public static void runSortJob(String... args) throws Exception {
Path input = new Path(args[0]);
Path output = new Path(args[1]);
JobConf job = new JobConf();
job.setNumReduceTasks(2);
job.setInputFormat(KeyValueTextInputFormat.class);
job.setOutputFormat(TextOutputFormat.class);
job.setMapOutputKeyClass(Text.class);
job.setMapOutputValueClass(Text.class);
FileInputFormat.setInputPaths(job, input);
FileOutputFormat.setOutputPath(job, output);
job.setJarByClass(SampleJob.class);
output.getFileSystem(job).delete(output, true);
JobClient jc = new JobClient(job);
JobClient.setTaskOutputFilter(job, JobClient.TaskStatusFilter.ALL);
RunningJob rj = jc.submitJob(job);
try {
if (!jc.monitorAndPrintJob(job, rj)) {
System.out.println("Job Failed: " + rj.getFailureInfo());
throw new IOException("Job failed!");
}
} catch (InterruptedException ie) {
Thread.currentThread().interrupt();
}
}
@Override
public int run(String[] args) throws Exception {
Configuration conf = getConf();
FileSystem fs = FileSystem.get(conf);
JobConf job = new JobConf(conf);
job.setJarByClass(Jacobi.class);
fs.delete(new Path("curX"), true);
job.setInputFormat(TextInputFormat.class);
job.setOutputFormat(TextOutputFormat.class);
job.setMapperClass(Map.class);
job.setReducerClass(Reduce.class);
job.setOutputKeyClass(IntWritable.class);
job.setOutputValueClass(DoubleWritable.class);
FileInputFormat.setInputPaths(job, new Path("preX"));
FileOutputFormat.setOutputPath(job, new Path("curX"));
JobClient.runJob(job);
return 1;
}
public int run(String[] argv) throws Exception {
JobConf job = new JobConf(getConf());
job.setJarByClass(GenericMRLoadGenerator.class);
job.setMapperClass(SampleMapper.class);
job.setReducerClass(SampleReducer.class);
if (!parseArgs(argv, job)) {
return -1;
}
if (null == FileOutputFormat.getOutputPath(job)) {
// No output dir? No writes
job.setOutputFormat(NullOutputFormat.class);
}
if (0 == FileInputFormat.getInputPaths(job).length) {
// No input dir? Generate random data
System.err.println("No input path; ignoring InputFormat");
confRandom(job);
} else if (null
!= job.getClass(
org.apache.hadoop.mapreduce.GenericMRLoadGenerator.INDIRECT_INPUT_FORMAT, null)) {
// specified IndirectInputFormat? Build src list
JobClient jClient = new JobClient(job);
Path tmpDir = new Path(jClient.getFs().getHomeDirectory(), ".staging");
Random r = new Random();
Path indirInputFile =
new Path(tmpDir, Integer.toString(r.nextInt(Integer.MAX_VALUE), 36) + "_files");
job.set(
org.apache.hadoop.mapreduce.GenericMRLoadGenerator.INDIRECT_INPUT_FILE,
indirInputFile.toString());
SequenceFile.Writer writer =
SequenceFile.createWriter(
tmpDir.getFileSystem(job),
job,
indirInputFile,
LongWritable.class,
Text.class,
SequenceFile.CompressionType.NONE);
try {
for (Path p : FileInputFormat.getInputPaths(job)) {
FileSystem fs = p.getFileSystem(job);
Stack<Path> pathstack = new Stack<Path>();
pathstack.push(p);
while (!pathstack.empty()) {
for (FileStatus stat : fs.listStatus(pathstack.pop())) {
if (stat.isDirectory()) {
if (!stat.getPath().getName().startsWith("_")) {
pathstack.push(stat.getPath());
}
} else {
writer.sync();
writer.append(
new LongWritable(stat.getLen()), new Text(stat.getPath().toUri().toString()));
}
}
}
}
} finally {
writer.close();
}
}
Date startTime = new Date();
System.out.println("Job started: " + startTime);
JobClient.runJob(job);
Date endTime = new Date();
System.out.println("Job ended: " + endTime);
System.out.println(
"The job took " + (endTime.getTime() - startTime.getTime()) / 1000 + " seconds.");
return 0;
}