kafka 和 rocketMQ 的数据存储

kafka 版本：1.1.1

一个分区对应一个文件夹，数据以 segment 文件存储，segment 默认 1G。

分区文件夹：

segment 文件：

segment 的命名规则是怎样的？

kafka roll segment 的逻辑：kafka.log.Log#roll

  /**
   * Roll the log over to a new active segment starting with the current logEndOffset.
   * This will trim the index to the exact size of the number of entries it currently contains.
   *
   * @return The newly rolled segment
   */
  def roll(expectedNextOffset: Option[Long] = None): LogSegment = {
    maybeHandleIOException(s"Error while rolling log segment for $topicPartition in dir ${dir.getParent}") {
      val start = time.hiResClockMs()
      lock synchronized {
        checkIfMemoryMappedBufferClosed()
        val newOffset = math.max(expectedNextOffset.getOrElse(0L), logEndOffset)
        // 00000000000030898257.log 文件
        val logFile = Log.logFile(dir, newOffset)

        if (segments.containsKey(newOffset)) {
          // segment with the same base offset already exists and loaded
          if (activeSegment.baseOffset == newOffset && activeSegment.size == 0) {
            // We have seen this happen (see KAFKA-6388) after shouldRoll() returns true for an
            // active segment of size zero because of one of the indexes is "full" (due to _maxEntries == 0).
            warn(s"Trying to roll a new log segment with start offset $newOffset " +
                 s"=max(provided offset = $expectedNextOffset, LEO = $logEndOffset) while it already " +
                 s"exists and is active with size 0. Size of time index: ${activeSegment.timeIndex.entries}," +
                 s" size of offset index: ${activeSegment.offsetIndex.entries}.")
            deleteSegment(activeSegment)
          } else {
            throw new KafkaException(s"Trying to roll a new log segment for topic partition $topicPartition with start offset $newOffset" +
                                     s" =max(provided offset = $expectedNextOffset, LEO = $logEndOffset) while it already exists. Existing " +
                                     s"segment is ${segments.get(newOffset)}.")
          }
        } else if (!segments.isEmpty && newOffset < activeSegment.baseOffset) {
          throw new KafkaException(
            s"Trying to roll a new log segment for topic partition $topicPartition with " +
            s"start offset $newOffset =max(provided offset = $expectedNextOffset, LEO = $logEndOffset) lower than start offset of the active segment $activeSegment")
        } else {
          val offsetIdxFile = offsetIndexFile(dir, newOffset)
          val timeIdxFile = timeIndexFile(dir, newOffset)
          val txnIdxFile = transactionIndexFile(dir, newOffset)
          for (file <- List(logFile, offsetIdxFile, timeIdxFile, txnIdxFile) if file.exists) {
            warn(s"Newly rolled segment file ${file.getAbsolutePath} already exists; deleting it first")
            Files.delete(file.toPath)
          }

          Option(segments.lastEntry).foreach(_.getValue.onBecomeInactiveSegment())
        }

        // take a snapshot of the producer state to facilitate recovery. It is useful to have the snapshot
        // offset align with the new segment offset since this ensures we can recover the segment by beginning
        // with the corresponding snapshot file and scanning the segment data. Because the segment base offset
        // may actually be ahead of the current producer state end offset (which corresponds to the log end offset),
        // we manually override the state offset here prior to taking the snapshot.
        producerStateManager.updateMapEndOffset(newOffset)
        producerStateManager.takeSnapshot()

        val segment = LogSegment.open(dir,
          baseOffset = newOffset,
          config,
          time = time,
          fileAlreadyExists = false,
          initFileSize = initFileSize,
          preallocate = config.preallocate)
        addSegment(segment)
        // We need to update the segment base offset and append position data of the metadata when log rolls.
        // The next offset should not change.
        updateLogEndOffset(nextOffsetMetadata.messageOffset)
        // schedule an asynchronous flush of the old segment
        scheduler.schedule("flush-log", () => flush(newOffset), delay = 0L)

        info(s"Rolled new log segment at offset $newOffset in ${time.hiResClockMs() - start} ms.")

        segment
      }
    }
  }

可以看到，segment 使用当前 logEndOffset 作为文件名。即 segment 文件用第一条消息的 offset 作文件名。

还有一个和 log 文件同名的 index 文件，index 文件内容是 offset/position，一个 entry 包含 2 个 int，一共 8 字节。

kafka.log.OffsetIndex#append

  /**
   * Append an entry for the given offset/location pair to the index. This entry must have a larger offset than all subsequent entries.
   */
  def append(offset: Long, position: Int) {
    inLock(lock) {
      require(!isFull, "Attempt to append to a full index (size = " + _entries + ").")
      if (_entries == 0 || offset > _lastOffset) {
        trace(s"Adding index entry $offset => $position to ${file.getAbsolutePath}")
        // 相对偏移量
        mmap.putInt((offset - baseOffset).toInt)
        // 消息在 log 文件中的物理地址
        mmap.putInt(position)
        _entries += 1
        _lastOffset = offset
        require(_entries * entrySize == mmap.position(), entries + " entries but file position in index is " + mmap.position() + ".")
      } else {
        throw new InvalidOffsetException(s"Attempt to append an offset ($offset) to position $entries no larger than" +
          s" the last offset appended (${_lastOffset}) to ${file.getAbsolutePath}.")
      }
    }
  }

