前面几节我们讲解的都是RocketMQ应用层面也有一些关于源码的分析讲解，消息的存储只是简单的分析没有深入的探讨其中的内部机制，我们都知道RocketMQ是基于文件存储的消息模型，文件存储给我们的直觉感受就是比较慢，为什么RocketMQ在消息的生产及消费都在毫秒级，真如我们所想象的文件存储就很慢吗？本节分析下消息的内部内存映射机制。

1、概述

我们首先看一下消息的发送流程图

image

前面我们在RocketMQ事务消息章节中消息发送中各个类的业务流程，我们只是分析到CommitLog.putMessage()方法，下面我们接着探讨。上图中就是该消息方法的流程。

public PutMessageResult putMessage(final MessageExtBrokerInner msg) {
    //设置消息存储到文件中的时间
    msg.setStoreTimestamp(System.currentTimeMillis());
    //设置消息的校验码CRC
    msg.setBodyCRC(UtilAll.crc32(msg.getBody()));
    AppendMessageResult result = null;
    StoreStatsService storeStatsService = this.defaultMessageStore.getStoreStatsService();

    String topic = msg.getTopic();
    int queueId = msg.getQueueId();

    final int tranType = MessageSysFlag.getTransactionValue(msg.getSysFlag());
    if (tranType == MessageSysFlag.TRANSACTION_NOT_TYPE
        || tranType == MessageSysFlag.TRANSACTION_COMMIT_TYPE) {
        // Delay Delivery消息的延迟级别是否大于0
        if (msg.getDelayTimeLevel() > 0) {
            //如果消息的延迟级别大于最大的延迟级别则置为最大延迟级别
            if (msg.getDelayTimeLevel() > this.defaultMessageStore.getScheduleMessageService().getMaxDelayLevel()) {
                msg.setDelayTimeLevel(this.defaultMessageStore.getScheduleMessageService().getMaxDelayLevel());
            }
            //将消息主题设置为SCHEDULE_TOPIC_XXXX
            topic = ScheduleMessageService.SCHEDULE_TOPIC;
            //将消息队列设置为延迟的消息队列的ID
            queueId = ScheduleMessageService.delayLevel2QueueId(msg.getDelayTimeLevel());
            //消息的原有的主题和消息队列存入属性中
            MessageAccessor.putProperty(msg, MessageConst.PROPERTY_REAL_TOPIC, msg.getTopic());
            MessageAccessor.putProperty(msg, MessageConst.PROPERTY_REAL_QUEUE_ID, String.valueOf(msg.getQueueId()));
            msg.setPropertiesString(MessageDecoder.messageProperties2String(msg.getProperties()));
            msg.setTopic(topic);
            msg.setQueueId(queueId);
        }
    }

    long eclipseTimeInLock = 0;
    MappedFile unlockMappedFile = null;
    //获取最后一个消息的映射文件，mappedFileQueue可看作是CommitLog文件夹下的一个个文件的映射
    MappedFile mappedFile = this.mappedFileQueue.getLastMappedFile();

    //写入消息之前先申请putMessageLock，也就是保证消息写入CommitLog文件中串行的
    putMessageLock.lock(); //spin or ReentrantLock ,depending on store config
    try {
        long beginLockTimestamp = this.defaultMessageStore.getSystemClock().now();
        this.beginTimeInLock = beginLockTimestamp;

        //设置消息的存储时间
        msg.setStoreTimestamp(beginLockTimestamp);
        //mappedFile==null标识CommitLog文件还未创建，第一次存消息则创建CommitLog文件
        //mappedFile.isFull()表示mappedFile文件已满，需要重新创建CommitLog文件
        if (null == mappedFile || mappedFile.isFull()) {
            //里面的参数0代表偏移量
            mappedFile = this.mappedFileQueue.getLastMappedFile(0); // Mark: NewFile may be cause noise
        }
        //mappedFile==null说明创建CommitLog文件失败抛出异常，创建失败可能是磁盘空间不足或者权限不够
        if (null == mappedFile) {
            log.error("create mapped file1 error, topic: " + msg.getTopic() + " clientAddr: " + msg.getBornHostString());
            beginTimeInLock = 0;
            return new PutMessageResult(PutMessageStatus.CREATE_MAPEDFILE_FAILED, null);
        }
        //mappedFile文件后面追加消息
        result = mappedFile.appendMessage(msg, this.appendMessageCallback);
        switch (result.getStatus()) {
            case PUT_OK:
                break;
            case END_OF_FILE:
                unlockMappedFile = mappedFile;
                // Create a new file, re-write the message
                mappedFile = this.mappedFileQueue.getLastMappedFile(0);
                if (null == mappedFile) {
                    // XXX: warn and notify me
                    log.error("create mapped file2 error, topic: " + msg.getTopic() + " clientAddr: " + msg.getBornHostString());
                    beginTimeInLock = 0;
                    return new PutMessageResult(PutMessageStatus.CREATE_MAPEDFILE_FAILED, result);
                }
                result = mappedFile.appendMessage(msg, this.appendMessageCallback);
                break;
            case MESSAGE_SIZE_EXCEEDED:
            case PROPERTIES_SIZE_EXCEEDED:
                beginTimeInLock = 0;
                return new PutMessageResult(PutMessageStatus.MESSAGE_ILLEGAL, result);
            case UNKNOWN_ERROR:
                beginTimeInLock = 0;
                return new PutMessageResult(PutMessageStatus.UNKNOWN_ERROR, result);
            default:
                beginTimeInLock = 0;
                return new PutMessageResult(PutMessageStatus.UNKNOWN_ERROR, result);
        }

