RocketMQ源码解析之Broker消息存储(刷盘机制)

  在上篇RocketMQ源码解析之Broker消息存储(消息存储)中分析了消息如何从Broker存储到MappedFile内存缓冲区,但还没有存储到文件中,即还没有刷盘,本篇将介绍RocketMQ如何进行刷盘的
  RocketMQ 的读写时基于NIO的内存映射机制的,进行消息存储时先将消息追加到MappedFile,然后根据不同的刷盘策略进行刷盘。同步刷盘是先将消息追加到内存后,将同步调用MappedByteBuffer的force()方法,同步等待刷盘结果,进行刷盘结果返回;异步刷盘在将消息追加到内存后,不等待刷盘结果立即将刷盘成功的消息返回给消息发送端;

CommitLog#submitFlushRequest 分别针对同步和异步情况进行刷盘
/**
 * 刷盘
 *
 * @param result
 * @param messageExt
 * @return
 */
public CompletableFuture<PutMessageStatus> submitFlushRequest(AppendMessageResult result, MessageExt messageExt) {
    /**
     * flushCommitLogService 在CommitLog构造函数中实例化,按照同步和异步分别实例化对应的子类
     */
    // Synchronization flush 同步刷盘
    if (FlushDiskType.SYNC_FLUSH == this.defaultMessageStore.getMessageStoreConfig().getFlushDiskType()) {
        // GroupCommitService 同步刷盘服务类
        /**将请求被提交给GroupCommitService后,GroupCommitService并不是立即处理,而是
         * 先放到内部的一个请求队列中
         * */
        final GroupCommitService service = (GroupCommitService) this.flushCommitLogService; // 刷盘操作,
        if (messageExt.isWaitStoreMsgOK()) {    // 进行同步刷盘
            // 数据准备
            GroupCommitRequest request = new GroupCommitRequest(result.getWroteOffset() + result.getWroteBytes(),
                    this.defaultMessageStore.getMessageStoreConfig().getSyncFlushTimeout());
            /**
             * 将数据对象放到requestWrite中
             */
            service.putRequest(request);
            // 利用future模式的原理,阻塞等待
            return request.future();
        } else {
            service.wakeup();
            return CompletableFuture.completedFuture(PutMessageStatus.PUT_OK);
        }
    }
    // Asynchronous flush 异步刷盘
    else {
        if (!this.defaultMessageStore.getMessageStoreConfig().isTransientStorePoolEnable()) {
            flushCommitLogService.wakeup();
        } else {
            commitLogService.wakeup();
        }
        return CompletableFuture.completedFuture(PutMessageStatus.PUT_OK);
    }
}
  • 第13行—第29行是针对同步刷盘,第32行—第36行是针对异步刷盘,同步刷盘的服务类是GroupCommitService ;异步刷盘服务类则是FlushRealTimeService

CommitLog在初始化的时候,会根据配置,启动两种不同的刷盘服务:

public CommitLog(final DefaultMessageStore defaultMessageStore) {
	......
    // 如果是同步刷盘
    if (FlushDiskType.SYNC_FLUSH == defaultMessageStore.getMessageStoreConfig().getFlushDiskType()) {
        this.flushCommitLogService = new GroupCommitService();
    } else {
        // 异步刷盘
        this.flushCommitLogService = new FlushRealTimeService();
    }
    ......
}
GroupCommitService#putRequest 将刷盘任务放入到requestsWrite中,

requestsWrite:写队列,主要用于向该线程中添加刷盘任务

requestsRead:读队列,主要用于执行特定的刷盘任务,读写分离

/**
 * GroupCommitRequest刷盘任务放入到 requestWrite 中,就返回了
 *
 * @param request
 */
public synchronized void putRequest(final GroupCommitRequest request) {
    synchronized (this.requestsWrite) { // 写队列
        this.requestsWrite.add(request);
    }
    /**
     * 设置成true,然后计数器-1 ,对应的run方法才会进行数据交换
     */
    this.wakeup();
}

public void wakeup() {
    /**
     * ** 设置成true,然后计数器-1 ,对应的run方法才会进行数据交换
     */
    if (hasNotified.compareAndSet(false, true)) {
        // waitPoint 是利用CountDownLatch自定义的CountDownLatch2对象
        waitPoint.countDown(); // notify
    }
}

