public static boolean checkJobContextIfRunningFromBackend(JobContext j) {
if (j.getConfiguration().get("mapred.task.id", "").equals("")
&& !("true".equals(j.getConfiguration().get("pig.illustrating")))) {
return false;
}
return true;
}
public static Map<FormatBundle, Map<Integer, List<Path>>> getFormatNodeMap(JobContext job) {
Map<FormatBundle, Map<Integer, List<Path>>> formatNodeMap = Maps.newHashMap();
Configuration conf = job.getConfiguration();
String crunchInputs = conf.get(CRUNCH_INPUTS);
if (crunchInputs == null || crunchInputs.isEmpty()) {
return ImmutableMap.of();
}
for (String input : Splitter.on(RECORD_SEP).split(crunchInputs)) {
List<String> fields = Lists.newArrayList(SPLITTER.split(input));
FormatBundle<InputFormat> inputBundle =
FormatBundle.fromSerialized(fields.get(0), job.getConfiguration());
if (!formatNodeMap.containsKey(inputBundle)) {
formatNodeMap.put(inputBundle, Maps.<Integer, List<Path>>newHashMap());
}
Integer nodeIndex = Integer.valueOf(fields.get(1));
if (!formatNodeMap.get(inputBundle).containsKey(nodeIndex)) {
formatNodeMap.get(inputBundle).put(nodeIndex, Lists.<Path>newLinkedList());
}
List<Path> formatNodePaths = formatNodeMap.get(inputBundle).get(nodeIndex);
String paths = fields.get(2);
for (String path : Splitter.on(PATH_SEP).split(paths)) {
formatNodePaths.add(new Path(path));
}
}
return formatNodeMap;
}
public SqoopOutputFormatLoadExecutorSpark(JobContext jobctx) {
context = jobctx;
loaderName = context.getConfiguration().get(MRJobConstants.JOB_ETL_LOADER);
writer = new SqoopRecordWriter();
// jackh: This must be conditional - Extract schema using credentials in case of MR and simply
// extract from the
// credentials object in case of Spark (due to known issue with Hadoop/Spark that the
// credentials are never added
// for serialization)
// matcher = MatcherFactory.getMatcher(
// MRConfigurationUtils.getConnectorSchema(Direction.FROM, context.getConfiguration()),
// MRConfigurationUtils.getConnectorSchema(Direction.TO, context.getConfiguration()));
matcher =
MatcherFactory.getMatcher(
MRConfigurationUtils.getConnectorSchemaUnsafe(
Direction.FROM, context.getConfiguration()),
MRConfigurationUtils.getConnectorSchemaUnsafe(
Direction.TO, context.getConfiguration()));
toDataFormat =
(IntermediateDataFormat<?>)
ClassUtils.instantiate(
context.getConfiguration().get(MRJobConstants.TO_INTERMEDIATE_DATA_FORMAT));
// Using the TO schema since the SqoopDataWriter in the SqoopMapper encapsulates the
// toDataFormat
toDataFormat.setSchema(matcher.getToSchema());
}
public static void checkOutputSpecs(JobContext jc) throws IOException, InterruptedException {
Map<String, OutputConfig> outputs = getNamedOutputs(jc.getConfiguration());
for (Map.Entry<String, OutputConfig> e : outputs.entrySet()) {
String namedOutput = e.getKey();
Job job = getJob(jc.getJobID(), e.getKey(), jc.getConfiguration());
OutputFormat fmt = getOutputFormat(namedOutput, job, e.getValue());
fmt.checkOutputSpecs(job);
}
}
@Test
public void testDeleteMissing() {
TaskAttemptContext taskAttemptContext = getTaskAttemptContext(config);
JobContext jobContext =
new JobContextImpl(
taskAttemptContext.getConfiguration(),
taskAttemptContext.getTaskAttemptID().getJobID());
Configuration conf = jobContext.getConfiguration();
String sourceBase;
String targetBase;
FileSystem fs = null;
try {
OutputCommitter committer = new CopyCommitter(null, taskAttemptContext);
fs = FileSystem.get(conf);
sourceBase = TestDistCpUtils.createTestSetup(fs, FsPermission.getDefault());
targetBase = TestDistCpUtils.createTestSetup(fs, FsPermission.getDefault());
String targetBaseAdd = TestDistCpUtils.createTestSetup(fs, FsPermission.getDefault());
fs.rename(new Path(targetBaseAdd), new Path(targetBase));
DistCpOptions options =
new DistCpOptions(Arrays.asList(new Path(sourceBase)), new Path("/out"));
