Package | Description |
---|---|
org.apache.hadoop.mapreduce.v2.app.job.impl |
Modifier and Type | Field and Description |
---|---|
protected static org.apache.hadoop.yarn.state.StateMachineFactory<JobImpl,JobStateInternal,JobEventType,JobEvent> |
JobImpl.stateMachineFactory |
Modifier and Type | Method and Description |
---|---|
protected org.apache.hadoop.mapreduce.split.JobSplit.TaskSplitMetaInfo[] |
JobImpl.InitTransition.createSplits(JobImpl job,
org.apache.hadoop.mapreduce.v2.api.records.JobId jobId) |
protected void |
JobImpl.InitTransition.setup(JobImpl job) |
JobStateInternal |
JobImpl.InitTransition.transition(JobImpl job,
JobEvent event)
Note that this transition method is called directly (and synchronously)
by MRAppMaster's init() method (i.e., no RPC, no thread-switching;
just plain sequential call within AM context), so we can trigger
modifications in AM state from here (at least, if AM is written that
way; MR version is).
|
void |
JobImpl.StartTransition.transition(JobImpl job,
JobEvent event)
This transition executes in the event-dispatcher thread, though it's
triggered in MRAppMaster's startJobs() method.
|
Copyright © 2008–2024 Apache Software Foundation. All rights reserved.