compbio.engine.cluster.drmaa

Class AsyncClusterRunner

java.lang.Object

compbio.engine.cluster.drmaa.AsyncClusterRunner

All Implemented Interfaces:

AsyncExecutor

public class AsyncClusterRunner
extends java.lang.Object
implements AsyncExecutor

Single cluster job runner class

Version:

1.0 August 2009 TODO after call to submitJob() no setters really work as the job template gets deleted, this needs to be taken into account in this class design!

Author:

pvtroshin

Field Summary

Fields

Modifier and Type	Field and Description
private static org.apache.log4j.Logger	log

Constructor Summary

Constructors

Constructor and Description
AsyncClusterRunner()

Method Summary

Modifier and Type	Method and Description
boolean	cancelJob(java.lang.String jobId) Stop running job.
boolean	cleanup(java.lang.String jobId) Remove all files and a job directory for a jobid.
JobStatus	getJobStatus(java.lang.String jobId) Query the status of the job
ConfiguredExecutable<?>	getResults(java.lang.String jobId) Retrieve the results of the job.
java.lang.String	getWorkDirectory(java.lang.String jobId)
java.lang.String	submitJob(ConfiguredExecutable<?> executable) Submits job for the execution Immediate execution is not guaranteed, this method puts the job in the queue.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

log

private static org.apache.log4j.Logger log

Constructor Detail

AsyncClusterRunner

public AsyncClusterRunner()

Method Detail

submitJob

public java.lang.String submitJob(ConfiguredExecutable<?> executable)
                           throws JobSubmissionException

Description copied from interface: AsyncExecutor

Submits job for the execution Immediate execution is not guaranteed, this method puts the job in the queue. All it guarantees that the job will be eventually executed. The start of execution will depend on the number of jobs in the queue.

Specified by:

submitJob in interface AsyncExecutor

Returns:

unique job identifier

Throws:

JobSubmissionException - if submission fails. This usually happens due to the problem on a server side.

cancelJob

public boolean cancelJob(java.lang.String jobId)

Description copied from interface: AsyncExecutor

Stop running job. Please not that this method does not guarantee to remove the job directory and files in it.

Specified by:

cancelJob in interface AsyncExecutor

Returns:

true if job was cancelled successfully, false otherwise

getJobStatus

public JobStatus getJobStatus(java.lang.String jobId)

Description copied from interface: AsyncExecutor

Query the status of the job

Specified by:

getJobStatus in interface AsyncExecutor

Returns:

The JobStatus object representing the status of the job

See Also:

JobStatus

cleanup

public boolean cleanup(java.lang.String jobId)

Description copied from interface: AsyncExecutor

Remove all files and a job directory for a jobid.

Specified by:

cleanup in interface AsyncExecutor

Returns:

true if job directory was successfully removed, false otherwise.

getResults

public ConfiguredExecutable<?> getResults(java.lang.String jobId)
                                   throws ResultNotAvailableException

Description copied from interface: AsyncExecutor

Retrieve the results of the job. Please not that current implementations of this method blocks if the task is running until the end of the calculation.

Specified by:

getResults in interface AsyncExecutor

Parameters:

jobId - job identifier obtained at the job submission

Returns:

ConfiguredExecutable object from which result can be obtained

Throws:

ResultNotAvailableException - if the result is not available for whatever reason. Could be due to execution failure, or due to the results being removed from the server at the time of request.

getWorkDirectory

public java.lang.String getWorkDirectory(java.lang.String jobId)

Specified by:

getWorkDirectory in interface AsyncExecutor

Parameters:

jobId - unique job identifier

Returns:

task working directory