options.setSyncFolder(true);
options.setDeleteMissing(true);
options.appendToConf(conf);
CopyListing listing = new GlobbedCopyListing(conf, CREDENTIALS);
Path listingFile = new Path("/tmp1/" + String.valueOf(rand.nextLong()));
listing.buildListing(listingFile, options);
conf.set(DistCpConstants.CONF_LABEL_TARGET_WORK_PATH, targetBase);
conf.set(DistCpConstants.CONF_LABEL_TARGET_FINAL_PATH, targetBase);
committer.commitJob(jobContext);
if (!TestDistCpUtils.checkIfFoldersAreInSync(fs, targetBase, sourceBase)) {
Assert.fail("Source and target folders are not in sync");
}
if (!TestDistCpUtils.checkIfFoldersAreInSync(fs, sourceBase, targetBase)) {
Assert.fail("Source and target folders are not in sync");
}
// Test for idempotent commit
committer.commitJob(jobContext);
if (!TestDistCpUtils.checkIfFoldersAreInSync(fs, targetBase, sourceBase)) {
Assert.fail("Source and target folders are not in sync");
}
if (!TestDistCpUtils.checkIfFoldersAreInSync(fs, sourceBase, targetBase)) {
Assert.fail("Source and target folders are not in sync");
}
} catch (Throwable e) {
LOG.error("Exception encountered while testing for delete missing", e);
Assert.fail("Delete missing failure");
} finally {
TestDistCpUtils.delete(fs, "/tmp1");
conf.set(DistCpConstants.CONF_LABEL_DELETE_MISSING, "false");
}
}
@Override
public void abortJob(JobContext jobContext, JobStatus.State state) throws IOException {
Configuration conf = jobContext.getConfiguration();
for (Map.Entry<String, OutputCommitter> e : committers.entrySet()) {
Job job = getJob(jobContext.getJobID(), e.getKey(), conf);
configureJob(e.getKey(), job, outputs.get(e.getKey()));
e.getValue().abortJob(job, state);
}
}
/**
* Generate the list of files and make them into FileSplits. This needs to be copied to insert a
* filter on acceptable data
*/
@Override
public List<InputSplit> getSplits(JobContext job) throws IOException {
long minSize = Math.max(getFormatMinSplitSize(), getMinSplitSize(job));
long maxSize = getMaxSplitSize(job);
long desiredMappers =
job.getConfiguration().getLong("org.systemsbiology.jxtandem.DesiredXMLInputMappers", 0);
// generate splits
List<InputSplit> splits = new ArrayList<InputSplit>();
List<FileStatus> fileStatuses = listStatus(job);
boolean forceNumberMappers = fileStatuses.size() == 1;
for (FileStatus file : fileStatuses) {
Path path = file.getPath();
if (!isPathAcceptable(path)) // filter acceptable data
continue;
FileSystem fs = path.getFileSystem(job.getConfiguration());
long length = file.getLen();
BlockLocation[] blkLocations = fs.getFileBlockLocations(file, 0, length);
if ((length != 0) && isSplitable(job, path)) {
long blockSize = file.getBlockSize();
// use desired mappers to force more splits
if (forceNumberMappers && desiredMappers > 0)
maxSize = Math.min(maxSize, (length / desiredMappers));
long splitSize = computeSplitSize(blockSize, minSize, maxSize);
long bytesRemaining = length;
while (withinSlop(splitSize, bytesRemaining)) {
int blkIndex = getBlockIndex(blkLocations, length - bytesRemaining);
splits.add(
new FileSplit(
path, length - bytesRemaining, splitSize, blkLocations[blkIndex].getHosts()));
bytesRemaining -= splitSize;
}
if (bytesRemaining != 0) {
splits.add(
new FileSplit(
path,
length - bytesRemaining,
bytesRemaining,
blkLocations[blkLocations.length - 1].getHosts()));
}
} else if (length != 0) {
splits.add(new FileSplit(path, 0, length, blkLocations[0].getHosts()));
} else {
// Create empty hosts array for zero length files
splits.add(new FileSplit(path, 0, length, new String[0]));
}
}
System.out.println("Total # of splits: " + splits.size());
// LOG.debug("Total # of splits: " + splits.size());
return splits;
}
/**
* Validates that a valid FIXED_RECORD_LENGTH config property has been set and if so, returns the
* splits. If the FIXED_RECORD_LENGTH property has not been set, this will throw an IOException.
