前言
本文從操作系統(tǒng)實(shí)際調(diào)用角度(以CentOS Linux release 7.5操作系統(tǒng)為示例),力求追根溯源看IO的每一步操作到底發(fā)生了什么。
關(guān)于如何查看系統(tǒng)調(diào)用,Linux可以使用 strace 來(lái)查看任何軟件的系統(tǒng)調(diào)動(dòng)(這是個(gè)很好的分析學(xué)習(xí)方法):strace -ff -o ./out java TestJava
一 BIO
- /**
- * Alipay.com Inc. Copyright (c) 2004-2020 All Rights Reserved.
- */
- package io;
- import java.io.*;
- import java.net.ServerSocket;
- import java.net.Socket;
- /**
- * @author xiangyong.ding
- * @version $Id: TestSocket.java, v 0.1 2020年08月02日 20:56 xiangyong.ding Exp $
- */
- public class BIOSocket {
- public static void main(String[] args) throws IOException {
- ServerSocket serverSocket = new ServerSocket(8090);
- System.out.println("step1: new ServerSocket ");
- while (true) {
- Socket client = serverSocket.accept();
- System.out.println("step2: client\t" + client.getPort());
- new Thread(() -> {
- try {
- InputStream in = client.getInputStream();
- BufferedReader reader = new BufferedReader(new InputStreamReader(in));
- while (true) {
- System.out.println(reader.readLine());
- }
- } catch (IOException e) {
- e.printStackTrace();
- }
- }).start();
- }
- }
- }
1 發(fā)生的系統(tǒng)調(diào)用
啟動(dòng)時(shí)
- socket(AF_INET, SOCK_STREAM, IPPROTO_IP) = 5
- bind(5, {sa_family=AF_INET, sin_port=htons(8090), sin_addr=inet_addr("0.0.0.0")}, 16) = 0
- listen(5, 50) = 0
- poll([{fd=5, events=POLLIN|POLLERR}], 1, -1) = 1 ([{fd=5, revents=POLLIN}])
poll函數(shù)會(huì)阻塞直到其中任何一個(gè)fd發(fā)生事件。
有客戶端連接后
- accept(5, {sa_family=AF_INET, sin_port=htons(10253), sin_addr=inet_addr("42.120.74.252")}, [16]) = 6
- clone(child_stack=0x7f013e5c4fb0, flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID, parent_tidptr=0x7f013e5c59d0, tls=0x7f013e5c5700, child_tidptr=0x7f013e5c59d0) = 13168
- poll([{fd=5, events=POLLIN|POLLERR}], 1, -1
拋出線程(即我們代碼里的 new Thread() )后,繼續(xù)poll阻塞等待連接。
clone出來(lái)的線程
- recvfrom(6, "hello,bio\n", 8192, 0, NULL, NULL) =
關(guān)于對(duì)recvfrom函數(shù)的說(shuō)明,其中第四個(gè)參數(shù)0 表示這是一個(gè)阻塞調(diào)用。
客戶端發(fā)送數(shù)據(jù)后
- recvfrom(6, "hello,bio\n", 8192, 0, NULL, NULL) = 10
2 優(yōu)缺點(diǎn)
優(yōu)點(diǎn)
代碼簡(jiǎn)單,邏輯清晰。
缺點(diǎn)
- 由于stream的read操作是阻塞讀,面對(duì)多個(gè)連接時(shí) 每個(gè)連接需要每線程。無(wú)法處理大量連接(C10K問(wèn)題)。
- 誤區(qū):可見(jiàn)JDK1.8中對(duì)于最初的BIO,在Linux OS下仍然使用的poll,poll本身也是相對(duì)比較高效的多路復(fù)用函數(shù)(支持非阻塞、多個(gè)socket同時(shí)檢查event),只是限于JDK最初的stream API限制,無(wú)法支持非阻塞讀取。
二 NIO(non block)
改進(jìn):使用NIO API,將阻塞變?yōu)榉亲枞?不需要大量線程。
- /**
- * Alipay.com Inc. Copyright (c) 2004-2020 All Rights Reserved.
