粘包与半包
粘包现象
粘包的问题出现是因为不知道一个用户消息的边界在哪，如果知道了边界在哪，接收方就可以通过边界来划分出有效的用户消息。
服务端代码
public class HelloWorldServer {
static final Logger log = LoggerFactory.getLogger(HelloWorldServer.class);
void start() {
NioEventLoopGroup boss = new NioEventLoopGroup(1);
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 ChannelInboundHandlerAdapter() {
@Override
public void channelActive(ChannelHandlerContext ctx) throws Exception {
log.debug("connected {}", ctx.channel());
super.channelActive(ctx);
}
@Override
public void channelInactive(ChannelHandlerContext ctx) throws Exception {
log.debug("disconnect {}", ctx.channel());
super.channelInactive(ctx);
}
});
}
});
ChannelFuture channelFuture = serverBootstrap.bind(8080);
log.debug("{} binding...", channelFuture.channel());
channelFuture.sync();
log.debug("{} bound...", channelFuture.channel());
channelFuture.channel().closeFuture().sync();
} catch (InterruptedException e) {
log.error("server error", e);
} finally {
boss.shutdownGracefully();
worker.shutdownGracefully();
log.debug("stoped");
}
}
public static void main(String[] args) {
new HelloWorldServer().start();
}
}
客户端代码希望发送 10 个消息，每个消息是 16 字节
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 ChannelInboundHandlerAdapter() {
@Override
//会在连接channel建立成功后，会触发Active事件
public void channelActive(ChannelHandlerContext ctx) throws Exception {
log.debug("sending...");
Random r = new Random();
char c = 'a';
for (int i = 0; i < 10; i++) {
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);
}
}
});
}
});
ChannelFuture channelFuture = bootstrap.connect("127.0.0.1", 8080).sync();
channelFuture.channel().closeFuture().sync();
} catch (InterruptedException e) {
log.error("client error", e);
} finally {
worker.shutdownGracefully();
}
}
}
服务器端的某次输出，可以看到一次就接收了 160 个字节，而期望的是一次16字节，分 10 次接收。这就出现了粘包现象
08:24:46 [DEBUG] [main] c.i.n.HelloWorldServer - [id: 0x81e0fda5] binding...
08:24:46 [DEBUG] [main] c.i.n.HelloWorldServer - [id: 0x81e0fda5, L:/0:0:0:0:0:0:0:0:8080] bound...
08:24:55 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x94132411, L:/127.0.0.1:8080 - R:/127.0.0.1:58177] REGISTERED
08:24:55 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x94132411, L:/127.0.0.1:8080 - R:/127.0.0.1:58177] ACTIVE
08:24:55 [DEBUG] [nioEventLoopGroup-3-1] c.i.n.HelloWorldServer - connected [id: 0x94132411, L:/127.0.0.1:8080 - R:/127.0.0.1:58177]
08:24:55 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x94132411, L:/127.0.0.1:8080 - R:/127.0.0.1:58177] READ: 160B
+-------------------------------------------------+
| 0 1 2 3 4 5 6 7 8 9 a b c d e f |
+--------+-------------------------------------------------+----------------+
|00000000| 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f |................|
|00000010| 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f |................|
|00000020| 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f |................|
|00000030| 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f |................|
|00000040| 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f |................|
|00000050| 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f |................|
|00000060| 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f |................|
|00000070| 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f |................|
|00000080| 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f |................|
|00000090| 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f |................|
+--------+-------------------------------------------------+----------------+
08:24:55 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x94132411, L:/127.0.0.1:8080 - R:/127.0.0.1:58177] READ COMPLETE
半包现象
半包是指 接收端只收到了部分数据，而非完整的数据的情况
客户端代码希望发送 1 个消息，这个消息是 160 字节，代码改为
ByteBuf buffer = ctx.alloc().buffer();
for (int i = 0; i < 10; i++) {
buffer.writeBytes(new byte[]{0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15});
}
ctx.writeAndFlush(buffer);
为现象明显，服务端修改一下接收缓冲区，其它代码不变
serverBootstrap.option(ChannelOption.SO_RCVBUF, 10);
服务器端的某次输出，可以看到接收的消息被分为两节，如 第一次 20 字节，第二次 140 字节
08:43:49 [DEBUG] [main] c.i.n.HelloWorldServer - [id: 0x4d6c6a84] binding...
08:43:49 [DEBUG] [main] c.i.n.HelloWorldServer - [id: 0x4d6c6a84, L:/0:0:0:0:0:0:0:0:8080] bound...