盗图一张：

http://rocketmq.cloud/zh-cn/docs/design-store.html

而 rocketMQ 的存储与 kafka 不同，分为 commitlog 和 consumequeue：

所有 topic 的消息存储在 commitlog 文件中，commitlog 默认按 1G 分段，文件名按物理偏移量命名。

而索引信息保存在 consumequeue/topic/queue 目录下，一个 entry 固定 20 字节，分别为 8 字节的 commitlog 物理偏移量、4 字节的消息长度、8 字节 tag hashcode。

从代码推出 commitLog 和 consumeQueue 的文件存储格式。

默认文件大小

// org.apache.rocketmq.store.config.MessageStoreConfig
// CommitLog file size, default is 1G
private int mapedFileSizeCommitLog = 1024 * 1024 * 1024;
// ConsumeQueue file size, default is 30W, 大小有 6M
private int mapedFileSizeConsumeQueue = 300000 * ConsumeQueue.CQ_STORE_UNIT_SIZE;

从这个方法可以清晰地看出 commitLog 的存储格式

// org.apache.rocketmq.store.CommitLog#calMsgLength
private static int calMsgLength(int bodyLength, int topicLength, int propertiesLength) {
    final int msgLen = 4 //TOTALSIZE
        + 4 //MAGICCODE
        + 4 //BODYCRC
        + 4 //QUEUEID
        + 4 //FLAG
        + 8 //QUEUEOFFSET
        + 8 //PHYSICALOFFSET
        + 4 //SYSFLAG
        + 8 //BORNTIMESTAMP
        + 8 //BORNHOST
        + 8 //STORETIMESTAMP
        + 8 //STOREHOSTADDRESS
        + 4 //RECONSUMETIMES
        + 8 //Prepared Transaction Offset
        + 4 + (bodyLength > 0 ? bodyLength : 0) //BODY
        + 1 + topicLength //TOPIC
        + 2 + (propertiesLength > 0 ? propertiesLength : 0) //propertiesLength
        + 0;
    return msgLen;
}

当使用分区 offset 拉取消息时，consumeQueue 类似于 index，一个 entry 20 字节，包括 commitLog offset，消息 size，tag 的 hashcode，对于延时消息，tag 字段存的是超时时间。

boolean result = this.putMessagePositionInfo(request.getCommitLogOffset(), request.getMsgSize(), tagsCode, request.getConsumeQueueOffset());

// org.apache.rocketmq.store.ConsumeQueue#putMessagePositionInfo
private boolean putMessagePositionInfo(final long offset, final int size, final long tagsCode, final long cqOffset) {
    if (offset <= this.maxPhysicOffset) {
        return true;
    }

    this.byteBufferIndex.flip();
    this.byteBufferIndex.limit(CQ_STORE_UNIT_SIZE);
    // 8 + 4 + 8 = 20
    this.byteBufferIndex.putLong(offset); // commitLog 的物理位置
    this.byteBufferIndex.putInt(size); // 消息大小
    this.byteBufferIndex.putLong(tagsCode); // 8 字节 tag 哈希值

    ...
}

 broker 为消息的 UNIQ_KEY 和 topic + "#" + key 建立索引，index 文件的结构本质上是一个 hashmap 

// org.apache.rocketmq.store.index.IndexFile
// 40 + 5000000*4 + 20000000*20
int fileTotalSize = IndexHeader.INDEX_HEADER_SIZE + (hashSlotNum * hashSlotSize) + (indexNum * indexSize);
// 一个索引文件大概 420M, 写满了则创建新文件

索引文件就是一个 hashmap，根据 key 查询消息时，遍历所有的 indexFile

文件结构：

文件头
哈希槽
数据部分

// org.apache.rocketmq.store.index.IndexFile#putKey
// 数据 entry 的大小为 20 字节：keyHash, phyOffset, timeDiff, slotValue
this.mappedByteBuffer.putInt(absIndexPos, keyHash);
this.mappedByteBuffer.putLong(absIndexPos + 4, phyOffset);
this.mappedByteBuffer.putInt(absIndexPos + 4 + 8, (int) timeDiff);
// 这里的 slotValue 是上一条索引的编号
this.mappedByteBuffer.putInt(absIndexPos + 4 + 8 + 4, slotValue);
// 当前索引的编号写到哈希槽
this.mappedByteBuffer.putInt(absSlotPos, this.indexHeader.getIndexCount());

rocketMQ 写完 commitLog 后，写 consumeQueue 和 indexFile 是一个异步的过程，在

org.apache.rocketmq.store.DefaultMessageStore.ReputMessageService#doReput

中触发

// org.apache.rocketmq.store.DefaultMessageStore#DefaultMessageStore
this.dispatcherList = new LinkedList<>();
this.dispatcherList.addLast(new CommitLogDispatcherBuildConsumeQueue());
this.dispatcherList.addLast(new CommitLogDispatcherBuildIndex());

// org.apache.rocketmq.store.DefaultMessageStore#doDispatch
public void doDispatch(DispatchRequest req) {
    for (CommitLogDispatcher dispatcher : this.dispatcherList) {
        dispatcher.dispatch(req);
    }
}