Spark主备切换机制原理

Master实际上可以配置两个，那么在spark原生的standalone上也是支持Master主备切换的，也就是说，当Active Master节点挂掉之后，我们可以将Standby Master切换为Active Master

Spark Master的主备切换可以基于两种切换机制，一种是文件系统，一种是基于Zookeeper,基于文件系统的机制，是Active Master挂掉后，需要我们手动去切换到Standby Master上，基于Zookeeper机制，呆以实现自动切换。

所以这里说的主备切换机制，其实指的是在Active Master挂掉之后，切换到Standby Master时，Master会做哪些操作

1.使用持久化引挚(FileSystemPersistence或者是ZookeeperPersisitence)去读取持久化的storedApps,storedDriver,storedWorker,

2.判断上面的三个持久化的storedApps,storedDriver,storedWorker,

有任何一个不为空，就将持久化有Application,Driver,Worker的信息重新注册，注册到Master内部的缓存结构中。

3.将Application和Worker的状态都修改为UNKNOWN,然后向Application对应的Driver，Worker发送Standby Master的地址.

4.Driver,Worker,理论上讲，如果他们目前是正常工作的话，那么在收到Master发送来的地址后，就会返回响应给新的Master。

5.此时，Master在陆续接收到Driver,Worker发送来的响应消息之后，会使用completeRecovery()对没有收到发送响应消息的Driver,Worker进行处理，过滤掉他们的信息。如下：

// Kill off any workers and apps that didn’t respond to us. workers.filter(.state == WorkerState.UNKNOWN).foreach(removeWorker) apps.filter(.state == ApplicationState.UNKNOWN).foreach(finishApplication) // Reschedule drivers which were not claimed by any workers drivers.filter(_.worker.isEmpty).foreach { d => logWarning(s"Driver ${d.id} was not found after master recovery") if (d.desc.supervise) { logWarning(s"Re-launching ${d.id}") relaunchDriver(d) } else { removeDriver(d.id, DriverState.ERROR, None) logWarning(s"Did not re-launch ${d.id} because it was not supervised") } 12345678910

6.调用Master的schedule()，对正在等待调度的Driver,Application进行调度，比如在某个Worker上启动Driver,或者为Application在Worker上启动Executor。

state = RecoveryState.ALIVE schedule()