08:44:23 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x1719abf7, L:/127.0.0.1:8080 - R:/127.0.0.1:59221] REGISTERED
08:44:23 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x1719abf7, L:/127.0.0.1:8080 - R:/127.0.0.1:59221] ACTIVE
08:44:23 [DEBUG] [nioEventLoopGroup-3-1] c.i.n.HelloWorldServer - connected [id: 0x1719abf7, L:/127.0.0.1:8080 - R:/127.0.0.1:59221]
08:44:24 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x1719abf7, L:/127.0.0.1:8080 - R:/127.0.0.1:59221] READ: 20B
+-------------------------------------------------+
| 0 1 2 3 4 5 6 7 8 9 a b c d e f |
+--------+-------------------------------------------------+----------------+
|00000000| 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f |................|
|00000010| 00 01 02 03 |.... |
+--------+-------------------------------------------------+----------------+
08:44:24 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x1719abf7, L:/127.0.0.1:8080 - R:/127.0.0.1:59221] READ COMPLETE
08:44:24 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x1719abf7, L:/127.0.0.1:8080 - R:/127.0.0.1:59221] READ: 140B
+-------------------------------------------------+
| 0 1 2 3 4 5 6 7 8 9 a b c d e f |
+--------+-------------------------------------------------+----------------+
|00000000| 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f 00 01 02 03 |................|
|00000010| 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f 00 01 02 03 |................|
|00000020| 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f 00 01 02 03 |................|
|00000030| 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f 00 01 02 03 |................|
|00000040| 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f 00 01 02 03 |................|
|00000050| 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f 00 01 02 03 |................|
|00000060| 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f 00 01 02 03 |................|
|00000070| 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f 00 01 02 03 |................|
|00000080| 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f |............ |
+--------+-------------------------------------------------+----------------+
08:44:24 [DEBUG] [nioEventLoopGroup-3-1] i.n.h.l.LoggingHandler - [id: 0x1719abf7, L:/127.0.0.1:8080 - R:/127.0.0.1:59221] READ COMPLETE
注意：serverBootstrap.option(ChannelOption.SO_RCVBUF, 10) 影响的底层接收缓冲区（即滑动窗口）大小，仅决定了 netty 读取的最小单位，netty 实际每次读取的一般是它的整数倍
现象分析
这里出现的粘包半包问题，并非是JavaNIO或Netty的问题，本质是TCP是流失协议，消息无边界。
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
解决方案
接下来看下Netty如何解决以上问题的：
- 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
- 每一条消息采用固定长度，缺点浪费空间
- 每一条消息采用分隔符，例如 \n，缺点需要转义
- 每一条消息分为 head 和 body，head 中包含 body 的长度
方法1：短链接（极不推荐）
以解决粘包为例
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
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
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
12:07:00 [DEBUG] [nioEventLoopGroup-2-1] c.i.n.HelloWorldClient - connetted...
12:07:00 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0x3c2ef3c2] REGISTERED
12:07:00 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0x3c2ef3c2] CONNECT: /192.168.0.103:9090
12:07:00 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0x3c2ef3c2, L:/192.168.0.103:53155 - R:/192.168.0.103:9090] ACTIVE
12:07:00 [DEBUG] [nioEventLoopGroup-2-1] c.i.n.HelloWorldClient - sending...
12:07:00 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0x3c2ef3c2, L:/192.168.0.103:53155 - R:/192.168.0.103:9090] WRITE: 80B
+-------------------------------------------------+
| 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....|
+--------+-------------------------------------------------+----------------+
12:07:00 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0x3c2ef3c2, L:/192.168.0.103:53155 - R:/192.168.0.103:9090] FLUSH
服务端输出
12:06:51 [DEBUG] [main] c.i.n.HelloWorldServer - [id: 0xe3d9713f] binding...
12:06:51 [DEBUG] [main] c.i.n.HelloWorldServer - [id: 0xe3d9713f, L:/192.168.0.103:9090] bound...
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] REGISTERED
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] ACTIVE
12:07:00 [DEBUG] [nioEventLoopGroup-3-1] c.i.n.HelloWorldServer - connected [id: 0xd739f137, L:/192.168.0.103:9090 - R:/192.168.0.103:53155]
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| 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.... |
+--------+-------------------------------------------------+----------------+
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 COMPLETE
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
ch.pipeline().addLast(new LineBasedFrameDecoder(1024));
客户端在每条消息之后，加入 \n 分隔符
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
14:08:18 [DEBUG] [nioEventLoopGroup-2-1] c.i.n.HelloWorldClient - connetted...
14:08:18 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0x1282d755] REGISTERED
14:08:18 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0x1282d755] CONNECT: /192.168.0.103:9090
14:08:18 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0x1282d755, L:/192.168.0.103:63641 - R:/192.168.0.103:9090] ACTIVE
14:08:18 [DEBUG] [nioEventLoopGroup-2-1] c.i.n.HelloWorldClient - sending...
14:08:18 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0x1282d755, L:/192.168.0.103:63641 - R:/192.168.0.103:9090] WRITE: 60B
+-------------------------------------------------+
| 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. |
+--------+-------------------------------------------------+----------------+
14:08:18 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0x1282d755, L:/192.168.0.103:63641 - R:/192.168.0.103:9090] FLUSH
服务端输出
14:08:18 [DEBUG] [nioEventLoopGroup-3-5] c.i.n.HelloWorldServer - connected [id: 0xa4b3be43, L:/192.168.0.103:9090 - R:/192.168.0.103:63641]
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: 1B
+-------------------------------------------------+
| 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
// 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
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
14:37:10 [DEBUG] [nioEventLoopGroup-2-1] c.i.n.HelloWorldClient - connetted...
14:37:10 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0xf0f347b8] REGISTERED
14:37:10 [DEBUG] [nioEventLoopGroup-2-1] i.n.h.l.LoggingHandler - [id: 0xf0f347b8] CONNECT: /192.168.0.103:9090
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] ACTIVE
14:37:10 [DEBUG] [nioEventLoopGroup-2-1] c.i.n.HelloWorldClient - sending...
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] WRITE: 97B
+-------------------------------------------------+
| 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
14:36:50 [DEBUG] [main] c.i.n.HelloWorldServer - [id: 0xdff439d3] binding...
14:36:51 [DEBUG] [main] c.i.n.HelloWorldServer - [id: 0xdff439d3, L:/192.168.0.103:9090] bound...