        eclipseTimeInLock = this.defaultMessageStore.getSystemClock().now() - beginLockTimestamp;
        beginTimeInLock = 0;
    } finally {
        //释放锁
        putMessageLock.unlock();
    }

    if (eclipseTimeInLock > 500) {
        log.warn("[NOTIFYME]putMessage in lock cost time(ms)={}, bodyLength={} AppendMessageResult={}", eclipseTimeInLock, msg.getBody().length, result);
    }

    if (null != unlockMappedFile && this.defaultMessageStore.getMessageStoreConfig().isWarmMapedFileEnable()) {
        this.defaultMessageStore.unlockMappedFile(unlockMappedFile);
    }

    PutMessageResult putMessageResult = new PutMessageResult(PutMessageStatus.PUT_OK, result);

    // Statistics
    storeStatsService.getSinglePutMessageTopicTimesTotal(msg.getTopic()).incrementAndGet();
    storeStatsService.getSinglePutMessageTopicSizeTotal(topic).addAndGet(result.getWroteBytes());
    //消息刷盘
    handleDiskFlush(result, putMessageResult, msg);
    //主从数据同步复制
    handleHA(result, putMessageResult, msg);
    return putMessageResult;
}

本章节我们重点分析三点，关于刷盘机制后面章节会介绍

1. 获取映射文件MappedFile
2. 创建映射文件MappedFile
3. 映射文件中写入消息

2、获取映射文件MappedFile

2.1、MappedFile和Commitlog的关系

我们知道消息的存储文件时Commitlog文件中，那与内存映射对象MappedFile又有什么联系呢？

image

从图中我们可以清晰的知道他们之间的关系，每个MappedFile对象对于一个Commitlog文件，我们分析下这个对应关系的业务操作发生在什么时候，我们分析下源码

Broker服务启动时会创建BrokerController对象并对其初始化initialize()该方法调用DefaultMessageStore.load()方法加载Commitlog文件和消费队列文件

public boolean load() {
    //省略代码...
    // 加载Commitlog文件
    result = result && this.commitLog.load();
    // 加载消费队列文件
    result = result && this.loadConsumeQueue();
    //省略代码...
}

我们分析下commitLog.load()调用mappedFileQueue.load()

public boolean load() {
    //消息存储路径
    File dir = new File(this.storePath);
    File[] files = dir.listFiles();
    if (files != null) {
        // 升序
        Arrays.sort(files);
        for (File file : files) {
            if (file.length() != this.mappedFileSize) {
                log.warn(file + "\t" + file.length()
                    + " length not matched message store config value, ignore it");
                return true;
            }
            try {
                MappedFile mappedFile = new MappedFile(file.getPath(), mappedFileSize);
                //当前文件的写指针
                mappedFile.setWrotePosition(this.mappedFileSize);
                //刷写到磁盘指针，该指针之前的数据持久化到磁盘中
                mappedFile.setFlushedPosition(this.mappedFileSize);
                //当前文件的提交指针
                mappedFile.setCommittedPosition(this.mappedFileSize);
                //添加到MappedFile文件集合中
                this.mappedFiles.add(mappedFile);
                log.info("load " + file.getPath() + " OK");
            } catch (IOException e) {
                log.error("load file " + file + " error", e);
                return false;
            }
        }
    }
    return true;
}

