持久化

zookeeper在将写请求应用到内存数据库之前，会首先记录这次事务，从而生成事务日志。当事务日志的数量达到一定规模之后，会对内存数据库打快照，从而生成快照文件。

持久化相关的类

TxnLog包含了对事务日志的操作方法
FileTnxLog实现了TxnLog接口
SnapShot包含了对快照日志的操作方法
FileSnapShot实现了SnapShot接口
FileTnxSnapLog包含TnxLog和SnapShot

事务日志

TxnLog

public interface TxnLog {
	
	// 设置服务端的统计信息
    void setServerStats(ServerStats serverStats);
    
     // 滚动当前正在写的日志文件
    void rollLog() throws IOException;
    
    // 追加指定日志
    boolean append(TxnHeader hdr, Record r) throws IOException;

    // 返回可以从指定事务id开始读取之后事务日志的迭代器
    TxnIterator read(long zxid) throws IOException;
    
    // 获取最近记录的事务日志的事务id
    long getLastLoggedZxid() throws IOException;
    
    // 截取事务日志到指定的事务id
    boolean truncate(long zxid) throws IOException;
    
    // 返回当前事务日志所属的dbid
    long getDbId() throws IOException;
    
    // 提交日志，保证日志能够持久化
    void commit() throws IOException;

    /**
     *
     * @return transaction log's elapsed sync time in milliseconds
     */
    long getTxnLogSyncElapsedTime();
   
    // 关闭日志
    void close() throws IOException;
    
    // 读取事务日志的迭代器
    public interface TxnIterator {
        // 获取事务记录的头
        TxnHeader getHeader();
        
        // 获取事务记录的内容
        Record getTxn();
     
        // 移动到下一个日志
        boolean next() throws IOException;
        
        // 关闭文件
        void close() throws IOException;
        
        // 获取当前迭代器能够返回的日志预估大小
        long getStorageSize() throws IOException;
    }
}

文件组成

一个事务日志文件一共有三个部分：
FileHeader TxnList ZeroPad

FileHeader

FileHeader {
	magic 4bytes
	version 4bytes
	dbid 8bytes
}

TxnList

TxnList存放事务记录，包含多个Txn

Txn {
	checksum 8bytes
	Txnlen 4bytes
	TxnHeader {
		sessionid 8bytes
        cxid 4bytes
        zxid 8bytes
        time 8bytes
        type 4bytes
	}
	Record
	0x42
}

ZeroPad

在文件末尾添加0

FileTxnLog

主要属性

// 文件头部的magic数字
public final static int TXNLOG_MAGIC =
        ByteBuffer.wrap("ZKLG".getBytes()).getInt();
// 上一个日志记录的事务id
long lastZxidSeen;

// 日志文件流
volatile BufferedOutputStream logStream = null;
volatile OutputArchive oa;
volatile FileOutputStream fos = null;

// 事务日志文件存放目录
File logDir;

// 当提交时是否确保同步到磁盘
private final boolean forceSync = !System.getProperty("zookeeper.forceSync", "yes").equals("no");
// 内存数据库id
long dbId;
// 等待flush的流
private LinkedList<FileOutputStream> streamsToFlush =
    new LinkedList<FileOutputStream>();

内部类

PositionInputStream

PositionInputStream是一个能够记录当前读入位置的输入流，实现方式就是每次从输入流中读取数据后，就会增加position

static class PositionInputStream extends FilterInputStream {
   long position;
   protected PositionInputStream(InputStream in) {
       super(in);
       position = 0;
   }

   @Override
   public int read() throws IOException {
       int rc = super.read();
       if (rc > -1) {
           position++;
       }
       return rc;
   }

   public int read(byte[] b) throws IOException {
       int rc = super.read(b);
       if (rc > 0) {
           position += rc;
       }
       return rc;
   }

   @Override
   public int read(byte[] b, int off, int len) throws IOException {
       int rc = super.read(b, off, len);
       if (rc > 0) {
           position += rc;
       }
       return rc;
   }