- */
- package io;
- import java.io.IOException;
- import java.net.InetSocketAddress;
- import java.nio.ByteBuffer;
- import java.nio.channels.ServerSocketChannel;
- import java.nio.channels.SocketChannel;
- import java.util.LinkedList;
- /**
- * @author xiangyong.ding
- * @version $Id: NioSocket.java, v 0.1 2020年08月09日 11:25 xiangyong.ding Exp $
- */
- public class NIOSocket {
- private static LinkedList<SocketChannel> clients = new LinkedList<>();
- private static void startClientChannelHandleThread(){
- new Thread(() -> {
- while (true){
- ByteBuffer buffer = ByteBuffer.allocateDirect(4096);
- //處理客戶端連接
- for (SocketChannel c : clients) {
- // 非阻塞, >0 表示讀取到的字節(jié)數(shù)量, 0或-1表示未讀取到或讀取異常
- int num = 0;
- try {
- num = c.read(buffer);
- } catch (IOException e) {
- e.printStackTrace();
- }
- if (num > 0) {
- buffer.flip();
- byte[] clientBytes = new byte[buffer.limit()];
- //從緩沖區(qū) 讀取到內(nèi)存中
- buffer.get(clientBytes);
- System.out.println(c.socket().getPort() + ":" + new String(clientBytes));
- //清空緩沖區(qū)
- buffer.clear();
- }
- }
- }
- }).start();
- }
- public static void main(String[] args) throws IOException {
- //new socket,開(kāi)啟監(jiān)聽(tīng)
- ServerSocketChannel socketChannel = ServerSocketChannel.open();
- socketChannel.bind(new InetSocketAddress(9090));
- //設(shè)置阻塞接受客戶端連接
- socketChannel.configureBlocking(true);
- //開(kāi)始client處理線程
- startClientChannelHandleThread();
- while (true) {
- //接受客戶端連接; 非阻塞,無(wú)客戶端返回null(操作系統(tǒng)返回-1)
- SocketChannel client = socketChannel.accept();
- if (client == null) {
- //System.out.println("no client");
- } else {
- //設(shè)置讀非阻塞
- client.configureBlocking(false);
- int port = client.socket().getPort();
- System.out.println("client port :" + port);
- clients.add(client);
- }
- }
- }
- }
1 發(fā)生的系統(tǒng)調(diào)用
主線程
- socket(AF_INET, SOCK_STREAM, IPPROTO_IP) = 4
- bind(4, {sa_family=AF_INET, sin_port=htons(9090), sin_addr=inet_addr("0.0.0.0")}, 16) = 0
- listen(4, 50) = 0
- fcntl(4, F_SETFL, O_RDWR|O_NONBLOCK) = 0
- accept(4, 0x7fe26414e680, 0x7fe26c376710) = -1 EAGAIN (Resource temporarily unavailable)
有連接后,子線程
- read(6, 0x7f3f415b1c50, 4096) = -1 EAGAIN (Resource temporarily unavailable)
- read(6, 0x7f3f415b1c50, 4096) = -1 EAGAIN (Resource temporarily unavailable)
- ...
資源使用情況:
2 優(yōu)缺點(diǎn)
優(yōu)點(diǎn)
線程數(shù)大大減少。
缺點(diǎn)
需要程序自己掃描 每個(gè)連接read,需要 O(n)時(shí)間復(fù)雜度系統(tǒng)調(diào)用 (此時(shí)可能只有一個(gè)連接發(fā)送了數(shù)據(jù)),高頻系統(tǒng)調(diào)用(導(dǎo)致CPU 用戶態(tài)內(nèi)核態(tài)切換)高。導(dǎo)致CPU消耗很高。
三 多路復(fù)用器(select、poll、epoll)
改進(jìn):不需要用戶掃描所有連接,由kernel 給出哪些連接有數(shù)據(jù),然后應(yīng)用從有數(shù)據(jù)的連接讀取數(shù)據(jù)。
1 epoll
- import java.net.InetSocketAddress;
- import java.nio.ByteBuffer;
- import java.nio.channels.SelectionKey;
- import java.nio.channels.Selector;
- import java.nio.channels.ServerSocketChannel;
- import java.nio.channels.SocketChannel;
- import java.util.Iterator;
- import java.util.LinkedList;
- import java.util.Set;
- /**
- * 多路復(fù)用socket
- *
- * @author xiangyong.ding
- * @version $Id: MultiplexingSocket.java, v 0.1 2020年08月09日 12:19 xiangyong.ding Exp $
- */
- public class MultiplexingSocket {
- static ByteBuffer buffer = ByteBuffer.allocateDirect(4096);
- public static void main(String[] args) throws Exception {
- LinkedList<SocketChannel> clients = new LinkedList<>();
- //1.啟動(dòng)server
- //new socket,開(kāi)啟監(jiān)聽(tīng)
- ServerSocketChannel socketChannel = ServerSocketChannel.open();
- socketChannel.bind(new InetSocketAddress(9090));
- //設(shè)置非阻塞,接受客戶端
- socketChannel.configureBlocking(false);
- //多路復(fù)用器(JDK包裝的代理,select /poll/epoll/kqueue)
- Selector selector = Selector.open(); //java自動(dòng)代理,默認(rèn)為epoll
- //Selector selector = PollSelectorProvider.provider().openSelector();//指定為poll
- //將服務(wù)端socket 注冊(cè)到 多路復(fù)用器
- socketChannel.register(selector, SelectionKey.OP_ACCEPT);
- //2. 輪訓(xùn)多路復(fù)用器
- // 先詢問(wèn)有沒(méi)有連接,如果有則返回?cái)?shù)量以及對(duì)應(yīng)的對(duì)象(fd)
- while (selector.select() > 0) {
- System.out.println();
- Set<SelectionKey> selectionKeys = selector.selectedKeys();
