.class);
        messageClasses.put(GroupChatRequestMessage, GroupChatRequestMessage.class);
        messageClasses.put(GroupChatResponseMessage, GroupChatResponseMessage.class);
        messageClasses.put(GroupMembersRequestMessage, GroupMembersRequestMessage.class);
        messageClasses.put(GroupMembersResponseMessage, GroupMembersResponseMessage.class);
    }
}

5.2 参数调优

1)CONNECT_TIMEOUT_MILLIS

  • 属于 SocketChannal 参数

  • 用在客户端建立连接时,如果在指定毫秒内无法连接,会抛出 timeout 异常

  • SO_TIMEOUT 主要用在阻塞 IO,阻塞 IO 中 accept,read 等都是无限等待的,如果不希望永远阻塞,使用它调整超时时间

java
@Slf4j
public class TestConnectionTimeout {
    public static void main(String[] args) {
        NioEventLoopGroup group = new NioEventLoopGroup();
        try {
            Bootstrap bootstrap = new Bootstrap()
                    .group(group)
                    .option(ChannelOption.CONNECT_TIMEOUT_MILLIS, 300)
                    .channel(NioSocketChannel.class)
                    .handler(new LoggingHandler());
            ChannelFuture future = bootstrap.connect("127.0.0.1", 8080);
            future.sync().channel().closeFuture().sync(); // 断点1
        } catch (Exception e) {
            e.printStackTrace();
            log.debug("timeout");
        } finally {
            group.shutdownGracefully();
        }
    }
}

另外源码部分 io.netty.channel.nio.AbstractNioChannel.AbstractNioUnsafe#connect

java
@Override
public final void connect(
        final SocketAddress remoteAddress, final SocketAddress localAddress, final ChannelPromise promise) {
    // ...
    // Schedule connect timeout.
    int connectTimeoutMillis = config().getConnectTimeoutMillis();
    if (connectTimeoutMillis > 0) {
        connectTimeoutFuture = eventLoop().schedule(new Runnable() {
            @Override
            public void run() {                
                ChannelPromise connectPromise = AbstractNioChannel.this.connectPromise;
                ConnectTimeoutException cause =
                    new ConnectTimeoutException("connection timed out: " + remoteAddress); // 断点2
                if (connectPromise != null && connectPromise.tryFailure(cause)) {
                    close(voidPromise());
                }
            }
        }, connectTimeoutMillis, TimeUnit.MILLISECONDS);
    }
	// ...
}

2)SO_BACKLOG

  • 属于 ServerSocketChannal 参数
mermaid
sequenceDiagram

participant c as client
participant s as server
participant sq as syns queue
participant aq as accept queue

s ->> s : bind()
s ->> s : listen()
c ->> c : connect()
c ->> s : 1. SYN
Note left of c : SYN_SEND
s ->> sq : put
Note right of s : SYN_RCVD
s ->> c : 2. SYN + ACK
Note left of c : ESTABLISHED
c ->> s : 3. ACK
sq ->> aq : put
Note right of s : ESTABLISHED
aq -->> s : 
s ->> s : accept()
  1. 第一次握手,client 发送 SYN 到 server,状态修改为 SYN_SEND,server 收到,状态改变为 SYN_REVD,并将该请求放入 sync queue 队列
  2. 第二次握手,server 回复 SYN + ACK 给 client,client 收到,状态改变为 ESTABLISHED,并发送 ACK 给 server
  3. 第三次握手,server 收到 ACK,状态改变为 ESTABLISHED,将该请求从 sync queue 放入 accept queue

其中

  • 在 linux 2.2 之前,backlog 大小包括了两个队列的大小,在 2.2 之后,分别用下面两个参数来控制

  • sync queue - 半连接队列

    • 大小通过 /proc/sys/net/ipv4/tcp_max_syn_backlog 指定,在 syncookies 启用的情况下,逻辑上没有最大值限制,这个设置便被忽略
  • accept queue - 全连接队列

    • 其大小通过 /proc/sys/net/core/somaxconn 指定,在使用 listen 函数时,内核会根据传入的 backlog 参数与系统参数,取二者的较小值
    • 如果 accpet queue 队列满了,server 将发送一个拒绝连接的错误信息到 client

netty 中

可以通过 option(ChannelOption.SO_BACKLOG, 值) 来设置大小

可以通过下面源码查看默认大小

java
public class DefaultServerSocketChannelConfig extends DefaultChannelConfig
                                              implements ServerSocketChannelConfig {

    private volatile int backlog = NetUtil.SOMAXCONN;
    // ...
}

课堂调试关键断点为:io.netty.channel.nio.NioEventLoop#processSelectedKey

oio 中更容易说明,不用 debug 模式

java
public class Server {
    public static void main(String[] args) throws IOException {
        ServerSocket ss = new ServerSocket(8888, 2);
        Socket accept = ss.accept();
        System.out.println(accept);
        System.in.read();
    }
}

客户端启动 4 个

java
public class Client {
    public static void main(String[] args) throws IOException {
        try {
            Socket s = new Socket();
            System.out.println(new Date()+" connecting...");
            s.connect(new InetSocketAddress("localhost", 8888),1000);
            System.out.println(new Date()+" connected...");
            s.getOutputStream().write(1);
            System.in.read();
        } catch (IOException e) {
            System.out.println(new Date()+" connecting timeout...");
            e.printStackTrace();
        }
    }
}

第 1,2,3 个客户端都打印,但除了第一个处于 accpet 外,其它两个都处于 accept queue 中

java
Tue Apr 21 20:30:28 CST 2020 connecting...
Tue Apr 21 20:30:28 CST 2020 connected...

第 4 个客户端连接时

Tue Apr 21 20:53:58 CST 2020 connecting...
Tue Apr 21 20:53:59 CST 2020 connecting timeout...
java.net.SocketTimeoutException: connect timed out

3)ulimit -n

  • 属于操作系统参数

4)TCP_NODELAY

  • 属于 SocketChannal 参数

5)SO_SNDBUF & SO_RCVBUF

  • SO_SNDBUF 属于 SocketChannal 参数
  • SO_RCVBUF 既可用于 SocketChannal 参数,也可以用于 ServerSocketChannal 参数(建议设置到 ServerSocketChannal 上)

6)ALLOCATOR

  • 属于 SocketChannal 参数
  • 用来分配 ByteBuf, ctx.alloc()

7)RCVBUF_ALLOCATOR

  • 属于 SocketChannal 参数
  • 控制 netty 接收缓冲区大小
  • 负责入站数据的分配,决定入站缓冲区的大小(并可动态调整),统一采用 direct 直接内存,具体池化还是非池化由 allocator 决定