   @Override
   public long skip(long n) throws IOException {
       long rc = super.skip(n);
       if (rc > 0) {
           position += rc;
       }
       return rc;
   }
   public long getPosition() {
       return position;
   }

   @Override
   public boolean markSupported() {
       return false;
   }

   @Override
   public void mark(int readLimit) {
       throw new UnsupportedOperationException("mark");
   }

   @Override
   public void reset() {
       throw new UnsupportedOperationException("reset");
   }
}

FileTxnIterator

FileTxnIterator的作用就是从事务日志中读取事务记录
首先，会解析日志目录下的文件，提取文件名中的zxid作为当前日志包含事务日志中的最小事务id
然后，将zxid大于等于指定事务id以及小于zxid的最大zxid事务日志文件保存下下来，并根据zxid递减排序

public static class FileTxnIterator implements TxnLog.TxnIterator {
	  // 日志文件目录
      File logDir;
      // 当前记录的事务id
	  long zxid;
	  // 当前记录的头
	  TxnHeader hdr;
	  // 当前记录的内容
	  Record record;
	  File logFile;
	  InputArchive ia;
	  static final String CRC_ERROR="CRC check failed";
	
	  PositionInputStream inputStream=null;
	  // 包含有大于指定事务id的事务文件
	  private ArrayList<File> storedFiles;
	  
	  public FileTxnIterator(File logDir, long zxid, boolean fastForward)
	          throws IOException {
	      this.logDir = logDir;
	      this.zxid = zxid;
	      // 找到可能包含指定事务id的所有日志，并且排序
	      init();
	
		  // 如果设置了fastForward，会快速定位到该zxid位置，否则会定位到可能包含该zxid的日志的头部
	      if (fastForward && hdr != null) {
	          while (hdr.getZxid() < zxid) {
	              if (!next())
	                  break;
	          }
	      }
	  }
	
	  /**
	   * create an iterator over a transaction database directory
	   * @param logDir the transaction database directory
	   * @param zxid the zxid to start reading from
	   * @throws IOException
	   */
	  public FileTxnIterator(File logDir, long zxid) throws IOException {
	      this(logDir, zxid, true);
	  }
	
	  void init() throws IOException {
	      storedFiles = new ArrayList<File>();
	      // 根据文件名中的zxid进行递减排序
	      List<File> files = Util.sortDataDir(FileTxnLog.getLogFiles(logDir.listFiles(), 0), LOG_FILE_PREFIX, false);
	      // 筛选出大于指定zxid的日志和第一个小于zxid的日志文件
	      for (File f: files) {
	          if (Util.getZxidFromName(f.getName(), LOG_FILE_PREFIX) >= zxid) {
	              storedFiles.add(f);
	          }
	          // add the last logfile that is less than the zxid
	          else if (Util.getZxidFromName(f.getName(), LOG_FILE_PREFIX) < zxid) {
	              storedFiles.add(f);
	              break;
	          }
	      }
	      // 创建一个指向最后一个文件的输入流
	      goToNextLog();
	      // 定位到该文件中的第一条记录
	      next();
	  }
	
	  // 返回当前还没有处理的日志的预估大小
	  public long getStorageSize() {
	      long sum = 0;
	      for (File f : storedFiles) {
	          sum += f.length();
	      }
	      return sum;
	  }
	
	  // 移动到下一个日志文件，按照zxid递增的顺序来遍历文件
	  private boolean goToNextLog() throws IOException {
	      if (storedFiles.size() > 0) {
	          this.logFile = storedFiles.remove(storedFiles.size()-1);
	          ia = createInputArchive(this.logFile);
	          return true;
	      }
	      return false;
	  }
	