- Iterator<SelectionKey> iter = selectionKeys.iterator();
- while (iter.hasNext()) {
- SelectionKey key = iter.next();
- iter.remove();
- //2.1 處理新的連接
- if (key.isAcceptable()) {
- //接受客戶端連接; 非阻塞,無(wú)客戶端返回null(操作系統(tǒng)返回-1)
- SocketChannel client = socketChannel.accept();
- //設(shè)置讀非阻塞
- client.configureBlocking(false);
- //同樣,把client也注冊(cè)到selector
- client.register(selector, SelectionKey.OP_READ);
- System.out.println("new client : " + client.getRemoteAddress());
- }
- //2.2 處理讀取數(shù)據(jù)
- else if (key.isReadable()) {
- readDataFromSocket(key);
- }
- }
- }
- }
- protected static void readDataFromSocket(SelectionKey key) throws Exception {
- SocketChannel socketChannel = (SocketChannel) key.channel();
- // 非阻塞, >0 表示讀取到的字節(jié)數(shù)量, 0或-1表示未讀取到或讀取異常
- // 請(qǐng)注意:這個(gè)例子降低復(fù)雜度,不考慮報(bào)文大于buffer size的情況
- int num = socketChannel.read(buffer);
- if (num > 0) {
- buffer.flip();
- byte[] clientBytes = new byte[buffer.limit()];
- //從緩沖區(qū) 讀取到內(nèi)存中
- buffer.get(clientBytes);
- System.out.println(socketChannel.socket().getPort() + ":" + new String(clientBytes));
- //清空緩沖區(qū)
- buffer.clear();
- }
- }
- }
2 發(fā)生的系統(tǒng)調(diào)用
啟動(dòng)
- socket(AF_INET, SOCK_STREAM, IPPROTO_IP) = 4
- bind(4, {sa_family=AF_INET, sin_port=htons(9090), sin_addr=inet_addr("0.0.0.0")}, 16) = 0
- listen(4, 50)
- fcntl(4, F_SETFL, O_RDWR|O_NONBLOCK) = 0
- epoll_create(256) = 7
- epoll_ctl(7, EPOLL_CTL_ADD, 5, {EPOLLIN, {u32=5, u64=4324783852322029573}}) = 0
- epoll_ctl(7, EPOLL_CTL_ADD, 4, {EPOLLIN, {u32=4, u64=158913789956}}) = 0
- epoll_wait(7
關(guān)于對(duì)epoll_create(對(duì)應(yīng)著Java的 Selector selector = Selector.open()) 的說(shuō)明,本質(zhì)上是在內(nèi)存的操作系統(tǒng)保留區(qū),創(chuàng)建一個(gè)epoll數(shù)據(jù)結(jié)構(gòu)。用于后面當(dāng)有client連接時(shí),向該epoll區(qū)中添加監(jiān)聽(tīng)。
有連接
- epoll_wait(7,[{EPOLLIN, {u32=4, u64=158913789956}}], 8192, -1) = 1
- accept(4, {sa_family=AF_INET, sin_port=htons(29597), sin_addr=inet_addr("42.120.74.252")}, [16]) = 8
- fcntl(8, F_SETFL, O_RDWR|O_NONBLOCK) = 0
- epoll_ctl(7, EPOLL_CTL_ADD, 8, {EPOLLIN, {u32=8, u64=3212844375897800712}}) = 0
關(guān)于epoll_ctl (對(duì)應(yīng)著Java的 client.register(selector, SelectionKey.OP_READ) )。其中 EPOLLIN 恰好對(duì)應(yīng)著Java的 SelectionKey.OP_READ 即監(jiān)聽(tīng)數(shù)據(jù)到達(dá)讀取事件。
客戶端發(fā)送數(shù)據(jù)
- epoll_wait(7,[{EPOLLIN, {u32=8, u64=3212844375897800712}}], 8192, -1) = 1
- read(8, "hello,multiplex\n", 4096) = 16
- epoll_wait(7,
note:epoll_wait第四個(gè)參數(shù)-1表示block。
poll 和 epoll 對(duì)比
根據(jù)“1.BIO”中的poll函數(shù)調(diào)用和epoll函數(shù)對(duì)比如下:
poll和epoll本質(zhì)上都是同步IO, 區(qū)別于BIO的是 多路復(fù)用充分降低了 system call,而epoll更進(jìn)一步,再次降低了system call的時(shí)間復(fù)雜度。
3 優(yōu)缺點(diǎn)
優(yōu)點(diǎn)
- 線程數(shù)同樣很少,甚至可以把a(bǔ)cceptor線程和worker線程使用同一個(gè)。
- 時(shí)間復(fù)雜度低,Java實(shí)現(xiàn)的Selector(在Linux OS下使用的epoll函數(shù))支持多個(gè)clientChannel事件的一次性獲取,且時(shí)間復(fù)雜度維持在O(1)。
- CPU使用低:得益于Selector,我們不用向 “2.NIO”中需要自己一個(gè)個(gè)ClientChannel手動(dòng)去檢查事件,因此使得CPU使用率大大降低。
缺點(diǎn)
- 數(shù)據(jù)處理麻煩:目前socketChannel.read 讀取數(shù)據(jù)完全是基于字節(jié)的,當(dāng)我們需要需要作為HTTP服務(wù)網(wǎng)關(guān)時(shí),對(duì)于HTTP協(xié)議的處理完全需要自己解析,這是個(gè)龐大、煩雜、容易出錯(cuò)的工作。
-
性能
- 現(xiàn)有socket數(shù)據(jù)的讀取(socketChannel.read(buffer))全部通過(guò)一個(gè)buffer 緩沖區(qū)來(lái)接受,一旦連接多起來(lái),這無(wú)疑是一個(gè)單線程讀取,性能無(wú)疑是個(gè)問(wèn)題。
- 那么此時(shí)buffer我們每次讀取都重新new出來(lái)呢?如果每次都new出來(lái),這樣的內(nèi)存碎片對(duì)于GC無(wú)疑是一場(chǎng)災(zāi)難。如何平衡地協(xié)調(diào)好buffer的共享,既保證性能,又保證線程安全,這是個(gè)難題。
四 Netty
1 研究的目標(biāo)源碼(netty提供的入門example)
TelnetServer
- package telnet;
- import io.netty.bootstrap.ServerBootstrap;
- import io.netty.channel.EventLoopGroup;
- import io.netty.channel.nio.NioEventLoopGroup;
- import io.netty.channel.socket.nio.NioServerSocketChannel;
- import io.netty.handler.logging.LogLevel;
- import io.netty.handler.logging.LoggingHandler;
- import io.netty.handler.ssl.SslContext;
- import io.netty.handler.ssl.SslContextBuilder;
- import io.netty.handler.ssl.util.SelfSignedCertificate;
- /**
- * Simplistic telnet server.