很明显此方法就是MappedFile对象和一个Commitlog文件建立的逻辑关系

循环消息存储路径文件夹中的Commitlog文件，升序排列，创建MappedFile对象设置基础参数数据，添加到MappedFile文件集合中,我们查看new MappedFile(),调用MappedFile.init()方法

private void init(final String fileName, final int fileSize) throws IOException {
    this.fileName = fileName;
    this.fileSize = fileSize;
    this.file = new File(fileName);
    //初始化的初始偏移量是文件名称
    this.fileFromOffset = Long.parseLong(this.file.getName());
    boolean ok = false;
    ensureDirOK(this.file.getParent());
    try {
        //创建读写文件通道NIO
        this.fileChannel = new RandomAccessFile(this.file, "rw").getChannel();
        //将文件映射到内存
        this.mappedByteBuffer = this.fileChannel.map(MapMode.READ_WRITE, 0, fileSize);
        TOTAL_MAPPED_VIRTUAL_MEMORY.addAndGet(fileSize);
        TOTAL_MAPPED_FILES.incrementAndGet();
        ok = true;
    } catch (FileNotFoundException e) {
        log.error("create file channel " + this.fileName + " Failed. ", e);
        throw e;
    } catch (IOException e) {
        log.error("map file " + this.fileName + " Failed. ", e);
        throw e;
    } finally {
        if (!ok && this.fileChannel != null) {
            this.fileChannel.close();
        }
    }
}

将文件映射到内存。

上面我们分析了mappedFile和commitlog的逻辑建立关系，将mappedFile加入mappedFileQueue中，并讲解了MappedFile初始化的过程。

2.2、获取mappedFileQueue中最后一个mappedFile

上面我们了解到commitlog和mappedFile一一对应的关系，我们需要存储消息就需要找到最后一个未存满消息的commitlog文件，即查找的是最后一个mappedFiled对象

public MappedFile getLastMappedFile() {
    MappedFile mappedFileLast = null;

    while (!this.mappedFiles.isEmpty()) {
        try {
            mappedFileLast = this.mappedFiles.get(this.mappedFiles.size() - 1);
            break;
        } catch (IndexOutOfBoundsException e) {
            //continue;
        } catch (Exception e) {
            log.error("getLastMappedFile has exception.", e);
            break;
        }
    }

    return mappedFileLast;
}

该方法比较简单就是从mappedFiles集合中获取最后一个MappedFile对象，2.1中我们分析了其初始化的过程将MappedFile对象放入mappedFiles集合中。

3、创建映射文件MappedFile

当获取的MappedFile对象不存在或者消息已经存满我们需要创建,this.mappedFileQueue.getLastMappedFile(0)

public MappedFile getLastMappedFile(final long startOffset, boolean needCreate) {
    //创建映射问价的起始偏移量
    long createOffset = -1;
    //获取最后一个映射文件，如果为null或者写满则会执行创建逻辑
    MappedFile mappedFileLast = getLastMappedFile();
    //最后一个映射文件为null,创建一个新的映射文件
    if (mappedFileLast == null) {
        //计算将要创建的映射文件的起始偏移量
        //如果startOffset<=mappedFileSize则起始偏移量为0
        //如果startOffset>mappedFileSize则起始偏移量为是mappedFileSize的倍数
        createOffset = startOffset - (startOffset % this.mappedFileSize);
    }
    //映射文件满了，创建新的映射文件
    if (mappedFileLast != null && mappedFileLast.isFull()) {
        //创建的映射文件的偏移量等于最后一个映射文件的起始偏移量  + 映射文件的大小（commitlog文件大小）
        createOffset = mappedFileLast.getFileFromOffset() + this.mappedFileSize;
    }
    //创建新的映射文件
    if (createOffset != -1 && needCreate) {
        //构造commitlog名称
        String nextFilePath = this.storePath + File.separator + UtilAll.offset2FileName(createOffset);
        String nextNextFilePath = this.storePath + File.separator
            + UtilAll.offset2FileName(createOffset + this.mappedFileSize);
        MappedFile mappedFile = null;
        //优先通过allocateMappedFileService中方式构建映射文件，预分配方式，性能高
        //如果上述方式失败则通过new创建映射文件
        if (this.allocateMappedFileService != null) {
            mappedFile = this.allocateMappedFileService.putRequestAndReturnMappedFile(nextFilePath,
                nextNextFilePath, this.mappedFileSize);
        } else {
            try {
                mappedFile = new MappedFile(nextFilePath, this.mappedFileSize);
            } catch (IOException e) {
                log.error("create mappedFile exception", e);
            }
        }
        if (mappedFile != null) {
            if (this.mappedFiles.isEmpty()) {
                mappedFile.setFirstCreateInQueue(true);
            }
            this.mappedFiles.add(mappedFile);
        }
        return mappedFile;
    }
    return mappedFileLast;
}