GroupCommitService 被唤醒后,便会将requestsWrite 中的请求交换到requestsRead中,避免产生锁竞争,doCommit函数会遍历requestsRead的请求进行处理

public void run() {
    CommitLog.log.info(this.getServiceName() + " service started");

    while (!this.isStopped()) {
        try {
            /**
             *  一个线程一直的处理同步刷写任务,
             */
            this.waitForRunning(10);
            /**
             * 真正的刷盘逻辑
             */
            this.doCommit();
        } catch (Exception e) {
            CommitLog.log.warn(this.getServiceName() + " service has exception. ", e);
        }
    }

    // Under normal circumstances shutdown, wait for the arrival of the
    // request, and then flush
    // 每处理一个循环后等待 10 毫秒,
    // 一旦新任务到达,立即唤醒执行任务
    try {
        Thread.sleep(10);
    } catch (InterruptedException e) {
        CommitLog.log.warn("GroupCommitService Exception, ", e);
    }

    synchronized (this) {
        this.swapRequests();
    }

    this.doCommit();

    CommitLog.log.info(this.getServiceName() + " service end");
}

private void doCommit() {
    synchronized (this.requestsRead) {
        if (!this.requestsRead.isEmpty()) {
            for (GroupCommitRequest req : this.requestsRead) {
                // There may be a message in the next file, so a maximum of
                // two times the flush
                boolean flushOK = CommitLog.this.mappedFileQueue.getFlushedWhere() >= req.getNextOffset();
                for (int i = 0; i < 2 && !flushOK; i++) {
                    /**
                     * 执行刷盘操作
                     */
                    CommitLog.this.mappedFileQueue.flush(0);
                    flushOK = CommitLog.this.mappedFileQueue.getFlushedWhere() >= req.getNextOffset();
                }
                // 唤醒用户线程
                req.wakeupCustomer(flushOK ? PutMessageStatus.PUT_OK : PutMessageStatus.FLUSH_DISK_TIMEOUT);
            }

            long storeTimestamp = CommitLog.this.mappedFileQueue.getStoreTimestamp();
            if (storeTimestamp > 0) {
                // 保存刷盘检测点的commitLog 文件刷盘时间
                             CommitLog.this.defaultMessageStore.getStoreCheckpoint().setPhysicMsgTimestamp(storeTimestamp);
            }

            this.requestsRead.clear();
        } else {
            // Because of individual messages is set to not sync flush, it
            // will come to this process
            CommitLog.this.mappedFileQueue.flush(0);
        }
    }
}
  • 第49行是具体的刷盘操作,详情解析见MappedFileQueue#flush
FlushRealTimeService 异步刷盘
public void run() {
    CommitLog.log.info(this.getServiceName() + " service started");

    /**
     * 周期性地将内存缓冲区中的内容刷到文件中
     */
    while (!this.isStopped()) {
        // 主要是休眠策略,下面会使用这个,true则使用Thread.sleep,false则使用waitForRunning,默认false
        boolean flushCommitLogTimed = CommitLog.this.defaultMessageStore.getMessageStoreConfig().isFlushCommitLogTimed();

        // interval:获取刷盘的间隔时间,默认500ms
        int interval = CommitLog.this.defaultMessageStore.getMessageStoreConfig().getFlushIntervalCommitLog();
        // flushPhysicQueueLeastPages:每次刷盘最少需要刷新的页,每页大小为4k,默认每次要刷新4页
        int flushPhysicQueueLeastPages = CommitLog.this.defaultMessageStore.getMessageStoreConfig().getFlushCommitLogLeastPages();
        // 两次刷写之间的最大时间间隔,默认10s
        int flushPhysicQueueThoroughInterval =
                CommitLog.this.defaultMessageStore.getMessageStoreConfig().getFlushCommitLogThoroughInterval();

        boolean printFlushProgress = false;