- */
- public final class TelnetServer {
- static final boolean SSL = System.getProperty("ssl") != null;
- static final int PORT = Integer.parseInt(System.getProperty("port", SSL? "8992" : "8023"));
- public static void main(String[] args) throws Exception {
- // Configure SSL.
- final SslContext sslCtx;
- if (SSL) {
- SelfSignedCertificate ssc = new SelfSignedCertificate();
- sslCtx = SslContextBuilder.forServer(ssc.certificate(), ssc.privateKey()).build();
- } else {
- sslCtx = null;
- }
- EventLoopGroup bossGroup = new NioEventLoopGroup(1);
- EventLoopGroup workerGroup = new NioEventLoopGroup();
- try {
- ServerBootstrap b = new ServerBootstrap();
- b.group(bossGroup, workerGroup)
- .channel(NioServerSocketChannel.class)
- .handler(new LoggingHandler(LogLevel.INFO))
- .childHandler(new TelnetServerInitializer(sslCtx));
- b.bind(PORT).sync().channel().closeFuture().sync();
- } finally {
- bossGroup.shutdownGracefully();
- workerGroup.shutdownGracefully();
- }
- }
- }
TelnetServerHandler
- package telnet;
- import io.netty.channel.ChannelFuture;
- import io.netty.channel.ChannelFutureListener;
- import io.netty.channel.ChannelHandler.Sharable;
- import io.netty.channel.ChannelHandlerContext;
- import io.netty.channel.SimpleChannelInboundHandler;
- import java.net.InetAddress;
- import java.util.Date;
- /**
- * Handles a server-side channel.
- */
- @Sharable
- public class TelnetServerHandler extends SimpleChannelInboundHandler<String> {
- @Override
- public void channelActive(ChannelHandlerContext ctx) throws Exception {
- // Send greeting for a new connection.
- ctx.write("Welcome to " + InetAddress.getLocalHost().getHostName() + "!\r\n");
- ctx.write("It is " + new Date() + " now.\r\n");
- ctx.flush();
- }
- @Override
- public void channelRead0(ChannelHandlerContext ctx, String request) throws Exception {
- // Generate and write a response.
- String response;
- boolean close = false;
- if (request.isEmpty()) {
- response = "Please type something.\r\n";
- } else if ("bye".equals(request.toLowerCase())) {
- response = "Have a good day!\r\n";
- close = true;
- } else {
- response = "Did you say '" + request + "'?\r\n";
- }
- // We do not need to write a ChannelBuffer here.
- // We know the encoder inserted at TelnetPipelineFactory will do the conversion.
- ChannelFuture future = ctx.write(response);
- // Close the connection after sending 'Have a good day!'
- // if the client has sent 'bye'.
- if (close) {
- future.addListener(ChannelFutureListener.CLOSE);
- }
- }
- @Override
- public void channelReadComplete(ChannelHandlerContext ctx) {
- ctx.flush();
- }
- @Override
- public void exceptionCaught(ChannelHandlerContext ctx, Throwable cause) {
- cause.printStackTrace();
- ctx.close();
- }
- }
TelnetServerInitializer
- package telnet;
- import io.netty.channel.ChannelInitializer;
- import io.netty.channel.ChannelPipeline;
- import io.netty.channel.socket.SocketChannel;
- import io.netty.handler.codec.DelimiterBasedFrameDecoder;
- import io.netty.handler.codec.Delimiters;
- import io.netty.handler.codec.string.StringDecoder;
- import io.netty.handler.codec.string.StringEncoder;
- import io.netty.handler.ssl.SslContext;
- /**
- * Creates a newly configured {@link ChannelPipeline} for a new channel.