allocateMappedFileService.putRequestAndReturnMappedFile()通过MappedFile服务类创建MappedFile

image

AllocateMappedFileService是创建MappedFile核心类，我们分析下该类

字段	类型	说明
waitTimeOut	int	等待创建映射文件的超时时间，默认5秒
requestTable	ConcurrentMap<String, AllocateRequest>	用来保存当前所有待处理的分配请求，其中KEY是filePath,VALUE是分配请求。如果分配请求被成功处理，即获取到映射文件则从请求会从requestTable中移除
requestQueue	PriorityBlockingQueue<AllocateRequest>	分配请求队列，注意是优先级队列，从该队列中获取请求，进而根据请求创建映射文件
hasException	boolean	标识是否发生异常
messageStore	DefaultMessageStore

分析其核心方法

public MappedFile putRequestAndReturnMappedFile(String nextFilePath, String nextNextFilePath, int fileSize) {
    //默认提交两个请求
    int canSubmitRequests = 2;
    //当transientStorePoolEnable为true，刷盘方式是ASYNC_FLUSH，broker不是SLAVE，才启动TransientStorePool
    if (this.messageStore.getMessageStoreConfig().isTransientStorePoolEnable()) {
        //启动快速失败策略时，计算TransientStorePool中剩余的buffer数量减去requestQueue中待分配的数量后，剩余的buffer数量
        if (this.messageStore.getMessageStoreConfig().isFastFailIfNoBufferInStorePool()
            && BrokerRole.SLAVE != this.messageStore.getMessageStoreConfig().getBrokerRole()) { //if broker is slave, don't fast fail even no buffer in pool
            canSubmitRequests = this.messageStore.getTransientStorePool().remainBufferNumbs() - this.requestQueue.size();
        }
    }

    AllocateRequest nextReq = new AllocateRequest(nextFilePath, fileSize);
    //判断requestTable中是否存在该路径的分配请求，如果存在则说明该请求已经在排队中
    boolean nextPutOK = this.requestTable.putIfAbsent(nextFilePath, nextReq) == null;

    //该路径没有在排队
    if (nextPutOK) {
        //如果剩余的buffer数量小于等于0则快速失败
        if (canSubmitRequests <= 0) {
            log.warn("[NOTIFYME]TransientStorePool is not enough, so create mapped file error, " +
                "RequestQueueSize : {}, StorePoolSize: {}", this.requestQueue.size(), this.messageStore.getTransientStorePool().remainBufferNumbs());
            this.requestTable.remove(nextFilePath);
            return null;
        }
        //将指定的元素插入到此优先级队列中
        boolean offerOK = this.requestQueue.offer(nextReq);
        if (!offerOK) {
            log.warn("never expected here, add a request to preallocate queue failed");
        }
        //剩余的buffer数量减1
        canSubmitRequests--;
    }

    //创建第二个映射文件
    AllocateRequest nextNextReq = new AllocateRequest(nextNextFilePath, fileSize);
    boolean nextNextPutOK = this.requestTable.putIfAbsent(nextNextFilePath, nextNextReq) == null;
    if (nextNextPutOK) {
        //检查buffer数量
        if (canSubmitRequests <= 0) {
            log.warn("[NOTIFYME]TransientStorePool is not enough, so skip preallocate mapped file, " +
                "RequestQueueSize : {}, StorePoolSize: {}", this.requestQueue.size(), this.messageStore.getTransientStorePool().remainBufferNumbs());
            this.requestTable.remove(nextNextFilePath);
        } else {
            //将指定的元素插入到此优先级队列中
            boolean offerOK = this.requestQueue.offer(nextNextReq);
            if (!offerOK) {
                log.warn("never expected here, add a request to preallocate queue failed");
            }
        }
    }
    if (hasException) {
        log.warn(this.getServiceName() + " service has exception. so return null");
        return null;
    }
    AllocateRequest result = this.requestTable.get(nextFilePath);
    try {
        if (result != null) {
            //等待
            boolean waitOK = result.getCountDownLatch().await(waitTimeOut, TimeUnit.MILLISECONDS);
            if (!waitOK) {
                log.warn("create mmap timeout " + result.getFilePath() + " " + result.getFileSize());
                return null;
            } else {
                this.requestTable.remove(nextFilePath);
                return result.getMappedFile();
            }
        } else {
            log.error("find preallocate mmap failed, this never happen");
        }
    } catch (InterruptedException e) {
        log.warn(this.getServiceName() + " service has exception. ", e);
    }
    return null;
}