	
	  // 移动到下一个事务日志
	  public boolean next() throws IOException {
	      if (ia == null) {
	          return false;
	      }
	      try {
	      	  // 从日志文件输入流中恢复出日志记录
	          long crcValue = ia.readLong("crcvalue");
	          byte[] bytes = Util.readTxnBytes(ia);
	          // Since we preallocate, we define EOF to be an
	          if (bytes == null || bytes.length==0) {
	              throw new EOFException("Failed to read " + logFile);
	          }
	          // EOF or corrupted record
	          // validate CRC
	          Checksum crc = makeChecksumAlgorithm();
	          crc.update(bytes, 0, bytes.length);
	          if (crcValue != crc.getValue())
	              throw new IOException(CRC_ERROR);
	          hdr = new TxnHeader();
	          record = SerializeUtils.deserializeTxn(bytes, hdr);
	      } catch (EOFException e) {
	          // EOFException异常代表当前文件读取完毕，尝试调用toToNextLog切换到下一个文件
	          LOG.debug("EOF exception " + e);
	          inputStream.close();
	          inputStream = null;
	          ia = null;
	          hdr = null;
	          // this means that the file has ended
	          // we should go to the next file
	          if (!goToNextLog()) {
	              return false;
	          }
	          // if we went to the next log file, we should call next() again
	          return next();
	      } catch (IOException e) {
	          inputStream.close();
	          throw e;
	      }
	      return true;
	  }
	
	  /**
	   * return the current header
	   * @return the current header that
	   * is read
	   */
	  public TxnHeader getHeader() {
	      return hdr;
	  }
	
	  /**
	   * return the current transaction
	   * @return the current transaction
	   * that is read
	   */
	  public Record getTxn() {
	      return record;
	  }
	
	  /**
	   * close the iterator
	   * and release the resources.
	   */
	  public void close() throws IOException {
	      if (inputStream != null) {
	          inputStream.close();
	      }
	  }
	}

重要方法

append

// FileTxnLog
public synchronized boolean append(TxnHeader hdr, Record txn)
        throws IOException
{
    if (hdr == null) {
        return false;
    }
    // 如果当前的事务id小于日志中最大的事务id
    // 从这里可以看出事务日志中的事务日志的事务id并不是一定递增的
    if (hdr.getZxid() <= lastZxidSeen) {
        LOG.warn("Current zxid " + hdr.getZxid()
                + " is <= " + lastZxidSeen + " for "
                + hdr.getType());
    } else {
    	// 更新日志事务id的最大值
        lastZxidSeen = hdr.getZxid();
    }
    // 当前没有创建日志文件
    if (logStream==null) {
       if(LOG.isInfoEnabled()){
            LOG.info("Creating new log file: " + Util.makeLogName(hdr.getZxid()));
       }
	   // 在logDir目录下创建log.当前事务记录id的日志文件
       logFileWrite = new File(logDir, Util.makeLogName(hdr.getZxid()));
       // 创建文件的输出流
       fos = new FileOutputStream(logFileWrite);
       logStream=new BufferedOutputStream(fos);
       oa = BinaryOutputArchive.getArchive(logStream);
       FileHeader fhdr = new FileHeader(TXNLOG_MAGIC,VERSION, dbId);
       // 将文件头写入到文件中
       fhdr.serialize(oa, "fileheader");
       // Make sure that the magic number is written before padding.
       logStream.flush();
       filePadding.setCurrentSize(fos.getChannel().position());
	   // 将当前流加入到需要flush的流列表中
       streamsToFlush.add(fos);
    }
    filePadding.padFile(fos.getChannel());
    // 将当前事务记录的头和内容序列化成字节数组
    byte[] buf = Util.marshallTxnEntry(hdr, txn);
    if (buf == null || buf.length == 0) {
        throw new IOException("Faulty serialization for header " +
                "and txn");
    }
    // 计算头部和内容的校验和
    Checksum crc = makeChecksumAlgorithm();
    crc.update(buf, 0, buf.length);
    // 将校验和写入到文件中
    oa.writeLong(crc.getValue(), "txnEntryCRC");
    // 将事务头部、内容以及EOR写入到文件中
    Util.writeTxnBytes(oa, buf);

    return true;
}

上述方法首先判断当前是否有正在写的日志文件，如果没有会使用当前事务id生成log.事务id的事务日志文件
然后按照日志的格式，将当前事务记录写入到文件中

commit

// FileTxnLog
public synchronized void commit() throws IOException {
	// 当前有正在写入的日志文件，将数据刷写到操作系统，此时并不保证已经落盘
    if (logStream != null) {
        logStream.flush();
    }
    // 遍历streamToFlush，当前正在写入的流也已经加入到了streamsToFlush
    for (FileOutputStream log : streamsToFlush) {
    	// 刷写到操作系统
        log.flush();
        // 如果需要保证落盘，开始落盘
        if (forceSync) {
            long startSyncNS = System.nanoTime();

