Zookeeper(五)持久化快照
用途
快照文件是指定时间间隔对zookeeper服务器上的节点数据的序列化后备份到磁盘中,快照文件不一定是最新的
如果zk集群挂了,可能会用到它来复原
基本术语
- DataTree:zookeeper中数据存储结构
http://www.cnblogs.com/leesf456/p/6179118.html
问题(别人的问题,我暂时没想到)
- 什么时候生成快照,快照什么时候被删除,会不会被删除
- 如果zk集群挂了是从哪里恢复,FileSnap还是FileTxn,FileSnap都不一定是最新的,zxid怎么保证
- 什么时候进行快照记录
实现
SnapShot
public interface SnapShot {
//从最新且有效的快照中反序列化一个dataTree
long deserialize(DataTree dt, Map<Long, Integer> sessions) throws IOException;
//持久化dataTree和会话到一个持久化存储中
void serialize(DataTree dt, Map<Long, Integer> sessions, File name, boolean fsync)
throws IOException;
//查询最近的快照文件
File findMostRecentSnapshot() throws IOException;
//立刻释放快照文件中的资源
void close() throws IOException;
}
FileSnap
public class FileSnap implements SnapShot {
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
//100个有效且最新的快照文件
List<File> snapList = findNValidSnapshots(100);
if (snapList.size() == 0) {
return -1L;
}
File snap = null;
boolean foundValid = false;
//找到一个有效的快照 就break
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);
}
//获取该有效快照的zxid
dt.lastProcessedZxid = Util.getZxidFromName(snap.getName(), SNAPSHOT_FILE_PREFIX);
return dt.lastProcessedZxid;
}
}