public static void main(String[] args) throws Exception {
final String NAME_NODE = "hdfs://sandbox.hortonworks.com:8020";
Configuration conf = new Configuration();
Job job = Job.getInstance(conf);
job.setJarByClass(WordCount.class);
job.setMapOutputKeyClass(Text.class);
job.setMapOutputValueClass(IntWritable.class);
job.setOutputKeyClass(IntWritable.class);
job.setOutputValueClass(NullWritable.class);
if (args.length > 2) {
job.setNumReduceTasks(Integer.parseInt(args[2]));
}
job.setMapperClass(CountMapper.class);
job.setReducerClass(CountReducer.class);
job.setJarByClass(WordCount.class);
job.setNumReduceTasks(1);
FileInputFormat.addInputPath(job, new Path(args[0] + "data/plot_summaries.txt"));
FileSystem fs = FileSystem.get(conf);
// handle (e.g. delete) existing output path
Path outputDestination = new Path(args[0] + args[1]);
if (fs.exists(outputDestination)) {
fs.delete(outputDestination, true);
}
// set output path & start job1
FileOutputFormat.setOutputPath(job, outputDestination);
int jobCompletionStatus = job.waitForCompletion(true) ? 0 : 1;
}
private static Path checkDest(String srcName, FileSystem dstFS, Path dst, boolean overwrite)
throws IOException {
if (dstFS.exists(dst)) {
FileStatus sdst = dstFS.getFileStatus(dst);
if (sdst.isDir()) {
if (null == srcName) {
throw new IOException("Target " + dst + " is a directory");
}
return checkDest(null, dstFS, new Path(dst, srcName), overwrite);
} else if (!overwrite) {
throw new IOException("Target " + dst + " already exists");
}
}
return dst;
}
/**
* The src file is under FS, and the dst is on the local disk. Copy it from FS control to the
* local dst name. If src and dst are directories, the copyCrc parameter determines whether to
* copy CRC files.
*/
public void copyToLocalFile(Path src, Path dst, boolean copyCrc) throws IOException {
if (!fs.isDirectory(src)) { // source is a file
fs.copyToLocalFile(src, dst);
FileSystem localFs = getLocal(getConf()).getRawFileSystem();
if (localFs.isDirectory(dst)) {
dst = new Path(dst, src.getName());
}
dst = getChecksumFile(dst);
if (localFs.exists(dst)) { // remove old local checksum file
localFs.delete(dst, true);
}
Path checksumFile = getChecksumFile(src);
if (copyCrc && fs.exists(checksumFile)) { // copy checksum file
fs.copyToLocalFile(checksumFile, dst);
}
} else {
FileStatus[] srcs = listStatus(src);
for (FileStatus srcFile : srcs) {
copyToLocalFile(srcFile.getPath(), new Path(dst, srcFile.getPath().getName()), copyCrc);
}
}
}
public static boolean copy(
FileSystem srcFS,
Path[] srcs,
FileSystem dstFS,
Path dst,
boolean deleteSource,
boolean overwrite,
Configuration conf)
throws IOException {
boolean gotException = false;
boolean returnVal = true;
StringBuffer exceptions = new StringBuffer();
if (srcs.length == 1) return copy(srcFS, srcs[0], dstFS, dst, deleteSource, overwrite, conf);
// Check if dest is directory
if (!dstFS.exists(dst)) {
throw new IOException("`" + dst + "': specified destination directory " + "doest not exist");
} else {
FileStatus sdst = dstFS.getFileStatus(dst);
if (!sdst.isDir())
throw new IOException(
"copying multiple files, but last argument `" + dst + "' is not a directory");
}
for (Path src : srcs) {
try {
if (!copy(srcFS, src, dstFS, dst, deleteSource, overwrite, conf)) returnVal = false;
} catch (IOException e) {
gotException = true;
exceptions.append(e.getMessage());
exceptions.append("\n");
}
}
if (gotException) {
throw new IOException(exceptions.toString());
}
return returnVal;
}
static void checkRecords(
Configuration defaults, int noMaps, int noReduces, Path sortInput, Path sortOutput)
throws IOException {
JobConf jobConf = new JobConf(defaults, RecordChecker.class);
jobConf.setJobName("sortvalidate-record-checker");
jobConf.setInputFormat(SequenceFileInputFormat.class);
jobConf.setOutputFormat(SequenceFileOutputFormat.class);
jobConf.setOutputKeyClass(BytesWritable.class);
jobConf.setOutputValueClass(IntWritable.class);
jobConf.setMapperClass(Map.class);
jobConf.setReducerClass(Reduce.class);
JobClient client = new JobClient(jobConf);
ClusterStatus cluster = client.getClusterStatus();
if (noMaps == -1) {
noMaps = cluster.getTaskTrackers() * jobConf.getInt("test.sortvalidate.maps_per_host", 10);
}
if (noReduces == -1) {