            FileChannel channel = log.getChannel();
            channel.force(false);
			
			// 记录当前文件落盘时间
            syncElapsedMS = TimeUnit.NANOSECONDS.toMillis(System.nanoTime() - startSyncNS);
            // 落盘时间超过指定的阈值，记录超出此时，并打印提示日志
            if (syncElapsedMS > fsyncWarningThresholdMS) {
                if(serverStats != null) {
                    serverStats.incrementFsyncThresholdExceedCount();
                }
                LOG.warn("fsync-ing the write ahead log in "
                        + Thread.currentThread().getName()
                        + " took " + syncElapsedMS
                        + "ms which will adversely effect operation latency. "
                        + "File size is " + channel.size() + " bytes. "
                        + "See the ZooKeeper troubleshooting guide");
            }
        }
    }
    // 关闭streamsToFlush中的流，并清空streamsToFlush
    while (streamsToFlush.size() > 1) {
        streamsToFlush.removeFirst().close();
    }
}

该方法的作用是保证生成的事务日志文件都刷写到操作系统中，如果设置了强制落盘，会落盘

read

返回一个迭代器，该迭代器定位到了大于指定zxid的事务记录

public TxnIterator read(long zxid, boolean fastForward) throws IOException {
   return new FileTxnIterator(logDir, zxid, fastForward);
}

rollLog

// FileTxnLog
public synchronized void rollLog() throws IOException {
	// flush当前写的事务日志文件流
    if (logStream != null) {
        this.logStream.flush();
        this.logStream = null;
        oa = null;
    }
}

滚动当前日志

close

// FileTxnLog
public synchronized void close() throws IOException {
	// 关闭日志文件流
    if (logStream != null) {
        logStream.close();
    }
    for (FileOutputStream log : streamsToFlush) {
        log.close();
    }
}

truncate

public boolean truncate(long zxid) throws IOException {
    FileTxnIterator itr = null;
    try {
        itr = new FileTxnIterator(this.logDir, zxid);
        PositionInputStream input = itr.inputStream;
        if(input == null) {
            throw new IOException("No log files found to truncate! This could " +
                    "happen if you still have snapshots from an old setup or " +
                    "log files were deleted accidentally or dataLogDir was changed in zoo.cfg.");
        }
        long pos = input.getPosition();
        // now, truncate at the current position
        // 截断当前文件
        RandomAccessFile raf=new RandomAccessFile(itr.logFile,"rw");
        raf.setLength(pos);
        raf.close();
        // 将之后的日志文件全部删除
        while(itr.goToNextLog()) {
            if (!itr.logFile.delete()) {
                LOG.warn("Unable to truncate {}", itr.logFile);
            }
        }
    } finally {
        close(itr);
    }
    return true;
}

快照日志

SnapShot

public interface SnapShot {
    
    // 从最近的快照中恢复DateTree和session
    long deserialize(DataTree dt, Map<Long, Integer> sessions) 
        throws IOException;
    
    // 将当前的DataTree和session序列化到文件中
    void serialize(DataTree dt, Map<Long, Integer> sessions, 
            File name) 
        throws IOException;
    
    // 返回最近生成的快照文件
    File findMostRecentSnapshot() throws IOException;
    