将创建请求插入到requestQueue和requestTable中，由于优先级队列中requestQueue存入的是AllocateRequest对象实现了compareTo方法，优先级的排序，由于创建MappedFile时传入的是预创建两个，我们需要创建最新的请求的结果，其他请求需要进行排队。

AllocateMappedFileService是个多线程类，内部实现了run()的核心方法mmapOperation()

private boolean mmapOperation() {
    boolean isSuccess = false;
    AllocateRequest req = null;
    try {
        //检索并删除此队列的头，如有必要，等待元素可用
        req = this.requestQueue.take();
        //
        AllocateRequest expectedRequest = this.requestTable.get(req.getFilePath());
        if (null == expectedRequest) {
            log.warn("this mmap request expired, maybe cause timeout " + req.getFilePath() + " "
                + req.getFileSize());
            return true;
        }
        if (expectedRequest != req) {
            log.warn("never expected here,  maybe cause timeout " + req.getFilePath() + " "
                + req.getFileSize() + ", req:" + req + ", expectedRequest:" + expectedRequest);
            return true;
        }

        if (req.getMappedFile() == null) {
            long beginTime = System.currentTimeMillis();

            MappedFile mappedFile;
            //判断TransientStorePoolEnable是否启用
            if (messageStore.getMessageStoreConfig().isTransientStorePoolEnable()) {
                try {
                    mappedFile = ServiceLoader.load(MappedFile.class).iterator().next();
                    mappedFile.init(req.getFilePath(), req.getFileSize(), messageStore.getTransientStorePool());
                } catch (RuntimeException e) {//默认方式创建
                    log.warn("Use default implementation.");
                    mappedFile = new MappedFile(req.getFilePath(), req.getFileSize(), messageStore.getTransientStorePool());
                }
            } else {
                mappedFile = new MappedFile(req.getFilePath(), req.getFileSize());
            }

            long eclipseTime = UtilAll.computeEclipseTimeMilliseconds(beginTime);
            if (eclipseTime > 10) {
                int queueSize = this.requestQueue.size();
                log.warn("create mappedFile spent time(ms) " + eclipseTime + " queue size " + queueSize
                    + " " + req.getFilePath() + " " + req.getFileSize());
            }

            // pre write mappedFile
            if (mappedFile.getFileSize() >= this.messageStore.getMessageStoreConfig().getMapedFileSizeCommitLog()
                && this.messageStore.getMessageStoreConfig().isWarmMapedFileEnable()) {
                //对MappedFile进行预热
                mappedFile.warmMappedFile(this.messageStore.getMessageStoreConfig().getFlushDiskType(),
                    this.messageStore.getMessageStoreConfig().getFlushLeastPagesWhenWarmMapedFile());
            }

            req.setMappedFile(mappedFile);
            this.hasException = false;
            isSuccess = true;
        }
    } catch (InterruptedException e) {
        log.warn(this.getServiceName() + " interrupted, possibly by shutdown.");
        this.hasException = true;
        return false;
    } catch (IOException e) {
        log.warn(this.getServiceName() + " service has exception. ", e);
        this.hasException = true;
        if (null != req) {
            requestQueue.offer(req);
            try {
                Thread.sleep(1);
            } catch (InterruptedException ignored) {
            }
        }
    } finally {
        if (req != null && isSuccess)
            req.getCountDownLatch().countDown();
    }
    return true;
}

image

关于MMAP详细分析请查看-》linux内存映射mmap原理分析一文

我们发现有两种方式创建mappedFile对象

1、mappedFile = new MappedFile(req.getFilePath(), req.getFileSize())

public MappedFile(final String fileName, final int fileSize) throws IOException {
    init(fileName, fileSize);
}
private void init(final String fileName, final int fileSize) throws IOException {
    this.fileName = fileName;
    this.fileSize = fileSize;
    this.file = new File(fileName);
    //初始化的初始偏移量是文件名称
    this.fileFromOffset = Long.parseLong(this.file.getName());
    boolean ok = false;
    ensureDirOK(this.file.getParent());
    try {
        //创建读写文件通道NIO
        this.fileChannel = new RandomAccessFile(this.file, "rw").getChannel();
        //将文件映射到内存
        this.mappedByteBuffer = this.fileChannel.map(MapMode.READ_WRITE, 0, fileSize);
        TOTAL_MAPPED_VIRTUAL_MEMORY.addAndGet(fileSize);
        TOTAL_MAPPED_FILES.incrementAndGet();
        ok = true;
    } catch (FileNotFoundException e) {
        log.error("create file channel " + this.fileName + " Failed. ", e);
        throw e;
    } catch (IOException e) {
        log.error("map file " + this.fileName + " Failed. ", e);
        throw e;
    } finally {
        if (!ok && this.fileChannel != null) {
            this.fileChannel.close();
        }
    }
}