        // Print flush progress
        long currentTimeMillis = System.currentTimeMillis();
        /**
         * flushPhysicQueueThoroughInterval:两次刷写之间的最大时间间隔
         * 如果当前时间距离上次刷盘时间超出设置的两次刷盘最大间隔
         * 则改变 flushPhysicQueueLeastPages =0,并每10次输出异常刷新进度
         * flushPhysicQueueLeastPages:每次刷盘最少需要刷新的页
         */
        if (currentTimeMillis >= (this.lastFlushTimestamp + flushPhysicQueueThoroughInterval)) {
            this.lastFlushTimestamp = currentTimeMillis;
            flushPhysicQueueLeastPages = 0;
            printFlushProgress = (printTimes++ % 10) == 0;
        }

        try {
            /**
             * 根据不同的休眠策略,进行休眠等待
             * 两者区别是啥-2021/10/19
             */
            if (flushCommitLogTimed) {
                Thread.sleep(interval);
            } else {
                this.waitForRunning(interval);
            }

            if (printFlushProgress) {
                this.printFlushProgress();
            }

            long begin = System.currentTimeMillis();
            // 调用刷盘操作
            CommitLog.this.mappedFileQueue.flush(flushPhysicQueueLeastPages);
            long storeTimestamp = CommitLog.this.mappedFileQueue.getStoreTimestamp();
            if (storeTimestamp > 0) {
                /**
                 * 设置检测点的StoreCheckpoint 的physicMsgTimestamp(commitlog文件的检测点,也就是记录最新刷盘的时间戳)
                 */
                CommitLog.this.defaultMessageStore.getStoreCheckpoint().setPhysicMsgTimestamp(storeTimestamp);
            }
            long past = System.currentTimeMillis() - begin;
            if (past > 500) {
                log.info("Flush data to disk costs {} ms", past);
            }
        } catch (Throwable e) {
            CommitLog.log.warn(this.getServiceName() + " service has exception. ", e);
            this.printFlushProgress();
        }
    }

    // Normal shutdown, to ensure that all the flush before exit
    boolean result = false;
    for (int i = 0; i < RETRY_TIMES_OVER && !result; i++) {
        result = CommitLog.this.mappedFileQueue.flush(0);
        CommitLog.log.info(this.getServiceName() + " service shutdown, retry " + (i + 1) + " times " + (result ? "OK" : "Not OK"));
    }

    this.printFlushProgress();

    CommitLog.log.info(this.getServiceName() + " service end");
}
MappedFileQueue#flush 刷盘具体实现
/**
 * 刷盘具体实现
 *
 * @param flushLeastPages
 * @return
 */
public boolean flush(final int flushLeastPages) {
    boolean result = true;
    /**
     * 根据上次刷新的位置,得到当前的MappedFile对象
     */
    MappedFile mappedFile = this.findMappedFileByOffset(this.flushedWhere, this.flushedWhere == 0);
    if (mappedFile != null) {
        long tmpTimeStamp = mappedFile.getStoreTimestamp();
        int offset = mappedFile.flush(flushLeastPages);     // 执行刷新,返回当前刷新后的位置
        long where = mappedFile.getFileFromOffset() + offset;   // 更新刷新的位置
        result = where == this.flushedWhere;
        this.flushedWhere = where;
        if (0 == flushLeastPages) {
            this.storeTimestamp = tmpTimeStamp;
        }
    }

    return result;
}
  • 第12行是根据上次刷盘的位置,重新得到当前的MappedFile对象,详情解析见MappedFileQueue#findMappedFileByOffset
  • 第15行—第18行,执行刷盘,并更新刷盘位置,详情解析见MappedFile#flush
MappedFileQueue#findMappedFileByOffset
/**
 * Finds a mapped file by offset.
 * 根据 committedWhere 找到需要刷盘的 MappedFile 文件
 * -2021/10/19
 *
 * @param offset                Offset.
 * @param returnFirstOnNotFound If the mapped file is not found, then return the first one.
 * @return Mapped file or null (when not found and returnFirstOnNotFound is <code>false</code>).
 */
public MappedFile findMappedFileByOffset(final long offset, final boolean returnFirstOnNotFound) {
    try {
        /**
         * 此处得到的第一个MappedFile的文件起始offset,
         * 不一定是0,之前的文件可能已经被清除了
         */
        MappedFile firstMappedFile = this.getFirstMappedFile();
        /**
         * 获取队列中最后一个映射文件
         */
        MappedFile lastMappedFile = this.getLastMappedFile();
        if (firstMappedFile != null && lastMappedFile != null) {
            if (offset < firstMappedFile.getFileFromOffset() || offset >= lastMappedFile.getFileFromOffset() + this.mappedFileSize) {
                LOG_ERROR.warn("Offset not matched. Request offset: {}, firstOffset: {}, lastOffset: {}, mappedFileSize: {}, mappedFiles count: {}",
                        offset,
                        firstMappedFile.getFileFromOffset(),
                        lastMappedFile.getFileFromOffset() + this.mappedFileSize,
                        this.mappedFileSize,
                        this.mappedFiles.size());
            } else {
                /**
                 * 找到映射文件在队列中的索引位置
                 */
                int index = (int) ((offset / this.mappedFileSize) - (firstMappedFile.getFileFromOffset() / this.mappedFileSize));
                MappedFile targetFile = null;
                try {
                    /**
                     * 获取索引文件大小
                     */
                    targetFile = this.mappedFiles.get(index);
                } catch (Exception ignored) {
                }
                // offset在目标文件的起始offset和结束offset范围内
                if (targetFile != null && offset >= targetFile.getFileFromOffset()
                        && offset < targetFile.getFileFromOffset() + this.mappedFileSize) {
                    return targetFile;
                }
                // 如果按索引在队列中找不到映射文件就遍历队列查找映射文件
                for (MappedFile tmpMappedFile : this.mappedFiles) {
                    if (offset >= tmpMappedFile.getFileFromOffset()
                            && offset < tmpMappedFile.getFileFromOffset() + this.mappedFileSize) {
                        return tmpMappedFile;
                    }
                }
            }