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] REGISTERED
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] ACTIVE
14:37:10 [DEBUG] [nioEventLoopGroup-3-1] c.i.n.HelloWorldServer - connected [id: 0x744f2b47, L:/192.168.0.103:9090 - R:/192.168.0.103:49979]
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| 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
协议设计与解析
为什么需要协议？
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
例如：在网络上传输
下雨天留客天留我不留
是中文一句著名的无标点符号句子，在没有标点符号情况下，这句话有数种拆解方式，而意思却是完全不同，所以常被用作讲述标点符号的重要性
一种解读
下雨天留客，天留，我不留
另一种解读
下雨天，留客天，留我不？留
如何设计协议呢？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
模拟 redis 客户端发送命令。
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());//*2 表示数组元素个数为2，即get aaa 是2串内容
buf.writeBytes(LINE);
buf.writeBytes("$3".getBytes());//规定用 $3 表示后续有3个字节
buf.writeBytes(LINE);
buf.writeBytes("get".getBytes());//输入gset命令
buf.writeBytes(LINE);
buf.writeBytes("$3".getBytes());
buf.writeBytes(LINE);
buf.writeBytes("aaa".getBytes());//输入key为 aaa
buf.writeBytes(LINE);
ctx.writeAndFlush(buf);
}
private void set(ChannelHandlerContext ctx) {
//以下redis命令为 set aaa bbb
ByteBuf buf = ctx.alloc().buffer();
buf.writeBytes("*3".getBytes());//*3 表示数组元素个数为3，即set aaa bbb是3串内容
buf.writeBytes(LINE);
buf.writeBytes("$3".getBytes());//规定用 $3 表示后续有3个字节
buf.writeBytes(LINE);
buf.writeBytes("set".getBytes());//输入set命令
buf.writeBytes(LINE);
buf.writeBytes("$3".getBytes());
buf.writeBytes(LINE);
buf.writeBytes("aaa".getBytes());//输入key为 aaa
buf.writeBytes(LINE);
buf.writeBytes("$3".getBytes());
buf.writeBytes(LINE);
buf.writeBytes("bbb".getBytes());//输入value为 bbb
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();
}
当然 netty提供了现成的这些协议，不需要我们自己来开发，这里是为了知其所以然
http 协议举例
模拟http服务端
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：约定好的，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
@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
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);
@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
@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);
}
}
扩展序列化算法
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
// 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
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，这里直接将具体实现加入了枚举类 Serializer.Algorithm 中
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
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
serializer.algorithm=Json
修改编解码器
/**
* 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
@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);
}
}
参数调优
相关源码待更新
客户端参数 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
@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
@Override
public final void connect(
final SocketAddress remoteAddress, final SocketAddress localAddress, final ChannelPromise promise) {
// ...
// 获取超时时间
int connectTimeoutMillis = config().getConnectTimeoutMillis();
if (connectTimeoutMillis > 0) {//如果超时时间大于0
connectTimeoutFuture = eventLoop().schedule(new Runnable() {//则启动定时任务
@Override
public void run() {
//connectPromise是两个线程间交换数据的对象
ChannelPromise connectPromise = AbstractNioChannel.this.connectPromise;
ConnectTimeoutException cause =
new ConnectTimeoutException("connection timed out: " + remoteAddress); // 断点2
//connectPromise.tryFailure(cause) 将异常放到connectPromise里，再唤醒主线程
if (connectPromise != null && connectPromise.tryFailure(cause)) {
close(voidPromise());
}
}
}, connectTimeoutMillis, TimeUnit.MILLISECONDS);//定时任务在 connectTimeoutMillis 时间后执行
}
// ...
}
服务端参数 SO_BACKLOG
这是属于 ServerSocketChannal 的参数
三次握手时有半连接队列和全连接队列，详情看这篇文章：TCP - 半连接队列和全连接队列
- 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
- 大小通过 /proc/sys/net/ipv4/tcp_max_syn_backlog 指定，在
- 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backlog 的大小
可以通过下面源码查看默认大小
public class DefaultServerSocketChannelConfig extends DefaultChannelConfig
implements ServerSocketChannelConfig {
private volatile int backlog = NetUtil.SOMAXCONN;
// ...
}
TCP_NODELAY
- 属于 SocketChannal 参数
立即发送，建议设置成true。false就是开启了nagle算法
SO_SNDBUF & SO_RCVBUF
设置滑动窗口的参数，在早些可能需要设置这些参数，但现在tcp会根据拥塞等对窗口进行自动调整，因此不建议手动设置这两个值。
- 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）
ALLOCATOR
- 属于 SocketChannal 参数
- 用来分配 ByteBuf， ctx.alloc()
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
源码详解
启动剖析
我们就来看看 netty 中对下面的代码是怎样进行处理的
//1 netty 中使用 NioEventLoopGroup （简称 nio boss 线程）来封装线程和 selector
Selector selector = Selector.open();
//2 创建 NioServerSocketChannel，同时会初始化它关联的 handler，以及为原生 ssc 存储 config
NioServerSocketChannel attachment = new NioServerSocketChannel();