- */
- public class TelnetServerInitializer extends ChannelInitializer<SocketChannel> {
- private static final StringDecoder DECODER = new StringDecoder();
- private static final StringEncoder ENCODER = new StringEncoder();
- private static final TelnetServerHandler SERVER_HANDLER = new TelnetServerHandler();
- private final SslContext sslCtx;
- public TelnetServerInitializer(SslContext sslCtx) {
- this.sslCtx = sslCtx;
- }
- @Override
- public void initChannel(SocketChannel ch) throws Exception {
- ChannelPipeline pipeline = ch.pipeline();
- if (sslCtx != null) {
- pipeline.addLast(sslCtx.newHandler(ch.alloc()));
- }
- // Add the text line codec combination first,
- pipeline.addLast(new DelimiterBasedFrameDecoder(8192, Delimiters.lineDelimiter()));
- // the encoder and decoder are static as these are sharable
- pipeline.addLast(DECODER);
- pipeline.addLast(ENCODER);
- // and then business logic.
- pipeline.addLast(SERVER_HANDLER);
- }
- }
2 啟動(dòng)后的系統(tǒng)調(diào)用
主線程(23109)
- ## 256無(wú)實(shí)際作用,這里只為了兼容舊版kernel api
- epoll_create(256) = 7epoll_ctl(7, EPOLL_CTL_ADD, 5, {EPOLLIN, {u32=5, u64=5477705356928876549}}) = 0
- epoll_create(256) = 10epoll_ctl(10, EPOLL_CTL_ADD, 8, {EPOLLIN, {u32=8, u64=17041805914081853448}}) = 0
- epoll_create(256) = 13
- epoll_ctl(13, EPOLL_CTL_ADD, 11, {EPOLLIN, {u32=11, u64=17042151607409573899}}) = 0
- epoll_create(256) = 16
- epoll_ctl(16, EPOLL_CTL_ADD, 14, {EPOLLIN, {u32=14, u64=17042497300737294350}}) = 0
- epoll_create(256) = 19
- epoll_ctl(19, EPOLL_CTL_ADD, 17, {EPOLLIN, {u32=17, u64=17042561450368827409}}) = 0
- epoll_create(256) = 10
- socket(AF_INET, SOCK_STREAM, IPPROTO_IP) = 20
- clone(child_stack=0x7fc3c509afb0, flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID, parent_tidptr=0x7fc3c509b9d0, tls=0x7fc3c509b700, child_tidptr=0x7fc3c509b9d0) = 23130
概括為:
- 向OS新建socket,并開(kāi)啟clone boss線程23130。
- 為BOSS創(chuàng)建了一個(gè)epoll(論證參見(jiàn)下面“boss”),每個(gè)worker創(chuàng)建一個(gè)epoll數(shù)據(jù)結(jié)構(gòu)(本質(zhì)上是在kernel內(nèi)存區(qū)創(chuàng)建了一個(gè)數(shù)據(jù)結(jié)構(gòu),用于后續(xù)監(jiān)聽(tīng))。
- 創(chuàng)建boss線程監(jiān)聽(tīng)的socket(本質(zhì)上在kernel中創(chuàng)建一個(gè)數(shù)據(jù)結(jié)構(gòu))。
boss(23130)
- bind(20, {sa_family=AF_INET, sin_port=htons(8023), sin_addr=inet_addr("0.0.0.0")}, 16) = 0
- listen(20, 128) = 0
- getsockname(20, {sa_family=AF_INET, sin_port=htons(8023), sin_addr=inet_addr("0.0.0.0")}, [16]) = 0
- getsockname(20, {sa_family=AF_INET, sin_port=htons(8023), sin_addr=inet_addr("0.0.0.0")}, [16]) = 0
- ##將fd為7號(hào)epoll和fd為20號(hào)的socket綁定,事件:epoll_ctl_add和epoll_ctl_mod
- epoll_ctl(7, EPOLL_CTL_ADD, 20, {EPOLLIN, {u32=20, u64=14198059139132817428}}) = 0
- epoll_ctl(7, EPOLL_CTL_MOD, 20, {EPOLLIN, {u32=20, u64=20}}) = 0
- epoll_wait(7, [{EPOLLIN, {u32=5, u64=17295150779149058053}}], 8192, 1000) = 1
- epoll_wait(7, [], 8192, 1000) = 0(不斷輪訓(xùn),1S超時(shí)一次)
概括為:
- 將上一步中main線程創(chuàng)建的fd:20綁定端口8023,并開(kāi)啟監(jiān)聽(tīng)(網(wǎng)卡負(fù)責(zé)監(jiān)聽(tīng)和接受連接和數(shù)據(jù),kernel則負(fù)責(zé)路由到具體進(jìn)程,具體參見(jiàn):關(guān)于socket和bind和listen,TODO )。
- 將7號(hào)socket對(duì)應(yīng)的fd綁定到20號(hào)對(duì)應(yīng)的epoll數(shù)據(jù)結(jié)構(gòu)上去(都是操作kernel中的內(nèi)存)。