2、mappedFile = ServiceLoader.load(MappedFile.class).iterator().next(); mappedFile.init(req.getFilePath(), req.getFileSize(), messageStore.getTransientStorePool())

//transientStorePoolEnable 为 true
public void init(final String fileName, final int fileSize,
    final TransientStorePool transientStorePool) throws IOException {
    init(fileName, fileSize);
    //初始化MappedFile的writeBuffer
    this.writeBuffer = transientStorePool.borrowBuffer();
    this.transientStorePool = transientStorePool;
}

同样调用了init(fileName, fileSize)方法。

image

TransientStorePool与MappedFile在数据处理上的差异在什么地方呢？分析其代码，TransientStorePool会通过ByteBuffer.allocateDirect调用直接申请对外内存，消息数据在写入内存的时候是写入预申请的内存中。在异步刷盘的时候，再由刷盘线程将这些内存中的修改写入文件。

那么与直接使用MappedByteBuffer相比差别在什么地方呢？修改MappedByteBuffer实际会将数据写入文件对应的Page Cache中，而TransientStorePool方案下写入的则为纯粹的内存。因此在消息写入操作上会更快，因此能更少的占用CommitLog.putMessageLock锁，从而能够提升消息处理量。使用TransientStorePool方案的缺陷主要在于在异常崩溃的情况下回丢失更多的消息。

创建完mappedFile对象后，有个预热操作，每个字节填充(byte) 0

public void warmMappedFile(FlushDiskType type, int pages) {
    long beginTime = System.currentTimeMillis();
    //创建一个新的字节缓冲区，其内容是此缓冲区内容的共享子序列
    ByteBuffer byteBuffer = this.mappedByteBuffer.slice();
    //记录上一次刷盘的字节数
    int flush = 0;
    long time = System.currentTimeMillis();
    for (int i = 0, j = 0; i < this.fileSize; i += MappedFile.OS_PAGE_SIZE, j++) {
        byteBuffer.put(i, (byte) 0);
        // 刷盘方式是同步策略时，进行刷盘操作
        // 每修改pages个分页刷一次盘，相当于4096*4k = 16M  每16M刷一次盘，1G文件 1024M/16M = 64次
        if (type == FlushDiskType.SYNC_FLUSH) {
            if ((i / OS_PAGE_SIZE) - (flush / OS_PAGE_SIZE) >= pages) {
                flush = i;
                mappedByteBuffer.force();
            }
        }

        // 防止垃圾回收GC
        if (j % 1000 == 0) {
            log.info("j={}, costTime={}", j, System.currentTimeMillis() - time);
            time = System.currentTimeMillis();
            try {
                Thread.sleep(0);
            } catch (InterruptedException e) {
                log.error("Interrupted", e);
            }
        }
    }
    // force flush when prepare load finished
    if (type == FlushDiskType.SYNC_FLUSH) {
        log.info("mapped file warm-up done, force to disk, mappedFile={}, costTime={}",
            this.getFileName(), System.currentTimeMillis() - beginTime);
        //刷盘，强制将此缓冲区内容的任何更改写入包含映射文件的存储设备
        mappedByteBuffer.force();
    }
    log.info("mapped file warm-up done. mappedFile={}, costTime={}", this.getFileName(),
        System.currentTimeMillis() - beginTime);

    this.mlock();
}

4、映射文件中写入消息

MappedFile.appendMessage()的核心方法MappedFile.appendMessagesInner()

public AppendMessageResult appendMessagesInner(final MessageExt messageExt, final AppendMessageCallback cb) {
    assert messageExt != null;
    assert cb != null;
    //获取当前写的指针
    int currentPos = this.wrotePosition.get();

    if (currentPos < this.fileSize) {
        //创建一个与MappedFile的共享内存区
        ByteBuffer byteBuffer = writeBuffer != null ? writeBuffer.slice() : this.mappedByteBuffer.slice();
        //设置指针
        byteBuffer.position(currentPos);
        AppendMessageResult result = null;
        if (messageExt instanceof MessageExtBrokerInner) {
            result = cb.doAppend(this.getFileFromOffset(), byteBuffer, this.fileSize - currentPos, (MessageExtBrokerInner) messageExt);
        } else if (messageExt instanceof MessageExtBatch) {
            result = cb.doAppend(this.getFileFromOffset(), byteBuffer, this.fileSize - currentPos, (MessageExtBatch) messageExt);
        } else {
            return new AppendMessageResult(AppendMessageStatus.UNKNOWN_ERROR);
        }
        this.wrotePosition.addAndGet(result.getWroteBytes());
        this.storeTimestamp = result.getStoreTimestamp();
        return result;
    }
    //当前写的指针大于文件的大小则抛出异常
    log.error("MappedFile.appendMessage return null, wrotePosition: {} fileSize: {}", currentPos, this.fileSize);
    return new AppendMessageResult(AppendMessageStatus.UNKNOWN_ERROR);
}

