Yes. you can. Here is code snippet from our project.
final JobConf conf = new JobConf( c_JOB_SETTING.getCustomizeConf() );
conf.setJar( c_JOB_SETTING.getJarName() );
conf.setJobName( "MortgageValTestinput=" + c_JOB_SETTING
.getInputPath() + " output=" + c_JOB_SETTING.getOutputPath() );
conf.setOutputKeyClass( Text.class );
conf.setMapOutputKeyClass( Text.class );
conf.setOutputValueClass( AttributeValueMap.class );
conf.setMapperClass( MyMapper.class );
conf.setCombinerClass( MyReducer.class );
conf.setReducerClass( MyReducer.class );
conf.setInputFormat(MyInputFormat.class );
conf.setOutputFormat( SequenceFileOutputFormat.class );
conf.setStrings( "io.serializations",
"org.apache.hadoop.io.serializer.WritableSerialization",
"org.apache.hadoop.io.serializer.JavaSerialization" );
conf.set( "mapred.create.symlink", "yes" );
JobClient.runJob( conf )
Zhu, Guojun
Modeling Sr Graduate
571-3824370
guojun_...@freddiemac.com
Financial Engineering
Freddie Mac
Félix López <jaaaelpum...@gmail.com>
06/27/2012 12:54 PM
Please respond to
mapreduce-user@hadoop.apache.org
To
mapreduce-user@hadoop.apache.org
cc
Subject
Send jobs remotely
How can I send jobs remotely? I have a cluster running and I would like to
execute a mapreduce task from another machine (outside the cluster) and
without to have to do this : bin/hadoop jar hadoop.jar main input output
every time.
I don't know if this is possible in hadoop or maybe I have to program a
service in the master that runs the job when it's called.
Maybe a good solution would be to run the job when a file is written in
the hdfs, I don't if there are hooks in hadoop
Thanks