*
* @inheritDoc
*/
@Override
public List<InputSplit> getSplits(JobContext job) throws IOException {
// fetch configuration
Configuration conf = job.getConfiguration();
// ensure recordLength is properly setup
try {
if (this.recordLength == -1) {
this.recordLength = getAndValidateRecordLength(job.getConfiguration());
}
LOG.info("FixedLengthInputFormat: my fixed record length is: " + recordLength);
} catch (Exception e) {
throw new IOException(
"FixedLengthInputFormat requires the"
+ " Configuration property:"
+ FIXED_RECORD_LENGTH
+ " to"
+ " be set to something > 0. Currently the value is 0 (zero)");
}
// ensure recordKey start/end is setup properly if it was defined by the user
if (this.recordKeyStartAt == -1) {
this.recordKeyStartAt = FixedLengthInputFormat.getRecordKeyStartAt(conf);
this.recordKeyEndAt = FixedLengthInputFormat.getRecordKeyEndAt(conf);
// if one is set, they BOTH must be set, this is an error
// if endAt < startAt, this is an error
// if either is > record length, this is an error
// if either are < -1 (default), this is an error
if ((recordKeyStartAt >= 0 && recordKeyEndAt == -1)
|| (recordKeyStartAt == -1 && recordKeyEndAt >= 0)
|| (recordKeyEndAt < recordKeyStartAt)
|| (recordKeyEndAt > recordLength)
|| (recordKeyStartAt > recordLength)
|| (recordKeyStartAt < -1)
|| (recordKeyEndAt < -1)) {
throw new IOException(
"FixedLengthInputFormat requires the"
+ " optional configuration properties:"
+ FIXED_RECORD_KEY_START_AT
+ " and"
+ FIXED_RECORD_KEY_END_AT
+ " to A) be less than the "
+ " fixed record length. B) both must be set together C) neither "
+ " can be less than 0. D) end at must be > start at.");
}
}
return super.getSplits(job);
}
private void setSplitSize(JobContext cx) {
super.setMaxSplitSize(
cx.getConfiguration()
.getLong(
COMPACTION_JOB_MAPRED_MAX_SPLIT_SIZE,
DEFAULT_COMPACTION_JOB_MAPRED_MAX_SPLIT_SIZE));
super.setMinSplitSizeNode(
cx.getConfiguration()
.getLong(
COMPACTION_JOB_MAPRED_MIN_SPLIT_SIZE,
DEFAULT_COMPACTION_JOB_MAPRED_MIN_SPLIT_SIZE));
}
public List<InputSplit> getSplits(JobContext job) throws IOException {
Configuration conf = job.getConfiguration();
int numMapTasks = conf.getInt("admm.iteration.num.map.tasks", 0);
if (0 == numMapTasks) {
return super.getSplits(job);
}
// generate splits
List<InputSplit> splits = new ArrayList<InputSplit>();
List<FileStatus> files = listStatus(job);
for (FileStatus file : files) {
Path path = file.getPath();
FileSystem fs = path.getFileSystem(job.getConfiguration());
long length = file.getLen();
BlockLocation[] blkLocations = fs.getFileBlockLocations(file, 0, length);
if ((length != 0) && isSplitable(job, path)) {
long blockSize = file.getBlockSize();
long splitSize = Math.max(computeSplitSize(JAVA_OPTS, numMapTasks, length), blockSize);
long splitLength = (long) (length / Math.ceil((double) length / splitSize));
long bytesRemaining = length;
while (((double) bytesRemaining) / splitLength > SPLIT_SLOP) {
int blkIndex = getBlockIndex(blkLocations, length - bytesRemaining);
splits.add(
new FileSplit(
path, length - bytesRemaining, splitLength, blkLocations[blkIndex].getHosts()));
bytesRemaining -= splitLength;
}
if (bytesRemaining != 0) {