追加消息的核心方法Commotlog.doAppend()

public AppendMessageResult doAppend(final long fileFromOffset, final ByteBuffer byteBuffer, final int maxBlank,
    final MessageExtBrokerInner msgInner) {
    // STORETIMESTAMP + STOREHOSTADDRESS + OFFSET <br>

    // PHY OFFSET
    //写入的位置
    long wroteOffset = fileFromOffset + byteBuffer.position();

    this.resetByteBuffer(hostHolder, 8);
    //创建全局唯一的消息ID，消息ID有16字节，4个字节IP+4个字节的端口号+8字节的消息偏移量
    String msgId = MessageDecoder.createMessageId(this.msgIdMemory, msgInner.getStoreHostBytes(hostHolder), wroteOffset);

    // Record ConsumeQueue information
    keyBuilder.setLength(0);
    keyBuilder.append(msgInner.getTopic());
    keyBuilder.append('-');
    keyBuilder.append(msgInner.getQueueId());
    String key = keyBuilder.toString();
    //从CommitLog中保存了主题和队列的组合      待写入的偏移量
    Long queueOffset = CommitLog.this.topicQueueTable.get(key);
    //可能是第一次还没有偏移量设置为0
    if (null == queueOffset) {
        queueOffset = 0L;
        CommitLog.this.topicQueueTable.put(key, queueOffset);
    }

    // Transaction messages that require special handling
    final int tranType = MessageSysFlag.getTransactionValue(msgInner.getSysFlag());
    switch (tranType) {
        // Prepared and Rollback message is not consumed, will not enter the
        // consumer queuec
        case MessageSysFlag.TRANSACTION_PREPARED_TYPE:
        case MessageSysFlag.TRANSACTION_ROLLBACK_TYPE:
            queueOffset = 0L;
            break;
        case MessageSysFlag.TRANSACTION_NOT_TYPE:
        case MessageSysFlag.TRANSACTION_COMMIT_TYPE:
        default:
            break;
    }

    /**
     * Serialize message
     */
    final byte[] propertiesData =
        msgInner.getPropertiesString() == null ? null : msgInner.getPropertiesString().getBytes(MessageDecoder.CHARSET_UTF8);

    final int propertiesLength = propertiesData == null ? 0 : propertiesData.length;

    if (propertiesLength > Short.MAX_VALUE) {
        log.warn("putMessage message properties length too long. length={}", propertiesData.length);
        return new AppendMessageResult(AppendMessageStatus.PROPERTIES_SIZE_EXCEEDED);
    }

    final byte[] topicData = msgInner.getTopic().getBytes(MessageDecoder.CHARSET_UTF8);
    final int topicLength = topicData.length;

    final int bodyLength = msgInner.getBody() == null ? 0 : msgInner.getBody().length;
    //计算：消息长度 = 消息体的长度+消息主题的长度 +消息属性的长度
    final int msgLen = calMsgLength(bodyLength, topicLength, propertiesLength);

    // Exceeds the maximum message

    if (msgLen > this.maxMessageSize) {
        CommitLog.log.warn("message size exceeded, msg total size: " + msgLen + ", msg body size: " + bodyLength
            + ", maxMessageSize: " + this.maxMessageSize);
        return new AppendMessageResult(AppendMessageStatus.MESSAGE_SIZE_EXCEEDED);
    }
    //如果消息的长度+END_FILE_MIN_BLANK_LENGTH大于剩余的空闲长度
    // Determines whether there is sufficient free space
    //每一个CommitLog文件至少会空闲8个字节，前4位记录当前文件剩余空间，后四位存储魔数（CommitLog.MESSAGE_MAGIC_CODE）
    if ((msgLen + END_FILE_MIN_BLANK_LENGTH) > maxBlank) {
        this.resetByteBuffer(this.msgStoreItemMemory, maxBlank);
        // 1 TOTALSIZE
        this.msgStoreItemMemory.putInt(maxBlank);
        // 2 MAGICCODE
        this.msgStoreItemMemory.putInt(CommitLog.BLANK_MAGIC_CODE);
        // 3 The remaining space may be any value
        // Here the length of the specially set maxBlank
        final long beginTimeMills = CommitLog.this.defaultMessageStore.now();
        byteBuffer.put(this.msgStoreItemMemory.array(), 0, maxBlank);
        return new AppendMessageResult(AppendMessageStatus.END_OF_FILE, wroteOffset, maxBlank, msgId, msgInner.getStoreTimestamp(),
            queueOffset, CommitLog.this.defaultMessageStore.now() - beginTimeMills);
    }

