public int run(String[] args) throws Exception {
Path inputPath = new Path("weblog_entries.txt");
Path outputPath = new Path("output");
Schema schema = ReflectData.get().getSchema(WeblogRecord.class);
Configuration conf = getConf();
Job weblogJob = Job.getInstance(conf);
weblogJob.setJobName("Avro Writer");
weblogJob.setJarByClass(getClass());
weblogJob.setNumReduceTasks(0);
weblogJob.setMapperClass(WeblogMapper_Ex_5.class);
weblogJob.setMapOutputKeyClass(AvroWrapper.class);
weblogJob.setMapOutputValueClass(NullWritable.class);
weblogJob.setInputFormatClass(TextInputFormat.class);
AvroJob.setOutputKeySchema(weblogJob, schema);
FileInputFormat.setInputPaths(weblogJob, inputPath);
FileOutputFormat.setOutputPath(weblogJob, outputPath);
if (weblogJob.waitForCompletion(true)) {
return 0;
}
return 1;
}
public int run(String[] args) throws Exception {
Configuration conf = new Configuration();
if (args.length != 2) {
System.err.printf(
"Usage: %s <comma separated paths> <output path>\n", this.getClass().getName());
return -1;
}
Job job = Job.getInstance();
job.setJobName("PasmJoin");
job.setJarByClass(PsamXY.class);
job.setMapOutputKeyClass(Text.class);
job.setMapOutputValueClass(AvroValue.class);
job.setOutputKeyClass(AvroKey.class);
job.setOutputValueClass(NullWritable.class);
job.setMapperClass(PsamXYMapper.class);
job.setReducerClass(PsamXYReducer.class);
job.setInputFormatClass(AvroKeyInputFormat.class);
job.setOutputFormatClass(AvroKeyOutputFormat.class);
FileInputFormat.setInputPaths(job, args[0]);
Path output = new Path(args[1]);
FileOutputFormat.setOutputPath(job, output);
FileSystem fs = FileSystem.get(conf);
fs.delete(output, true);
AvroJob.setOutputKeySchema(job, outputSchema);
AvroJob.setMapOutputValueSchema(job, outputSchema);
// DistributedCache.addCacheFile(new Path("BM_TERM_TYPE_DMT.avro").toUri(),
// job.getConfiguration());
job.setNumReduceTasks(1);
job.submit();
job.waitForCompletion(true);
return 0;
}