    // 关闭序列化
    void close() throws IOException;
} 

FileSnap

FileSnap是SnapShot的实现

主要属性

// 快照文件目录
File snapDir;
// 是否关闭快照
private volatile boolean close = false;
// 版本
private static final int VERSION = 2;
private static final long dbId = -1;
private static final Logger LOG = LoggerFactory.getLogger(FileSnap.class);
// 快照文件头部的magic字段
public final static int SNAP_MAGIC
        = ByteBuffer.wrap("ZKSN".getBytes()).getInt();

// 快照文件前缀
public static final String SNAPSHOT_FILE_PREFIX = "snapshot";

重要方法

serialize

// FileSnap
public synchronized void serialize(DataTree dt, Map<Long, Integer> sessions, File snapShot)
            throws IOException {
    if (!close) {
        try (OutputStream sessOS = new BufferedOutputStream(new FileOutputStream(snapShot));
        	// CheckedOutputStream是带有校验的输出流，一边写一边更新校验值
             CheckedOutputStream crcOut = new CheckedOutputStream(sessOS, new Adler32())) {
            //CheckedOutputStream cout = new CheckedOutputStream()
            OutputArchive oa = BinaryOutputArchive.getArchive(crcOut);
            // 文件头
            FileHeader header = new FileHeader(SNAP_MAGIC, VERSION, dbId);
            // 实际进行序列化
            serialize(dt, sessions, oa, header);
            long val = crcOut.getChecksum().getValue();
            // 写出校验值
            oa.writeLong(val, "val");
            oa.writeString("/", "path");
            sessOS.flush();
        }
    }
}

// FileSnap
protected void serialize(DataTree dt,Map<Long, Integer> sessions,
            OutputArchive oa, FileHeader header) throws IOException {
    // this is really a programmatic error and not something that can
    // happen at runtime
    if(header==null)
        throw new IllegalStateException(
                "Snapshot's not open for writing: uninitialized header");
    // 写入文件头
    header.serialize(oa, "fileheader");
    // 写入DataTree 和 session
    SerializeUtils.serializeSnapshot(dt,oa,sessions);
}

// SerializeUtils
public static void serializeSnapshot(DataTree dt,OutputArchive oa,
            Map<Long, Integer> sessions) throws IOException {
    HashMap<Long, Integer> sessSnap = new HashMap<Long, Integer>(sessions);
    // 写入当前事务的个数
    oa.writeInt(sessSnap.size(), "count");
    // 遍历事务，写入事务的id和过期时间
    for (Entry<Long, Integer> entry : sessSnap.entrySet()) {
        oa.writeLong(entry.getKey().longValue(), "id");
        oa.writeInt(entry.getValue().intValue(), "timeout");
    }
    // 写入dataTree
    dt.serialize(oa, "tree");
}

下面看下如何序列化DataTree

public void serialize(OutputArchive oa, String tag) throws IOException {
	// 序列化aclCache
    aclCache.serialize(oa);
    // 序列化节点
    serializeNode(oa, new StringBuilder(""));
    // / marks end of stream
    // we need to check if clear had been called in between the snapshot.
    if (root != null) {
        oa.writeString("/", "path");
    }
}

void serializeNode(OutputArchive oa, StringBuilder path) throws IOException {
    String pathString = path.toString();
    // 获取路径对应的节点
    DataNode node = getNode(pathString);
    if (node == null) {
        return;
    }
    String children[] = null;
    // 拷贝当前节点
    DataNode nodeCopy;
    synchronized (node) {
        StatPersisted statCopy = new StatPersisted();
        copyStatPersisted(node.stat, statCopy);
        //we do not need to make a copy of node.data because the contents
        //are never changed
        nodeCopy = new DataNode(node.data, node.acl, statCopy);
        Set<String> childs = node.getChildren();
        children = childs.toArray(new String[childs.size()]);
    }
    // 序列化节点
    serializeNodeData(oa, pathString, nodeCopy);
    path.append('/');
    int off = path.length();
    // 遍历子节点并进行序列化
    for (String child : children) {
        // since this is single buffer being resused
        // we need
        // to truncate the previous bytes of string.
        path.delete(off, Integer.MAX_VALUE);
        path.append(child);
        serializeNode(oa, path);
    }
}