splits.add(
new FileSplit(
path,
length - bytesRemaining,
bytesRemaining,
blkLocations[blkLocations.length - 1].getHosts()));
}
} else if (length != 0) {
splits.add(new FileSplit(path, 0, length, blkLocations[0].getHosts()));
} else {
splits.add(new FileSplit(path, 0, length, new String[0]));
}
}
// Save the number of input files in the job-conf
job.getConfiguration().setLong(NUM_INPUT_FILES, files.size());
job.getConfiguration().setInt("admm.iteration.num.map.tasks", splits.size());
return splits;
}
/**
* List input directories. Subclasses may override to, e.g., select only files matching a regular
* expression.
*
* @param job the job to list input paths for
* @return array of FileStatus objects
* @throws IOException if zero items.
*/
protected List<FileStatus> listStatus(JobContext job) throws IOException {
Path[] dirs = getInputPaths(job);
if (dirs.length == 0) {
throw new IOException("No input paths specified in job");
}
// get tokens for all the required FileSystems..
TokenCache.obtainTokensForNamenodes(job.getCredentials(), dirs, job.getConfiguration());
// Whether we need to recursive look into the directory structure
boolean recursive = getInputDirRecursive(job);
// creates a MultiPathFilter with the hiddenFileFilter and the
// user provided one (if any).
List<PathFilter> filters = new ArrayList<PathFilter>();
filters.add(hiddenFileFilter);
PathFilter jobFilter = getInputPathFilter(job);
if (jobFilter != null) {
filters.add(jobFilter);
}
PathFilter inputFilter = new MultiPathFilter(filters);
List<FileStatus> result = null;
int numThreads =
job.getConfiguration().getInt(LIST_STATUS_NUM_THREADS, DEFAULT_LIST_STATUS_NUM_THREADS);
Stopwatch sw = new Stopwatch().start();
if (numThreads == 1) {
result = singleThreadedListStatus(job, dirs, inputFilter, recursive);
} else {
Iterable<FileStatus> locatedFiles = null;
try {
LocatedFileStatusFetcher locatedFileStatusFetcher =
new LocatedFileStatusFetcher(
job.getConfiguration(), dirs, recursive, inputFilter, true);
locatedFiles = locatedFileStatusFetcher.getFileStatuses();
} catch (InterruptedException e) {
throw new IOException("Interrupted while getting file statuses");
}
result = Lists.newArrayList(locatedFiles);
}
sw.stop();
if (LogGlobal.isDebugEnabled()) {
/* LOG.debug("Time taken to get FileStatuses: "+sw.elapsedMillis()) */
LOG.time_taken_get_filestatuses(String.valueOf(sw.elapsedMillis())).tag("methodCall").debug();
}
/* LOG.info("Total input paths to process : "+result.size()) */
LOG.total_input_paths_process(String.valueOf(result.size())).tag("methodCall").info();
return result;
}
@Override
public List<InputSplit> getSplits(JobContext job) throws IOException {
List<InputSplit> splitList = new ArrayList<InputSplit>();
int totalRows = job.getConfiguration().getInt("LoadSplit.TOTAL_ROWS", 0);
int clients = job.getConfiguration().getInt("LoadSplit.CLIENTS", 1);
int numRows = totalRows / clients;
for (int ii = 0; ii < clients; ++ii) {
int startRow = ii * numRows;
LoadSplit split = new LoadSplit(startRow, numRows, ii);
splitList.add(split);
}
return splitList;
}
@Override
public void checkOutputSpecs(JobContext job) throws IOException {