- 開(kāi)始1S中一次阻塞等待epoll有任何連接或數(shù)據(jù)到達(dá)。
3 客戶端連接
boss (23130)
- accept(20, {sa_family=AF_INET, sin_port=htons(11144), sin_addr=inet_addr("42.120.74.122")}, [16]) = 24
- getsockname(24, {sa_family=AF_INET, sin_port=htons(8023), sin_addr=inet_addr("192.168.0.120")}, [16]) = 0
- getsockname(24, {sa_family=AF_INET, sin_port=htons(8023), sin_addr=inet_addr("192.168.0.120")}, [16]) = 0
- setsockopt(24, SOL_TCP, TCP_NODELAY, [1], 4) = 0
- getsockopt(24, SOL_SOCKET, SO_SNDBUF, [87040], [4]) = 0
- getsockopt(24, SOL_SOCKET, SO_SNDBUF, [87040], [4]) = 0
- ##拋出 work線程
- clone(child_stack=0x7fc3c4c98fb0, flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID, parent_tidptr=0x7fc3c4c999d0, tls=0x7fc3c4c99700, child_tidptr=0x7fc3c4c999d0) = 2301
worker (2301)
- writev(24, [{"Welcome to iZbp14e1g9ztpshfrla9m"..., 37}, {"It is Sun Aug 23 15:44:14 CST 20"..., 41}], 2) = 78
- epoll_ctl(13, EPOLL_CTL_ADD, 24, {EPOLLIN, {u32=24, u64=24}}) = 0
- epoll_ctl(13, EPOLL_CTL_MOD, 24, {EPOLLIN, {u32=24, u64=14180008216221450264}}) = 0
- epoll_wait(13, [{EPOLLIN, {u32=11, u64=17042151607409573899}}], 8192, 1000) = 1
- read(11, "\1", 128) = 1
- ##開(kāi)始無(wú)限loop
- epoll_wait(13, [], 8192, 1000) = 0
- epoll_wait(13, [{EPOLLIN, {u32=24, u64=24}}], 8192, 1000) = 1
概括:
- 當(dāng)BOSS輪訓(xùn)epoll_wait等到了連接后,首先accept得到該socket對(duì)應(yīng)的fd。
- 連接建立后 BOSS立馬拋出一個(gè)線程(clone函數(shù))。
- worker(即新建的線程)寫入了一段數(shù)據(jù)(這里是業(yè)務(wù)邏輯)。
- worker將該client對(duì)應(yīng)的fd綁定到了13號(hào)epoll上。
- worker繼續(xù)輪訓(xùn)監(jiān)聽(tīng)13號(hào)epoll。
4 客戶端主動(dòng)發(fā)送數(shù)據(jù)
worker(2301)
- read(24, "i am daojian\r\n", 1024) = 14
- write(24, "Did you say 'i am daojian'?\r\n", 29) = 29
- ##繼續(xù)無(wú)限loop
- epoll_wait(13, [], 8192, 1000) = 0
概括為:
- wait到數(shù)據(jù)后,立即read到用戶控件內(nèi)存中(讀取1024個(gè)字節(jié)到 用戶控件某個(gè)buff中)。
- 寫入數(shù)據(jù)(業(yè)務(wù)邏輯,不必太關(guān)注)。
- 繼續(xù)輪訓(xùn)等待13號(hào)epoll。
5 客戶端發(fā)送bye報(bào)文,服務(wù)器斷開(kāi)TCP連接
worker(2301)
- read(24, "bye\r\n", 1024) = 5
- write(24, "Have a good day!\r\n", 18) = 18
- getsockopt(24, SOL_SOCKET, SO_LINGER, {onoff=0, linger=0}, [8]) = 0
- dup2(25, 24) = 24
- ##從epoll數(shù)據(jù)結(jié)構(gòu)中(OS)中刪除fd為24的socket
- epoll_ctl(13, EPOLL_CTL_DEL, 24, 0x7f702dd531e0) = -1 ENOENT
- ##關(guān)閉24 socket
- close(24) = 0
- ##繼續(xù)等待13 epoll數(shù)據(jù)
- epoll_wait(13, [], 8192, 1000) = 0
斷開(kāi)客戶端連接概括為:
- 從epoll中刪除該客戶端對(duì)應(yīng)的fd(這里觸發(fā)源頭沒(méi)找到,可能是boss)。
- close關(guān)閉客戶端24號(hào)fd。
- 繼續(xù)輪訓(xùn)epoll。
6 五個(gè)客戶端同時(shí)連接
boss線程(23130)
- accept(20, {sa_family=AF_INET, sin_port=htons(1846), sin_addr=inet_addr("42.120.74.122")}, [16]) = 24
- clone(child_stack=0x7f702cc51fb0, flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID, parent_tidptr=0x7f702cc529d0, tls=0x7f702cc52700, child_tidptr=0x7f702cc529d0) = 10035
- accept(20, {sa_family=AF_INET, sin_port=htons(42067), sin_addr=inet_addr("42.120.74.122")}, [16]) = 26
- clone(child_stack=0x7f702cb50fb0, flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID, parent_tidptr=0x7f702cb519d0, tls=0x7f702cb51700, child_tidptr=0x7f702cb519d0) = 10067
- ...