下面看下如何序列化DataTree中的一个节点

public void serializeNodeData(OutputArchive oa, String path, DataNode node) throws IOException {
	// 首先写入路径
    oa.writeString(path, "path");
    // 写入节点内容
    oa.writeRecord(node, "node");
}

// DataNode
synchronized public void serialize(OutputArchive archive, String tag)
            throws IOException {
   archive.startRecord(this, "node");
   // 写入数据
   archive.writeBuffer(data, "data");
   // 写入acl值
   archive.writeLong(acl, "acl");
   // 写入状态，比如修改时间，创建时间等
   stat.serialize(archive, "statpersisted");
   archive.endRecord(this, "node");
}

findNValidSnapshots

寻找最近的n个快照文件

private List<File> findNValidSnapshots(int n) throws IOException {
	// 遍历快照日志下的文件，从文件名中提取事务id，按照事务id递减排序文件
    List<File> files = Util.sortDataDir(snapDir.listFiles(), SNAPSHOT_FILE_PREFIX, false);
    int count = 0;
    List<File> list = new ArrayList<File>();
    // 从这些文件中找到校验通过的n个文件
    for (File f : files) {
        // we should catch the exceptions
        // from the valid snapshot and continue
        // until we find a valid one
        try {
            if (Util.isValidSnapshot(f)) {
                list.add(f);
                count++;
                if (count == n) {
                    break;
                }
            }
        } catch (IOException e) {
            LOG.info("invalid snapshot " + f, e);
        }
    }
    return list;
}

findNRecentSnapshots

遍历快照目录下的快照文件，寻找最近的n个快照

public List<File> findNRecentSnapshots(int n) throws IOException {
    List<File> files = Util.sortDataDir(snapDir.listFiles(), SNAPSHOT_FILE_PREFIX, false);
    int count = 0;
    List<File> list = new ArrayList<File>();
    for (File f: files) {
        if (count == n)
            break;
        if (Util.getZxidFromName(f.getName(), SNAPSHOT_FILE_PREFIX) != -1) {
            count++;
            list.add(f);
        }
    }
    return list;
}

findMostRecentSnapshot

遍历快照目录下的快照文件，找到最近的快照文件

public File findMostRecentSnapshot() throws IOException {
    List<File> files = findNValidSnapshots(1);
    if (files.size() == 0) {
        return null;
    }
    return files.get(0);
}

deserialize

public long deserialize(DataTree dt, Map<Long, Integer> sessions)
            throws IOException {
     // we run through 100 snapshots (not all of them)
     // if we cannot get it running within 100 snapshots
     // we should  give up
     // 找到最近有效的100个快照文件
     // 这里找到的文件并不一定是完全有效的
     List<File> snapList = findNValidSnapshots(100);
     if (snapList.size() == 0) {
         return -1L;
     }
     File snap = null;
     boolean foundValid = false;
     // 遍历找到的快照文件，读取文件，并且计算校验和，校验文件是否有效
     // 序列化该文件到DateTree session
     for (int i = 0, snapListSize = snapList.size(); i < snapListSize; i++) {
         snap = snapList.get(i);
         LOG.info("Reading snapshot " + snap);
         try (InputStream snapIS = new BufferedInputStream(new FileInputStream(snap));
              CheckedInputStream crcIn = new CheckedInputStream(snapIS, new Adler32())) {
             InputArchive ia = BinaryInputArchive.getArchive(crcIn);
             deserialize(dt, sessions, ia);
             long checkSum = crcIn.getChecksum().getValue();
             long val = ia.readLong("val");
             if (val != checkSum) {
                 throw new IOException("CRC corruption in snapshot :  " + snap);
             }
             foundValid = true;
             break;
         } catch (IOException e) {
             LOG.warn("problem reading snap file " + snap, e);
         }
     }
     if (!foundValid) {
         throw new IOException("Not able to find valid snapshots in " + snapDir);
     }
     dt.lastProcessedZxid = Util.getZxidFromName(snap.getName(), SNAPSHOT_FILE_PREFIX);
     return dt.lastProcessedZxid;
 }