//3 创建 NioServerSocketChannel 时，创建了 java 原生的 ServerSocketChannel
ServerSocketChannel serverSocketChannel = ServerSocketChannel.open();
serverSocketChannel.configureBlocking(false);
//4 启动 nio boss 线程执行接下来的操作
//5 注册（仅关联 selector 和 NioServerSocketChannel），未关注事件
SelectionKey selectionKey = serverSocketChannel.register(selector, 0, attachment);
//6 head -> 初始化器 -> ServerBootstrapAcceptor -> tail，初始化器是一次性的，只为添加 acceptor
//7 绑定端口
serverSocketChannel.bind(new InetSocketAddress(8080));
//8 触发 channel active 事件，在 head 中关注 op_accept 事件
selectionKey.interestOps(SelectionKey.OP_ACCEPT);
入口 io.netty.bootstrap.ServerBootstrap#bind
关键代码 io.netty.bootstrap.AbstractBootstrap#doBind
这个函数是由哪些线程处理的呢？可以先有个概念，再往下看：
- init & register regFuture 处理
- init：由main处理
- 创建NioServerSocketChannel：由main处理
- 添加 NioServerSocketChannel 初始化 handler ：由main处理
- 初始化 handler 等待调用
- init：由main处理
- register
- 启动 nio boss 线程 ：由main处理
- 原生 ssc 注册至 selector 未关注事件：由nio-thread处理
- 执行 NioServerSocketChannel 初始化 handler：由nio-thread处理
- regFuture 等待回调 doBind0：由nio-thread处理
- 原生 ServerSocketChannel 绑定：由nio-thread处理
- 触发NioServerSocketChannel active 事件：由nio-thread处理
private ChannelFuture doBind(final SocketAddress localAddress) {
// 1. 执行初始化和注册 regFuture 会由 initAndRegister 设置其是否完成，从而回调 3.2 处代码
// init就相当于 ServerSocketChannel ssc= ServerSocketChannel.open();
// Register就相当于 SelectionKey selectionKey=ssc.register(selector, 0, nettySsc);
final ChannelFuture regFuture = initAndRegister();
final Channel channel = regFuture.channel();
if (regFuture.cause() != null) {
return regFuture;
}
// 2. 因为是 initAndRegister 异步执行，需要分两种情况来看，调试时也需要通过 suspend 断点类型加以区分
// 2.1 如果已经完成
if (regFuture.isDone()) {
ChannelPromise promise = channel.newPromise();
// 3.1 立刻调用 doBind0
doBind0(regFuture, channel, localAddress, promise);
return promise;
}
// 2.2 还没有完成
else {
final PendingRegistrationPromise promise = new PendingRegistrationPromise(channel);
// 3.2 回调 doBind0
regFuture.addListener(new ChannelFutureListener() {
@Override
public void operationComplete(ChannelFuture future) throws Exception {
Throwable cause = future.cause();
if (cause != null) {
// 处理异常...
promise.setFailure(cause);
} else {
promise.registered();
// 3. 由注册线程去执行 doBind0
doBind0(regFuture, channel, localAddress, promise);
}
}
});
return promise;
}
}
关键代码 io.netty.bootstrap.AbstractBootstrap#initAndRegister
final ChannelFuture initAndRegister() {
Channel channel = null;
try {
channel = channelFactory.newChannel();
// 1.1 初始化 - 做的事就是添加一个初始化器 ChannelInitializer
init(channel);
} catch (Throwable t) {
// 处理异常...
return new DefaultChannelPromise(new FailedChannel(), GlobalEventExecutor.INSTANCE).setFailure(t);
}
// 1.2 注册 - 做的事就是将原生 channel 注册到 selector 上
ChannelFuture regFuture = config().group().register(channel);
if (regFuture.cause() != null) {
// 处理异常...
}
return regFuture;
}
关键代码 io.netty.bootstrap.ServerBootstrap#init
// 这里 channel 实际上是 NioServerSocketChannel
void init(Channel channel) throws Exception {
final Map<ChannelOption<?>, Object> options = options0();
synchronized (options) {
setChannelOptions(channel, options, logger);
}
final Map<AttributeKey<?>, Object> attrs = attrs0();
synchronized (attrs) {
for (Entry<AttributeKey<?>, Object> e: attrs.entrySet()) {
@SuppressWarnings("unchecked")
AttributeKey<Object> key = (AttributeKey<Object>) e.getKey();
channel.attr(key).set(e.getValue());
}
}
ChannelPipeline p = channel.pipeline();
final EventLoopGroup currentChildGroup = childGroup;
final ChannelHandler currentChildHandler = childHandler;
final Entry<ChannelOption<?>, Object>[] currentChildOptions;
final Entry<AttributeKey<?>, Object>[] currentChildAttrs;
synchronized (childOptions) {
currentChildOptions = childOptions.entrySet().toArray(newOptionArray(0));
}
synchronized (childAttrs) {
currentChildAttrs = childAttrs.entrySet().toArray(newAttrArray(0));
}
// 为 NioServerSocketChannel 添加初始化器
p.addLast(new ChannelInitializer<Channel>() {
@Override
public void initChannel(final Channel ch) throws Exception {
final ChannelPipeline pipeline = ch.pipeline();
ChannelHandler handler = config.handler();
if (handler != null) {
pipeline.addLast(handler);
}
// 初始化器的职责是将 ServerBootstrapAcceptor 加入至 NioServerSocketChannel
ch.eventLoop().execute(new Runnable() {
@Override
public void run() {
pipeline.addLast(new ServerBootstrapAcceptor(
ch, currentChildGroup, currentChildHandler, currentChildOptions, currentChildAttrs));
}
});
}
});
}
关键代码 io.netty.channel.AbstractChannel.AbstractUnsafe#register
public final void register(EventLoop eventLoop, final ChannelPromise promise) {
// 一些检查，略...
