org.apache.flink.runtime.highavailability

Interface RunningJobsRegistry

All Known Implementing Classes:

StandaloneRunningJobsRegistry, ZooKeeperRunningJobsRegistry

public interface RunningJobsRegistry

A simple registry that tracks if a certain job is pending execution, running, or completed.

This registry is used in highly-available setups with multiple master nodes, to determine whether a new leader should attempt to recover a certain job (because the job is still running), or whether the job has already finished successfully (in case of a finite job) and the leader has only been granted leadership because the previous leader quit cleanly after the job was finished.

In addition, the registry can help to determine whether a newly assigned leader JobManager should attempt reconciliation with running TaskManagers, or immediately schedule the job from the latest checkpoint/savepoint.

Nested Class Summary

Nested Classes

Modifier and Type	Interface and Description
static class	RunningJobsRegistry.JobSchedulingStatus The scheduling status of a job, as maintained by the RunningJobsRegistry.

Method Summary

Modifier and Type	Method and Description
void	clearJob(org.apache.flink.api.common.JobID jobID) Clear job state form the registry, usually called after job finish.
RunningJobsRegistry.JobSchedulingStatus	getJobSchedulingStatus(org.apache.flink.api.common.JobID jobID) Gets the scheduling status of a job.
void	setJobFinished(org.apache.flink.api.common.JobID jobID) Marks a job as completed.
void	setJobRunning(org.apache.flink.api.common.JobID jobID) Marks a job as running.

Method Detail

setJobRunning

void setJobRunning(org.apache.flink.api.common.JobID jobID)
            throws IOException

Marks a job as running. Requesting the job's status via the getJobSchedulingStatus(JobID) method will return RunningJobsRegistry.JobSchedulingStatus.RUNNING.

Parameters:

jobID - The id of the job.

Throws:

IOException - Thrown when the communication with the highly-available storage or registry failed and could not be retried.

setJobFinished

void setJobFinished(org.apache.flink.api.common.JobID jobID)
             throws IOException

Marks a job as completed. Requesting the job's status via the getJobSchedulingStatus(JobID) method will return RunningJobsRegistry.JobSchedulingStatus.DONE.

Parameters:

jobID - The id of the job.

Throws:

IOException - Thrown when the communication with the highly-available storage or registry failed and could not be retried.

getJobSchedulingStatus

RunningJobsRegistry.JobSchedulingStatus getJobSchedulingStatus(org.apache.flink.api.common.JobID jobID)
                                                        throws IOException

Gets the scheduling status of a job.

Parameters:

jobID - The id of the job to check.

Returns:

The job scheduling status.

Throws:

IOException - Thrown when the communication with the highly-available storage or registry failed and could not be retried.

clearJob

void clearJob(org.apache.flink.api.common.JobID jobID)
       throws IOException

Clear job state form the registry, usually called after job finish.

Parameters:

jobID - The id of the job to check.

Throws:

IOException - Thrown when the communication with the highly-available storage or registry failed and could not be retried.