FileTxnSnapLog

主要属性

// 事务日志目录
private final File dataDir;
// 快照文件目录
private final File snapDir;
// 事务日志
private TxnLog txnLog;
// 快照
private SnapShot snapLog;
private final boolean trustEmptySnapshot;
public final static int VERSION = 2;
public final static String version = "version-";

private static final Logger LOG = LoggerFactory.getLogger(FileTxnSnapLog.class);
// 当目录不存在时是否自动创建目录
public static final String ZOOKEEPER_DATADIR_AUTOCREATE =
        "zookeeper.datadir.autocreate";

public static final String ZOOKEEPER_DATADIR_AUTOCREATE_DEFAULT = "true";

重要方法

构造方法

// FileTxnSnapLog
public FileTxnSnapLog(File dataDir, File snapDir) throws IOException {
    LOG.debug("Opening datadir:{} snapDir:{}", dataDir, snapDir);
	
	// 事务日志文件目录和快照文件目录
	// ${dataDir}/version-2
    this.dataDir = new File(dataDir, version + VERSION);
    this.snapDir = new File(snapDir, version + VERSION);

	// 判断是否自动创建目录
    boolean enableAutocreate = Boolean.valueOf(
            System.getProperty(ZOOKEEPER_DATADIR_AUTOCREATE,
                    ZOOKEEPER_DATADIR_AUTOCREATE_DEFAULT));

    trustEmptySnapshot = Boolean.getBoolean(ZOOKEEPER_SNAPSHOT_TRUST_EMPTY);
    LOG.info(ZOOKEEPER_SNAPSHOT_TRUST_EMPTY + " : " + trustEmptySnapshot);

	// 创建目录
    if (!this.dataDir.exists()) {
        if (!enableAutocreate) {
            throw new DatadirException("Missing data directory "
                    + this.dataDir
                    + ", automatic data directory creation is disabled ("
                    + ZOOKEEPER_DATADIR_AUTOCREATE
                    + " is false). Please create this directory manually.");
        }

        if (!this.dataDir.mkdirs()) {
            throw new DatadirException("Unable to create data directory "
                    + this.dataDir);
        }
    }
    if (!this.dataDir.canWrite()) {
        throw new DatadirException("Cannot write to data directory " + this.dataDir);
    }

    if (!this.snapDir.exists()) {
        // by default create this directory, but otherwise complain instead
        // See ZOOKEEPER-1161 for more details
        if (!enableAutocreate) {
            throw new DatadirException("Missing snap directory "
                    + this.snapDir
                    + ", automatic data directory creation is disabled ("
                    + ZOOKEEPER_DATADIR_AUTOCREATE
                    + " is false). Please create this directory manually.");
        }

        if (!this.snapDir.mkdirs()) {
            throw new DatadirException("Unable to create snap directory "
                    + this.snapDir);
        }
    }
    if (!this.snapDir.canWrite()) {
        throw new DatadirException("Cannot write to snap directory " + this.snapDir);
    }

    // check content of transaction log and snapshot dirs if they are two different directories
    // See ZOOKEEPER-2967 for more details
    if(!this.dataDir.getPath().equals(this.snapDir.getPath())){
    	// 当事务日志目录和快照目录不同时，进行校验
    	// 判断日志目录下是否有快照文件
        checkLogDir();
        // 判断快照目录下是否有日志文件
        checkSnapDir();
    }

    txnLog = new FileTxnLog(this.dataDir);
    snapLog = new FileSnap(this.snapDir);
}

构造方法主要就是创建TxnLog和SnapLog

append

将请求透传给TxnLog

public boolean append(Request si) throws IOException {
    return txnLog.append(si.getHdr(), si.getTxn());
}

save

使用DataTree中最新的事务id生成快照文件名，使用SnapLog来生成快照

public void save(DataTree dataTree,
            ConcurrentHashMap<Long, Integer> sessionsWithTimeouts)
        throws IOException {
    long lastZxid = dataTree.lastProcessedZxid;
    File snapshotFile = new File(snapDir, Util.makeSnapshotName(lastZxid));
    LOG.info("Snapshotting: 0x{} to {}", Long.toHexString(lastZxid),
            snapshotFile);
    snapLog.serialize(dataTree, sessionsWithTimeouts, snapshotFile);