AbstractChannel.this.eventLoop = eventLoop;
if (eventLoop.inEventLoop()) {
register0(promise);
} else {
try {
// 首次执行 execute 方法时，会启动 nio 线程，之后注册等操作在 nio 线程上执行
// 因为只有一个 NioServerSocketChannel 因此，也只会有一个 boss nio 线程
// 这行代码完成的事实是 main -> nio boss 线程的切换
eventLoop.execute(new Runnable() {
@Override
public void run() {
register0(promise);
}
});
} catch (Throwable t) {
// 日志记录...
closeForcibly();
closeFuture.setClosed();
safeSetFailure(promise, t);
}
}
}
io.netty.channel.AbstractChannel.AbstractUnsafe#register0
private void register0(ChannelPromise promise) {
try {
if (!promise.setUncancellable() || !ensureOpen(promise)) {
return;
}
boolean firstRegistration = neverRegistered;
// 1.2.1 原生的 nio channel 绑定到 selector 上，注意此时没有注册 selector 关注事件，附件为 NioServerSocketChannel
doRegister();
neverRegistered = false;
registered = true;
// 1.2.2 执行 NioServerSocketChannel 初始化器的 initChannel
pipeline.invokeHandlerAddedIfNeeded();
// 回调 3.2 io.netty.bootstrap.AbstractBootstrap#doBind0
safeSetSuccess(promise);
pipeline.fireChannelRegistered();
// 对应 server socket channel 还未绑定，isActive 为 false
if (isActive()) {
if (firstRegistration) {
pipeline.fireChannelActive();
} else if (config().isAutoRead()) {
beginRead();
}
}
} catch (Throwable t) {
// Close the channel directly to avoid FD leak.
closeForcibly();
closeFuture.setClosed();
safeSetFailure(promise, t);
}
}
关键代码 io.netty.channel.ChannelInitializer#initChannel
private boolean initChannel(ChannelHandlerContext ctx) throws Exception {
if (initMap.add(ctx)) { // Guard against re-entrance.
try {
// 1.2.2.1 执行初始化
initChannel((C) ctx.channel());
} catch (Throwable cause) {
exceptionCaught(ctx, cause);
} finally {
// 1.2.2.2 移除初始化器
ChannelPipeline pipeline = ctx.pipeline();
if (pipeline.context(this) != null) {
pipeline.remove(this);
}
}
return true;
}
return false;
}
关键代码 io.netty.bootstrap.AbstractBootstrap#doBind0
// 3.1 或 3.2 执行 doBind0
private static void doBind0(
final ChannelFuture regFuture, final Channel channel,
final SocketAddress localAddress, final ChannelPromise promise) {
channel.eventLoop().execute(new Runnable() {
@Override
public void run() {
if (regFuture.isSuccess()) {
channel.bind(localAddress, promise).addListener(ChannelFutureListener.CLOSE_ON_FAILURE);
} else {
promise.setFailure(regFuture.cause());
}
}
});
}
关键代码 io.netty.channel.AbstractChannel.AbstractUnsafe#bind
public final void bind(final SocketAddress localAddress, final ChannelPromise promise) {
assertEventLoop();
if (!promise.setUncancellable() || !ensureOpen(promise)) {
return;
}
if (Boolean.TRUE.equals(config().getOption(ChannelOption.SO_BROADCAST)) &&
localAddress instanceof InetSocketAddress &&
!((InetSocketAddress) localAddress).getAddress().isAnyLocalAddress() &&
!PlatformDependent.isWindows() && !PlatformDependent.maybeSuperUser()) {
// 记录日志...
}
boolean wasActive = isActive();
try {
// 3.3 执行端口绑定
doBind(localAddress);
} catch (Throwable t) {
safeSetFailure(promise, t);
closeIfClosed();
return;
}
if (!wasActive && isActive()) {
invokeLater(new Runnable() {
@Override
public void run() {
// 3.4 触发 active 事件
pipeline.fireChannelActive();
}
});
}
safeSetSuccess(promise);
}
关键代码 io.netty.channel.socket.nio.NioServerSocketChannel#doBind
protected void doBind(SocketAddress localAddress) throws Exception {
if (PlatformDependent.javaVersion() >= 7) {
javaChannel().bind(localAddress, config.getBacklog());
} else {
javaChannel().socket().bind(localAddress, config.getBacklog());
}
}
关键代码 io.netty.channel.DefaultChannelPipeline.HeadContext#channelActive
public void channelActive(ChannelHandlerContext ctx) {
ctx.fireChannelActive();
// 触发 read (NioServerSocketChannel 上的 read 不是读取数据，只是为了触发 channel 的事件注册)
readIfIsAutoRead();
}
关键代码 io.netty.channel.nio.AbstractNioChannel#doBeginRead
protected void doBeginRead() throws Exception {
// Channel.read() or ChannelHandlerContext.read() was called
final SelectionKey selectionKey = this.selectionKey;
if (!selectionKey.isValid()) {
return;
}
readPending = true;
final int interestOps = selectionKey.interestOps();
// readInterestOp 取值是 16，在 NioServerSocketChannel 创建时初始化好，代表关注 accept 事件
if ((interestOps & readInterestOp) == 0) {
selectionKey.interestOps(interestOps | readInterestOp);
}
}
NioEventLoop 剖析
NioEventLoop 线程不仅要处理 IO 事件，还要处理 Task（包括普通任务和定时任务），
提交任务代码 io.netty.util.concurrent.SingleThreadEventExecutor#execute
public void execute(Runnable task) {
if (task == null) {
throw new NullPointerException("task");
}
boolean inEventLoop = inEventLoop();
// 添加任务，其中队列使用了 jctools 提供的 mpsc 无锁队列
addTask(task);
if (!inEventLoop) {
// inEventLoop 如果为 false 表示由其它线程来调用 execute，即首次调用，这时需要向 eventLoop 提交首个任务，启动死循环，会执行到下面的 doStartThread
startThread();
if (isShutdown()) {
// 如果已经 shutdown，做拒绝逻辑，代码略...