            if (returnFirstOnNotFound) {
                return firstMappedFile;
            }
        }
    } catch (Exception e) {
        log.error("findMappedFileByOffset Exception", e);
    }

    return null;
}
MappedFile#flush
/**
 * @return The current flushed position 返回当前刷新的位置
 * 就是调用FileChannel或者mappedByteBuffer的force方法
 */
public int flush(final int flushLeastPages) {
    // 判断是否满足刷盘条件:即当前剩余未刷盘内容长度是否超过最小刷盘长度
    if (this.isAbleToFlush(flushLeastPages)) {
        if (this.hold()) {
            int value = getReadPosition();

            try {
                //We only append data to fileChannel or mappedByteBuffer, never both.
                // 如果writeBuffer不为空,说明开启了transientStorePoolEnable,
                // 消息会先写入堆外内存,然后提交到PageCache并最终刷写到磁盘,
                // 则表明消息是先提交到writeBuffer中的
                // (如果writeBuffer不为空,则在调用MappedFile(,,transientStorePool)#init时就已经初始化writeBuffer了)
                // 已经将消息从writeBuffer提交到fileChannel,直接调用fileChannel.force()
                if (writeBuffer != null || this.fileChannel.position() != 0) {
                    // 刷盘 NIO
                    this.fileChannel.force(false);
                } else {
                    // 反之,消息是直接存储在文件映射缓存区MappedByteBuffer中,直接调用它的force()即可
                    // 刷盘 NIO
                    this.mappedByteBuffer.force();
                }
            } catch (Throwable e) {
                log.error("Error occurred when force data to disk.", e);
            }

            this.flushedPosition.set(value);
            this.release();
        } else {
            log.warn("in flush, hold failed, flush offset = " + this.flushedPosition.get());
            this.flushedPosition.set(getReadPosition());
        }
    }
    // 当前刷新到磁盘的指针
    return this.getFlushedPosition();
}
  • 第7行:判断当前剩余未刷盘内容长度是否超过最小刷盘长度,是否满足刷盘条件
  • 第18行至第24行,如果开启了transientStorePoolEnable模式,消息会先写入堆外内存,然后提交到PageCache并最终刷写到磁盘.如果消息是直接存储在文件映射缓存区MappedByteBuffer中,直接调用它的force()即可
  • 第38行,返回当前刷新到磁盘的指针
FlushRealTimeService(异步不使用暂存池刷盘) 和 CommitRealTimeService(异步使用暂存池刷盘)的区别?
TransientStorePool的作用

TransientStorePool 相当于在内存层面做了读写分离,写走内存磁盘,读走pagecache,同时最大程度消除了page cache的锁竞争,降低了毛刺。它还使用了锁机制,避免直接内存被交换到swap分区
参考:https://github.com/apache/rocketmq/issues/2466

FileChannel.force VS MappedByteBuffer.force区别
posted @ 2021-10-19 20:25  枫叶艾辰  阅读(137)  评论(0编辑  收藏  举报