}

rollLog

public void rollLog() throws IOException {
   txnLog.rollLog();
}

close

public void close() throws IOException {
   txnLog.close();
    snapLog.close();
}

truncateLog

关闭TxnLog和SnapShot，然后截断日志文件，重新创建TxnLog SnapShot

public boolean truncateLog(long zxid) throws IOException {
    // close the existing txnLog and snapLog
    close();

    // truncate it
    FileTxnLog truncLog = new FileTxnLog(dataDir);
    boolean truncated = truncLog.truncate(zxid);
    truncLog.close();

    // re-open the txnLog and snapLog
    // I'd rather just close/reopen this object itself, however that 
    // would have a big impact outside ZKDatabase as there are other
    // objects holding a reference to this object.
    txnLog = new FileTxnLog(dataDir);
    snapLog = new FileSnap(snapDir);

    return truncated;
}

restore

public long restore(DataTree dt, Map<Long, Integer> sessions,
                        PlayBackListener listener) throws IOException {
    // 使用最近的有效快照文件来还原DataTree和session
    long deserializeResult = snapLog.deserialize(dt, sessions);
    FileTxnLog txnLog = new FileTxnLog(dataDir);
    if (-1L == deserializeResult) {
        /* this means that we couldn't find any snapshot, so we need to
         * initialize an empty database (reported in ZOOKEEPER-2325) */
        if (txnLog.getLastLoggedZxid() != -1) {
            // ZOOKEEPER-3056: provides an escape hatch for users upgrading
            // from old versions of zookeeper (3.4.x, pre 3.5.3).
            if (!trustEmptySnapshot) {
                throw new IOException(EMPTY_SNAPSHOT_WARNING + "Something is broken!");
            } else {
                LOG.warn(EMPTY_SNAPSHOT_WARNING + "This should only be allowed during upgrading.");
            }
        }
        /* TODO: (br33d) we should either put a ConcurrentHashMap on restore()
         *       or use Map on save() */
        save(dt, (ConcurrentHashMap<Long, Integer>)sessions);
        /* return a zxid of zero, since we the database is empty */
        return 0;
    }
	
	// 接着使用事务日志来恢复
    return fastForwardFromEdits(dt, sessions, listener);
}

public long fastForwardFromEdits(DataTree dt, Map<Long, Integer> sessions,
                                     PlayBackListener listener) throws IOException {
    // 经过快照还原后，dt.lastProcessedZxid即是快照文件中存放的最大事务id
    // 从事务日志中接着该事务id接着还原
    TxnIterator itr = txnLog.read(dt.lastProcessedZxid+1);
    long highestZxid = dt.lastProcessedZxid;
    TxnHeader hdr;
    try {
        while (true) {
            // iterator points to
            // the first valid txn when initialized
            hdr = itr.getHeader();
            if (hdr == null) {
                //empty logs
                return dt.lastProcessedZxid;
            }
            // 更新目前处理过的最大事务id
            if (hdr.getZxid() < highestZxid && highestZxid != 0) {
                LOG.error("{}(highestZxid) > {}(next log) for type {}",
                        highestZxid, hdr.getZxid(), hdr.getType());
            } else {
                highestZxid = hdr.getZxid();
            }
            try {
            	// 处理事务
                processTransaction(hdr,dt,sessions, itr.getTxn());
            } catch(KeeperException.NoNodeException e) {
               throw new IOException("Failed to process transaction type: " +
                     hdr.getType() + " error: " + e.getMessage(), e);
            }
            listener.onTxnLoaded(hdr, itr.getTxn());
            if (!itr.next())
                break;
        }
    } finally {
        if (itr != null) {
            itr.close();
        }
    }
    return highestZxid;
}