}
}
if (!addTaskWakesUp && wakesUpForTask(task)) {
// 如果线程由于 IO select 阻塞了，添加的任务的线程需要负责唤醒 NioEventLoop 线程
wakeup(inEventLoop);
}
}
唤醒 select 阻塞线程io.netty.channel.nio.NioEventLoop#wakeup
@Override
protected void wakeup(boolean inEventLoop) {
if (!inEventLoop && wakenUp.compareAndSet(false, true)) {
selector.wakeup();
}
}
启动 EventLoop 主循环 io.netty.util.concurrent.SingleThreadEventExecutor#doStartThread
private void doStartThread() {
assert thread == null;
executor.execute(new Runnable() {
@Override
public void run() {
// 将线程池的当前线程保存在成员变量中，以便后续使用
thread = Thread.currentThread();
if (interrupted) {
thread.interrupt();
}
boolean success = false;
updateLastExecutionTime();
try {
// 调用外部类 SingleThreadEventExecutor 的 run 方法，进入死循环，run 方法见下
SingleThreadEventExecutor.this.run();
success = true;
} catch (Throwable t) {
logger.warn("Unexpected exception from an event executor: ", t);
} finally {
// 清理工作，代码略...
}
}
});
}
io.netty.channel.nio.NioEventLoop#run 主要任务是执行死循环，不断看有没有新任务，有没有 IO 事件
protected void run() {
for (;;) {
try {
try {
// calculateStrategy 的逻辑如下：
// 有任务，会执行一次 selectNow，清除上一次的 wakeup 结果，无论有没有 IO 事件，都会跳过 switch
// 没有任务，会匹配 SelectStrategy.SELECT，看是否应当阻塞
switch (selectStrategy.calculateStrategy(selectNowSupplier, hasTasks())) {
case SelectStrategy.CONTINUE:
continue;
case SelectStrategy.BUSY_WAIT:
case SelectStrategy.SELECT:
// 因为 IO 线程和提交任务线程都有可能执行 wakeup，而 wakeup 属于比较昂贵的操作，因此使用了一个原子布尔对象 wakenUp，它取值为 true 时，表示该由当前线程唤醒
// 进行 select 阻塞，并设置唤醒状态为 false
boolean oldWakenUp = wakenUp.getAndSet(false);
// 如果在这个位置，非 EventLoop 线程抢先将 wakenUp 置为 true，并 wakeup
// 下面的 select 方法不会阻塞
// 等 runAllTasks 处理完成后，到再循环进来这个阶段新增的任务会不会及时执行呢?
// 因为 oldWakenUp 为 true，因此下面的 select 方法就会阻塞，直到超时
// 才能执行，让 select 方法无谓阻塞
select(oldWakenUp);
if (wakenUp.get()) {
selector.wakeup();
}
default:
}
} catch (IOException e) {
rebuildSelector0();
handleLoopException(e);
continue;
}
cancelledKeys = 0;
needsToSelectAgain = false;
// ioRatio 默认是 50
final int ioRatio = this.ioRatio;
if (ioRatio == 100) {
try {
processSelectedKeys();
} finally {
// ioRatio 为 100 时，总是运行完所有非 IO 任务
runAllTasks();
}
} else {
final long ioStartTime = System.nanoTime();
try {
processSelectedKeys();
} finally {
// 记录 io 事件处理耗时
final long ioTime = System.nanoTime() - ioStartTime;
// 运行非 IO 任务，一旦超时会退出 runAllTasks
runAllTasks(ioTime * (100 - ioRatio) / ioRatio);
}
}
} catch (Throwable t) {
handleLoopException(t);
}
try {
if (isShuttingDown()) {
closeAll();
if (confirmShutdown()) {
return;
}
}
} catch (Throwable t) {
handleLoopException(t);
}
}
}
注意
这里有个费解的地方就是 wakeup，它既可以由提交任务的线程来调用（比较好理解），也可以由 EventLoop 线程来调用（比较费解），这里要知道 wakeup 方法的效果：
- 由非 EventLoop 线程调用，会唤醒当前在执行 select 阻塞的 EventLoop 线程
- 由 EventLoop 自己调用，会本次的 wakeup 会取消下一次的 select 操作
io.netty.channel.nio.NioEventLoop#select
private void select(boolean oldWakenUp) throws IOException {
Selector selector = this.selector;
try {
int selectCnt = 0;
long currentTimeNanos = System.nanoTime();
// 计算等待时间
// * 没有 scheduledTask，超时时间为 1s
// * 有 scheduledTask，超时时间为 `下一个定时任务执行时间 - 当前时间`
long selectDeadLineNanos = currentTimeNanos + delayNanos(currentTimeNanos);
for (;;) {
long timeoutMillis = (selectDeadLineNanos - currentTimeNanos + 500000L) / 1000000L;
// 如果超时，退出循环
if (timeoutMillis <= 0) {
if (selectCnt == 0) {
selector.selectNow();
selectCnt = 1;
}
break;
}
// 如果期间又有 task 退出循环，如果没这个判断，那么任务就会等到下次 select 超时时才能被执行
// wakenUp.compareAndSet(false, true) 是让非 NioEventLoop 不必再执行 wakeup
if (hasTasks() && wakenUp.compareAndSet(false, true)) {
selector.selectNow();
selectCnt = 1;
break;
}
// select 有限时阻塞
// 注意 nio 有 bug，当 bug 出现时，select 方法即使没有时间发生，也不会阻塞住，导致不断空轮询，cpu 占用 100%
int selectedKeys = selector.select(timeoutMillis);
// 计数加 1
selectCnt ++;
// 醒来后，如果有 IO 事件、或是由非 EventLoop 线程唤醒，或者有任务，退出循环
if (selectedKeys != 0 || oldWakenUp || wakenUp.get() || hasTasks() || hasScheduledTasks()) {
break;
}
if (Thread.interrupted()) {
// 线程被打断，退出循环
// 记录日志
selectCnt = 1;
break;
}
long time = System.nanoTime();
if (time - TimeUnit.MILLISECONDS.toNanos(timeoutMillis) >= currentTimeNanos) {
// 如果超时，计数重置为 1，下次循环就会 break
selectCnt = 1;
}
// 计数超过阈值，由 io.netty.selectorAutoRebuildThreshold 指定，默认 512
// 这是为了解决 nio 空轮询 bug
else if (SELECTOR_AUTO_REBUILD_THRESHOLD > 0 &&
selectCnt >= SELECTOR_AUTO_REBUILD_THRESHOLD) {