    // Initialization of storage space
    this.resetByteBuffer(msgStoreItemMemory, msgLen);
    // 1 TOTALSIZE 该消息条目的总长度，4字节
    this.msgStoreItemMemory.putInt(msgLen);
    // 2 MAGICCODE 魔数 ，4字节
    this.msgStoreItemMemory.putInt(CommitLog.MESSAGE_MAGIC_CODE);
    // 3 BODYCRC 消息体crc校验码 4字节
    this.msgStoreItemMemory.putInt(msgInner.getBodyCRC());
    // 4 QUEUEID 消息消费队列的ID 4字节
    this.msgStoreItemMemory.putInt(msgInner.getQueueId());
    // 5 FLAG
    this.msgStoreItemMemory.putInt(msgInner.getFlag());
    // 6 QUEUEOFFSET 消息在消息消费队列的偏移量，8字节
    this.msgStoreItemMemory.putLong(queueOffset);
    // 7 PHYSICALOFFSET 消息在CommitLog文件中的偏移量 8字节
    this.msgStoreItemMemory.putLong(fileFromOffset + byteBuffer.position());
    // 8 SYSFLAG
    this.msgStoreItemMemory.putInt(msgInner.getSysFlag());
    // 9 BORNTIMESTAMP 消息生产者调用消息发送的API的时间戳 8字节
    this.msgStoreItemMemory.putLong(msgInner.getBornTimestamp());
    // 10 BORNHOST  消息发送者的ip、端口号 8字节
    this.resetByteBuffer(hostHolder, 8);
    this.msgStoreItemMemory.put(msgInner.getBornHostBytes(hostHolder));
    // 11 STORETIMESTAMP  消息存储时间戳，8字节
    this.msgStoreItemMemory.putLong(msgInner.getStoreTimestamp());
    // 12 STOREHOSTADDRESS  broker服务器的IP+端口号 8字节
    this.resetByteBuffer(hostHolder, 8);
    this.msgStoreItemMemory.put(msgInner.getStoreHostBytes(hostHolder));
    //this.msgBatchMemory.put(msgInner.getStoreHostBytes());
    // 13 RECONSUMETIMES 消息重试的次数，4字节
    this.msgStoreItemMemory.putInt(msgInner.getReconsumeTimes());
    // 14 Prepared Transaction Offset 事务消息物理偏移量，8字节
    this.msgStoreItemMemory.putLong(msgInner.getPreparedTransactionOffset());
    // 15 BODY 消息体内容，bodyLength的长度
    this.msgStoreItemMemory.putInt(bodyLength);
    if (bodyLength > 0)
        this.msgStoreItemMemory.put(msgInner.getBody());
    // 16 TOPIC 主题
    this.msgStoreItemMemory.put((byte) topicLength);
    this.msgStoreItemMemory.put(topicData);
    // 17 PROPERTIES 消息属性
    this.msgStoreItemMemory.putShort((short) propertiesLength);
    if (propertiesLength > 0)
        this.msgStoreItemMemory.put(propertiesData);

    final long beginTimeMills = CommitLog.this.defaultMessageStore.now();
    // Write messages to the queue buffer
    //写到消息队列缓存中
    byteBuffer.put(this.msgStoreItemMemory.array(), 0, msgLen);

    AppendMessageResult result = new AppendMessageResult(AppendMessageStatus.PUT_OK, wroteOffset, msgLen, msgId,
        msgInner.getStoreTimestamp(), queueOffset, CommitLog.this.defaultMessageStore.now() - beginTimeMills);

    switch (tranType) {
        case MessageSysFlag.TRANSACTION_PREPARED_TYPE:
        case MessageSysFlag.TRANSACTION_ROLLBACK_TYPE:
            break;
        case MessageSysFlag.TRANSACTION_NOT_TYPE:
        case MessageSysFlag.TRANSACTION_COMMIT_TYPE:
            // The next update ConsumeQueue information
            CommitLog.this.topicQueueTable.put(key, ++queueOffset);
            break;
        default:
            break;
    }
    return result;
}

构建消息的基础参数，返回放入缓存的状态及写指针的位置。