woker線程(10035,第一個(gè)連接)
- epoll_ctl(13, EPOLL_CTL_ADD, 24, {EPOLLIN, {u32=24, u64=24}}) = 0
- epoll_ctl(13, EPOLL_CTL_MOD, 24, {EPOLLIN, {u32=24, u64=3226004877247250456}}) = 0
- epoll_wait(13, [{EPOLLIN, {u32=11, u64=17042151607409573899}}], 8192, 1000) = 1 = 1
- epoll_wait(13, [], 8192, 1000) = 0
worker線程(10067,第二個(gè)連接)
- epoll_ctl(16, EPOLL_CTL_ADD, 26, {EPOLLIN, {u32=26, u64=26}}) = 0
- epoll_ctl(16, EPOLL_CTL_MOD, 26, {EPOLLIN, {u32=26, u64=3221483685433835546}}) = 0
- epoll_wait(16, [{EPOLLIN, {u32=14, u64=17042497300737294350}}], 8192, 1000) = 1
- epoll_wait(16, [], 8192, 1000) = 0
- epoll_wait(16, [], 8192, 1000) = 0
worker線程(10067,第二個(gè)連接)
- epoll_ctl(19, EPOLL_CTL_ADD, 27, {EPOLLIN, {u32=27, u64=27}}) = 0
- epoll_ctl(19, EPOLL_CTL_MOD, 27, {EPOLLIN, {u32=27, u64=3216966479350071323}}) = 0
worker線程(8055,第四個(gè)連接)
- epoll_ctl(10, EPOLL_CTL_ADD, 28, {EPOLLIN, {u32=28, u64=28}}) = 0
- epoll_ctl(10, EPOLL_CTL_MOD, 28, {EPOLLIN, {u32=28, u64=3302604828697427996}}) = 0
worker線程(10035,第五個(gè)連接,不在clone線程,而是復(fù)用了第一個(gè)epoll對(duì)應(yīng)的worker)
- epoll_ctl(13, EPOLL_CTL_ADD, 29, {EPOLLIN, {u32=29, u64=29}}) = 0
- epoll_ctl(13, EPOLL_CTL_MOD, 29, {EPOLLIN, {u32=29, u64=29}}) = 0
概括為:
- epoll和boss、worker之間的關(guān)系:一共有4個(gè)worker對(duì)應(yīng)著4個(gè)epoll對(duì)象,boss和每個(gè)worker都有對(duì)應(yīng)自己的epoll。
- boss根據(jù)epoll數(shù)量,平衡分配連接到每個(gè)worker對(duì)應(yīng)的epoll中。
7 總結(jié)
下圖通過(guò)對(duì)系統(tǒng)調(diào)用的調(diào)查得出 netty 和 kernel 交互圖:
初始化直接創(chuàng)建5個(gè)epoll,其中7號(hào)為boss使用,專門用于處理和客戶端連接;其余4個(gè)用來(lái)給worker使用,用戶處理和客戶端的數(shù)據(jù)交互。
work的線程數(shù)量,取決于初始化時(shí)創(chuàng)建了幾個(gè)epoll,worker的復(fù)用本質(zhì)上是epoll的復(fù)用。
work之間為什么要獨(dú)立使用epoll?為什么不共享?
- 為了避免各個(gè)worker之間發(fā)生爭(zhēng)搶連接處理,netty直接做了物理隔離,避免競(jìng)爭(zhēng)。各個(gè)worker只負(fù)責(zé)處理自己管理的連接,并且后續(xù)該worker中的每個(gè)client的讀寫操作完全由 該線程單獨(dú)處理,天然避免了資源競(jìng)爭(zhēng),避免了鎖。
- worker單線程,性能考慮:worker不僅僅要epoll_wait,還是處理read、write邏輯,加入worker處理了過(guò)多的連接,勢(shì)必造成這部分消耗時(shí)間片過(guò)多,來(lái)不及處理更多連接,性能下降。
8 優(yōu)缺點(diǎn)
優(yōu)點(diǎn)
- 數(shù)據(jù)處理:netty提供了大量成熟的數(shù)據(jù)處理組件(ENCODER、DECODER),HTTP、POP3拿來(lái)即用。
- 編碼復(fù)雜度、可維護(hù)性:netty充分使得業(yè)務(wù)邏輯與網(wǎng)絡(luò)處理解耦,只需要少量的BootStrap配置即可,更多的集中在業(yè)務(wù)邏輯處理上。
- 性能:netty提供了的ByteBuf(底層Java原生的ByteBuffer),提供了池化的ByteBuf,兼顧讀取性能和ByteBuf內(nèi)存分配(在后續(xù)文檔中會(huì)再做詳解)。
缺點(diǎn)
- 入門有一定難度。
五 AIO
1 啟動(dòng)
main線程
- epoll_create(256) = 5
- epoll_ctl(5, EPOLL_CTL_ADD, 6, {EPOLLIN, {u32=6, u64=11590018039084482566}}) = 0
- ##創(chuàng)建BOSS 線程(Proactor)
- clone(child_stack=0x7f340ac06fb0, flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID, parent_tidptr=0x7f340ac079d0, tls=0x7f340ac07700, child_tidptr=0x7f340ac079d0) = 22704
- socket(AF_INET6, SOCK_STREAM, IPPROTO_IP) = 8