// 重建 selector
selector = selectRebuildSelector(selectCnt);
selectCnt = 1;
break;
}
currentTimeNanos = time;
}
if (selectCnt > MIN_PREMATURE_SELECTOR_RETURNS) {
// 记录日志
}
} catch (CancelledKeyException e) {
// 记录日志
}
}
处理 keys io.netty.channel.nio.NioEventLoop#processSelectedKeys
private void processSelectedKeys() {
if (selectedKeys != null) {
// 通过反射将 Selector 实现类中的就绪事件集合替换为 SelectedSelectionKeySet
// SelectedSelectionKeySet 底层为数组实现，可以提高遍历性能（原本为 HashSet）
processSelectedKeysOptimized();
} else {
processSelectedKeysPlain(selector.selectedKeys());
}
}
io.netty.channel.nio.NioEventLoop#processSelectedKey
private void processSelectedKey(SelectionKey k, AbstractNioChannel ch) {
final AbstractNioChannel.NioUnsafe unsafe = ch.unsafe();
// 当 key 取消或关闭时会导致这个 key 无效
if (!k.isValid()) {
// 无效时处理...
return;
}
try {
int readyOps = k.readyOps();
// 连接事件
if ((readyOps & SelectionKey.OP_CONNECT) != 0) {
int ops = k.interestOps();
ops &= ~SelectionKey.OP_CONNECT;
k.interestOps(ops);
unsafe.finishConnect();
}
// 可写事件
if ((readyOps & SelectionKey.OP_WRITE) != 0) {
ch.unsafe().forceFlush();
}
// 可读或可接入事件
if ((readyOps & (SelectionKey.OP_READ | SelectionKey.OP_ACCEPT)) != 0 || readyOps == 0) {
// 如果是可接入 io.netty.channel.nio.AbstractNioMessageChannel.NioMessageUnsafe#read
// 如果是可读 io.netty.channel.nio.AbstractNioByteChannel.NioByteUnsafe#read
unsafe.read();
}
} catch (CancelledKeyException ignored) {
unsafe.close(unsafe.voidPromise());
}
}
accept 剖析
nio 中如下代码，在 netty 中的流程
//1 阻塞直到事件发生
selector.select();
Iterator<SelectionKey> iter = selector.selectedKeys().iterator();
while (iter.hasNext()) {
//2 拿到一个事件
SelectionKey key = iter.next();
//3 如果是 accept 事件
if (key.isAcceptable()) {
//4 执行 accept
SocketChannel channel = serverSocketChannel.accept();
channel.configureBlocking(false);
//5 关注 read 事件
channel.register(selector, SelectionKey.OP_READ);
}
// ...
}
先来看可接入事件处理（accept）
io.netty.channel.nio.AbstractNioMessageChannel.NioMessageUnsafe#read
public void read() {
assert eventLoop().inEventLoop();
final ChannelConfig config = config();
final ChannelPipeline pipeline = pipeline();
final RecvByteBufAllocator.Handle allocHandle = unsafe().recvBufAllocHandle();
allocHandle.reset(config);
boolean closed = false;
Throwable exception = null;
try {
try {
do {
// doReadMessages 中执行了 accept 并创建 NioSocketChannel 作为消息放入 readBuf
// readBuf 是一个 ArrayList 用来缓存消息
int localRead = doReadMessages(readBuf);
if (localRead == 0) {
break;
}
if (localRead < 0) {
closed = true;
break;
}
// localRead 为 1，就一条消息，即接收一个客户端连接
allocHandle.incMessagesRead(localRead);
} while (allocHandle.continueReading());
} catch (Throwable t) {
exception = t;
}
int size = readBuf.size();
for (int i = 0; i < size; i ++) {
readPending = false;
// 触发 read 事件，让 pipeline 上的 handler 处理，这时是处理
// io.netty.bootstrap.ServerBootstrap.ServerBootstrapAcceptor#channelRead
pipeline.fireChannelRead(readBuf.get(i));
}
readBuf.clear();
allocHandle.readComplete();
pipeline.fireChannelReadComplete();
if (exception != null) {
closed = closeOnReadError(exception);
pipeline.fireExceptionCaught(exception);
}
if (closed) {
inputShutdown = true;
if (isOpen()) {
close(voidPromise());
}
}
} finally {
if (!readPending && !config.isAutoRead()) {
removeReadOp();
}
}
}
关键代码 io.netty.bootstrap.ServerBootstrap.ServerBootstrapAcceptor#channelRead
public void channelRead(ChannelHandlerContext ctx, Object msg) {
// 这时的 msg 是 NioSocketChannel
final Channel child = (Channel) msg;
// NioSocketChannel 添加 childHandler 即初始化器
child.pipeline().addLast(childHandler);
// 设置选项
setChannelOptions(child, childOptions, logger);
for (Entry<AttributeKey<?>, Object> e: childAttrs) {
child.attr((AttributeKey<Object>) e.getKey()).set(e.getValue());
}
try {
// 注册 NioSocketChannel 到 nio worker 线程，接下来的处理也移交至 nio worker 线程
childGroup.register(child).addListener(new ChannelFutureListener() {
@Override
public void operationComplete(ChannelFuture future) throws Exception {
if (!future.isSuccess()) {
forceClose(child, future.cause());
}
}
});
} catch (Throwable t) {
forceClose(child, t);
}
}
又回到了熟悉的 io.netty.channel.AbstractChannel.AbstractUnsafe#register 方法
public final void register(EventLoop eventLoop, final ChannelPromise promise) {
// 一些检查，略...
