类 ml.shifu.guagua.master.MasterContext
的使用

使用 MasterContext 的软件包

ml.shifu.guagua.master

This package contains guagua master implementation.

ml.shifu.guagua.master 中 MasterContext 的使用

返回 MasterContext 的 ml.shifu.guagua.master 中的方法

MasterContext<MASTER_RESULT,WORKER_RESULT>

GuaguaMasterService.getContext()

参数类型为 MasterContext 的 ml.shifu.guagua.master 中的方法

MASTER_RESULT	MasterComputable.compute(MasterContext<MASTER_RESULT,WORKER_RESULT> context) Master computation for each iteration.
protected MASTER_RESULT	GuaguaMasterService.iterate(MasterContext<MASTER_RESULT,WORKER_RESULT> context, int iteration, Progressable progress) Call each iteration computation and preIteration, postIteration in interceptors.
void	TraceMasterInterceptor.postApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	NettyMasterCoordinator.postApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context) Wait for unregister message for all workers and then clean all znodes existing for this job.
void	MemoryStatsMasterInterceptor.postApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	MasterTimer.postApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	MasterInterceptor.postApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context) The hook point after any computation logic.
void	MasterComputableTimer.postApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	BasicMasterInterceptor.postApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	AbstractMasterCoordinator.postApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	TraceMasterInterceptor.postIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	NettyMasterCoordinator.postIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context) Update master computable result to master znode.
void	MemoryStatsMasterInterceptor.postIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	MasterTimer.postIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	MasterInterceptor.postIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context) The hook point after computation of each iteration.
void	MasterComputableTimer.postIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	LocalMasterCoordinator.postIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	BasicMasterInterceptor.postIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	AbstractMasterCoordinator.postIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	TraceMasterInterceptor.preApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	SyncMasterCoordinator.preApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	NettyMasterCoordinator.preApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context) Do initialization and fail-over checking before all iterations.
void	MemoryStatsMasterInterceptor.preApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	MasterTimer.preApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	MasterInterceptor.preApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context) The hook point before any computation logic.
void	MasterComputableTimer.preApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	LocalMasterCoordinator.preApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	BasicMasterInterceptor.preApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	AsyncMasterCoordinator.preApplication(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	TraceMasterInterceptor.preIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	SyncMasterCoordinator.preIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	NettyMasterCoordinator.preIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context) Wait for all workers done in current iteration.
void	MemoryStatsMasterInterceptor.preIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	MasterTimer.preIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	MasterInterceptor.preIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context) The hook point before computation of each iteration.
void	MasterComputableTimer.preIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	LocalMasterCoordinator.preIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	BasicMasterInterceptor.preIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
void	AsyncMasterCoordinator.preIteration(MasterContext<MASTER_RESULT,WORKER_RESULT> context)
protected void	AbstractMasterCoordinator.setWorkerResults(MasterContext<MASTER_RESULT,WORKER_RESULT> context, String appCurrentWorkersNode, String appId, int iteration) Set worker results from znodes.
protected void	AbstractMasterCoordinator.updateMasterHaltStatus(MasterContext<MASTER_RESULT,WORKER_RESULT> context) Check whether GuaguaConstants.GUAGUA_WORKER_HALT_ENABLE) is enabled, if yes, check whether all workers are halted and update master status.

参数类型为 MasterContext 的 ml.shifu.guagua.master 中的构造方法

AbstractMasterCoordinator.FailOverCommand(MasterContext<MASTER_RESULT,WORKER_RESULT> context)