- setsockopt(8, SOL_IPV6, IPV6_V6ONLY, [0], 4) = 0
- setsockopt(8, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
- bind(8, {sa_family=AF_INET6, sin6_port=htons(9090), inet_pton(AF_INET6, "::", &sin6_addr), sin6_flowinfo=0, sin6_scope_id=0}, 28) = 0
- listen(8, 50)
- accept(8, 0x7f67d01b3120, 0x7f67d9246690) = -1
- epoll_ctl(5, EPOLL_CTL_MOD, 8, {EPOLLIN|EPOLLONESHOT, {u32=8, u64=15380749440025362440}}) = -1 ENOENT (No such file or directory)
- epoll_ctl(5, EPOLL_CTL_ADD, 8, {EPOLLIN|EPOLLONESHOT, {u32=8, u64=15380749440025362440}}) = 0
- read(0,
22704(BOSS 線程(Proactor))
- epoll_wait(5, <unfinished ...>
2 請(qǐng)求連接
22704(BOSS 線程(Proactor))處理連接
- epoll_wait(5,[{EPOLLIN, {u32=9, u64=4294967305}}], 512, -1) = 1
- accept(8, {sa_family=AF_INET6, sin6_port=htons(55320), inet_pton(AF_INET6, "::ffff:36.24.32.140", &sin6_addr), sin6_flowinfo=0, sin6_scope_id=0}, [28]) = 9
- clone(child_stack=0x7ff35c99ffb0, flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID, parent_tidptr=0x7ff35c9a09d0, tls=0x7ff35c9a0700, child_tidptr=0x7ff35c9a09d0) = 26241
- epoll_wait(5, <unfinished ...>
26241
- #將client 連接的FD加入到BOSS的epoll中,以便BOSS線程監(jiān)聽(tīng)網(wǎng)絡(luò)事件
- epoll_ctl(5, EPOLL_CTL_MOD, 9, {EPOLLIN|EPOLLONESHOT, {u32=9, u64=4398046511113}}) = -1 ENOENT (No such file or directory)
- epoll_ctl(5, EPOLL_CTL_ADD, 9, {EPOLLIN|EPOLLONESHOT, {u32=9, u64=4398046511113}}) = 0
- accept(8, 0x7ff3440008c0, 0x7ff35c99f4d0) = -1 EAGAIN (Resource temporarily unavailable)
- epoll_ctl(5, EPOLL_CTL_MOD, 8, {EPOLLIN|EPOLLONESHOT, {u32=8, u64=8}}) = 0
3 客戶端發(fā)送數(shù)據(jù)
22704(BOSS 線程(Proactor))處理連接
- epoll_wait(5,[{EPOLLIN, {u32=9, u64=4294967305}}], 512, -1) = 1
- ##數(shù)據(jù)讀出
- read(9, "daojian111\r\n", 1024) = 12
- ##數(shù)據(jù)處理交給其他線程,這里由于線程池為空,需要先clone線程
- clone(child_stack=0x7ff35c99ffb0, flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID, parent_tidptr=0x7ff35c9a09d0, tls=0x7ff35c9a0700, child_tidptr=0x7ff35c9a09d0) = 26532
復(fù)制線程處理,線程號(hào)26532
- write(1, "pool-1-thread-2-10received : dao"..., 41) = 41
- write(1, "\n", 1)
- accept(8, 0x7f11c400b5f0, 0x7f11f42fd4d0) = -1 EAGAIN (Resource temporarily unavailable)
- epoll_ctl(5, EPOLL_CTL_MOD, 8, {EPOLLIN|EPOLLONESHOT, {u32=8, u64=8}}) = 0
4 總結(jié)
- 從系統(tǒng)調(diào)用角度,Java的AIO事實(shí)上是以多路復(fù)用(Linux上為epoll)等同步IO為基礎(chǔ),自行實(shí)現(xiàn)了異步事件分發(fā)。
- BOSS Thread負(fù)責(zé)處理連接,并分發(fā)事件。
- WORKER Thread只負(fù)責(zé)從BOSS接收的事件執(zhí)行,不負(fù)責(zé)任何網(wǎng)絡(luò)事件監(jiān)聽(tīng)。
5 優(yōu)缺點(diǎn)
優(yōu)點(diǎn)
相比于前面的BIO、NIO,AIO已經(jīng)封裝好了任務(wù)調(diào)度,使用時(shí)只需關(guān)心任務(wù)處理。
缺點(diǎn)
- 事件處理完全由Thread Pool完成,對(duì)于同一個(gè)channel的多個(gè)事件可能會(huì)出現(xiàn)并發(fā)問(wèn)題。
- 相比netty,buffer API不友好容易出錯(cuò);編解碼工作復(fù)雜。
原文鏈接:https://zhuanlan.51cto.com/art/202106/669120.htm