AbstractChannel.this.eventLoop = eventLoop;
if (eventLoop.inEventLoop()) {
register0(promise);
} else {
try {
// 这行代码完成的事实是 nio boss -> nio worker 线程的切换
eventLoop.execute(new Runnable() {
@Override
public void run() {
register0(promise);
}
});
} catch (Throwable t) {
// 日志记录...
closeForcibly();
closeFuture.setClosed();
safeSetFailure(promise, t);
}
}
}
io.netty.channel.AbstractChannel.AbstractUnsafe#register0
private void register0(ChannelPromise promise) {
try {
if (!promise.setUncancellable() || !ensureOpen(promise)) {
return;
}
boolean firstRegistration = neverRegistered;
doRegister();
neverRegistered = false;
registered = true;
// 执行初始化器，执行前 pipeline 中只有 head -> 初始化器 -> tail
pipeline.invokeHandlerAddedIfNeeded();
// 执行后就是 head -> logging handler -> my handler -> tail
safeSetSuccess(promise);
pipeline.fireChannelRegistered();
if (isActive()) {
if (firstRegistration) {
// 触发 pipeline 上 active 事件
pipeline.fireChannelActive();
} else if (config().isAutoRead()) {
beginRead();
}
}
} catch (Throwable t) {
closeForcibly();
closeFuture.setClosed();
safeSetFailure(promise, t);
}
}
回到了熟悉的代码 io.netty.channel.DefaultChannelPipeline.HeadContext#channelActive
public void channelActive(ChannelHandlerContext ctx) {
ctx.fireChannelActive();
// 触发 read (NioSocketChannel 这里 read，只是为了触发 channel 的事件注册，还未涉及数据读取)
readIfIsAutoRead();
}
io.netty.channel.nio.AbstractNioChannel#doBeginRead
protected void doBeginRead() throws Exception {
// Channel.read() or ChannelHandlerContext.read() was called
final SelectionKey selectionKey = this.selectionKey;
if (!selectionKey.isValid()) {
return;
}
readPending = true;
// 这时候 interestOps 是 0
final int interestOps = selectionKey.interestOps();
if ((interestOps & readInterestOp) == 0) {
// 关注 read 事件
selectionKey.interestOps(interestOps | readInterestOp);
}
}
read 剖析
再来看可读事件 io.netty.channel.nio.AbstractNioByteChannel.NioByteUnsafe#read，注意发送的数据未必能够一次读完，因此会触发多次 nio read 事件，一次事件内会触发多次 pipeline read，一次事件会触发一次 pipeline read complete
public final void read() {
final ChannelConfig config = config();
if (shouldBreakReadReady(config)) {
clearReadPending();
return;
}
final ChannelPipeline pipeline = pipeline();
// io.netty.allocator.type 决定 allocator 的实现
final ByteBufAllocator allocator = config.getAllocator();
// 用来分配 byteBuf，确定单次读取大小
final RecvByteBufAllocator.Handle allocHandle = recvBufAllocHandle();
allocHandle.reset(config);
ByteBuf byteBuf = null;
boolean close = false;
try {
do {
byteBuf = allocHandle.allocate(allocator);
// 读取
allocHandle.lastBytesRead(doReadBytes(byteBuf));
if (allocHandle.lastBytesRead() <= 0) {
byteBuf.release();
byteBuf = null;
close = allocHandle.lastBytesRead() < 0;
if (close) {
readPending = false;
}
break;
}
allocHandle.incMessagesRead(1);
readPending = false;
// 触发 read 事件，让 pipeline 上的 handler 处理，这时是处理 NioSocketChannel 上的 handler
pipeline.fireChannelRead(byteBuf);
byteBuf = null;
}
// 是否要继续循环
while (allocHandle.continueReading());
allocHandle.readComplete();
// 触发 read complete 事件
pipeline.fireChannelReadComplete();
if (close) {
closeOnRead(pipeline);
}
} catch (Throwable t) {
handleReadException(pipeline, byteBuf, t, close, allocHandle);
} finally {
if (!readPending && !config.isAutoRead()) {
removeReadOp();
}
}
}
io.netty.channel.DefaultMaxMessagesRecvByteBufAllocator.MaxMessageHandle#continueReading(io.netty.util.UncheckedBooleanSupplier)
public boolean continueReading(UncheckedBooleanSupplier maybeMoreDataSupplier) {
return
// 一般为 true
config.isAutoRead() &&
// respectMaybeMoreData 默认为 true
// maybeMoreDataSupplier 的逻辑是如果预期读取字节与实际读取字节相等，返回 true
(!respectMaybeMoreData || maybeMoreDataSupplier.get()) &&
// 小于最大次数，maxMessagePerRead 默认 16
totalMessages < maxMessagePerRead &&
// 实际读到了数据
totalBytesRead > 0;
}