noReduces = (int) (cluster.getMaxReduceTasks() * 0.9);
String sortReduces = jobConf.get("test.sortvalidate.reduces_per_host");
if (sortReduces != null) {
noReduces = cluster.getTaskTrackers() * Integer.parseInt(sortReduces);
}
}
jobConf.setNumMapTasks(noMaps);
jobConf.setNumReduceTasks(noReduces);
FileInputFormat.setInputPaths(jobConf, sortInput);
FileInputFormat.addInputPath(jobConf, sortOutput);
Path outputPath = new Path("/tmp/sortvalidate/recordchecker");
FileSystem fs = FileSystem.get(defaults);
if (fs.exists(outputPath)) {
fs.delete(outputPath, true);
}
FileOutputFormat.setOutputPath(jobConf, outputPath);
// Uncomment to run locally in a single process
// job_conf.set("mapred.job.tracker", "local");
Path[] inputPaths = FileInputFormat.getInputPaths(jobConf);
System.out.println(
"\nSortValidator.RecordChecker: Running on "
+ cluster.getTaskTrackers()
+ " nodes to validate sort from "
+ inputPaths[0]
+ ", "
+ inputPaths[1]
+ " into "
+ FileOutputFormat.getOutputPath(jobConf)
+ " with "
+ noReduces
+ " reduces.");
Date startTime = new Date();
System.out.println("Job started: " + startTime);
JobClient.runJob(jobConf);
Date end_time = new Date();
System.out.println("Job ended: " + end_time);
System.out.println(
"The job took " + (end_time.getTime() - startTime.getTime()) / 1000 + " seconds.");
}
static void checkRecords(Configuration defaults, Path sortInput, Path sortOutput)
throws IOException {
FileSystem inputfs = sortInput.getFileSystem(defaults);
FileSystem outputfs = sortOutput.getFileSystem(defaults);
FileSystem defaultfs = FileSystem.get(defaults);
JobConf jobConf = new JobConf(defaults, RecordStatsChecker.class);
jobConf.setJobName("sortvalidate-recordstats-checker");
int noSortReduceTasks = outputfs.listStatus(sortOutput, sortPathsFilter).length;
jobConf.setInt("sortvalidate.sort.reduce.tasks", noSortReduceTasks);
int noSortInputpaths = inputfs.listStatus(sortInput).length;
jobConf.setInputFormat(NonSplitableSequenceFileInputFormat.class);
jobConf.setOutputFormat(SequenceFileOutputFormat.class);
jobConf.setOutputKeyClass(IntWritable.class);
jobConf.setOutputValueClass(RecordStatsChecker.RecordStatsWritable.class);
jobConf.setMapperClass(Map.class);
jobConf.setCombinerClass(Reduce.class);
jobConf.setReducerClass(Reduce.class);
jobConf.setNumMapTasks(noSortReduceTasks);
jobConf.setNumReduceTasks(1);
FileInputFormat.setInputPaths(jobConf, sortInput);
FileInputFormat.addInputPath(jobConf, sortOutput);
Path outputPath = new Path("/tmp/sortvalidate/recordstatschecker");
if (defaultfs.exists(outputPath)) {
defaultfs.delete(outputPath, true);
}
FileOutputFormat.setOutputPath(jobConf, outputPath);
// Uncomment to run locally in a single process
// job_conf.set("mapred.job.tracker", "local");
Path[] inputPaths = FileInputFormat.getInputPaths(jobConf);
System.out.println(
"\nSortValidator.RecordStatsChecker: Validate sort "
+ "from "
+ inputPaths[0]
+ " ("
+ noSortInputpaths
+ " files), "
+ inputPaths[1]
+ " ("
+ noSortReduceTasks
+ " files) into "
+ FileOutputFormat.getOutputPath(jobConf)
+ " with 1 reducer.");
Date startTime = new Date();
System.out.println("Job started: " + startTime);
JobClient.runJob(jobConf);
Date end_time = new Date();
System.out.println("Job ended: " + end_time);
System.out.println(
"The job took " + (end_time.getTime() - startTime.getTime()) / 1000 + " seconds.");
// Check to ensure that the statistics of the
// framework's sort-input and sort-output match
SequenceFile.Reader stats =
new SequenceFile.Reader(defaultfs, new Path(outputPath, "part-00000"), defaults);
IntWritable k1 = new IntWritable();
IntWritable k2 = new IntWritable();
RecordStatsWritable v1 = new RecordStatsWritable();
RecordStatsWritable v2 = new RecordStatsWritable();
if (!stats.next(k1, v1)) {
throw new IOException("Failed to read record #1 from reduce's output");
}
if (!stats.next(k2, v2)) {
throw new IOException("Failed to read record #2 from reduce's output");
}
if ((v1.getBytes() != v2.getBytes())
|| (v1.getRecords() != v2.getRecords())
|| v1.getChecksum() != v2.getChecksum()) {
throw new IOException(
"("
+ v1.getBytes()
+ ", "
+ v1.getRecords()
+ ", "
+ v1.getChecksum()
+ ") v/s ("
+ v2.getBytes()
+ ", "
+ v2.getRecords()
+ ", "
+ v2.getChecksum()
+ ")");
}
}