String hosts = job.getConfiguration().get(REDIS_HOSTS_CONF);
if (hosts == null || hosts.isEmpty()) {
throw new IOException(REDIS_HOSTS_CONF + " is not set in configuration.");
}
String hashKey = job.getConfiguration().get(REDIS_HASH_KEY_CONF);
if (hashKey == null || hashKey.isEmpty()) {
throw new IOException(REDIS_HASH_KEY_CONF + " is not set in configuration.");
}
}
@Test
public void testPreserveStatus() {
TaskAttemptContext taskAttemptContext = getTaskAttemptContext(config);
JobContext jobContext =
new JobContextImpl(
taskAttemptContext.getConfiguration(),
taskAttemptContext.getTaskAttemptID().getJobID());
Configuration conf = jobContext.getConfiguration();
String sourceBase;
String targetBase;
FileSystem fs = null;
try {
OutputCommitter committer = new CopyCommitter(null, taskAttemptContext);
fs = FileSystem.get(conf);
FsPermission sourcePerm = new FsPermission((short) 511);
FsPermission initialPerm = new FsPermission((short) 448);
sourceBase = TestDistCpUtils.createTestSetup(fs, sourcePerm);
targetBase = TestDistCpUtils.createTestSetup(fs, initialPerm);
DistCpOptions options =
new DistCpOptions(Arrays.asList(new Path(sourceBase)), new Path("/out"));
options.preserve(FileAttribute.PERMISSION);
options.appendToConf(conf);
CopyListing listing = new GlobbedCopyListing(conf, CREDENTIALS);
Path listingFile = new Path("/tmp1/" + String.valueOf(rand.nextLong()));
listing.buildListing(listingFile, options);
conf.set(DistCpConstants.CONF_LABEL_TARGET_WORK_PATH, targetBase);
committer.commitJob(jobContext);
if (!checkDirectoryPermissions(fs, targetBase, sourcePerm)) {
Assert.fail("Permission don't match");
}
// Test for idempotent commit
committer.commitJob(jobContext);
if (!checkDirectoryPermissions(fs, targetBase, sourcePerm)) {
Assert.fail("Permission don't match");
}
} catch (IOException e) {
LOG.error("Exception encountered while testing for preserve status", e);
Assert.fail("Preserve status failure");
} finally {
TestDistCpUtils.delete(fs, "/tmp1");
}
}
/**
* Creates splits with multiple indexes per split (if they are smaller than
* maxCombinedIndexSizePerSplit). It is possible for a split to be larger than
* maxCombinedIndexSizePerSplit, if it consists of a single index that is larger than
* maxCombinedIndexSizePerSplit.
*
* <p>All inputPaths will be searched for indexes recursively
*
* <p>The bin-packing problem of combining splits is solved naively:
*
* <ol>
* <li>Sort all indexes by size
* <li>Begin packing indexes into splits until adding the next split would cause the split to
* exceed maxCombinedIndexSizePerSplit
* <li>Begin packing subsequent indexes into the next split, and so on
* </ol>
*/
@Override
public List<InputSplit> getSplits(JobContext job) throws IOException, InterruptedException {
// load settings from job conf
loadConfig(job.getConfiguration());
// find all the index dirs and create a split for each
PriorityQueue<LuceneIndexInputSplit> splits = findSplits(job.getConfiguration());
// combine the splits based on maxCombineSplitSize
List<InputSplit> combinedSplits =
combineSplits(splits, maxCombinedIndexSizePerSplit, maxNumIndexesPerSplit);
return combinedSplits;
}
private String[] getActiveServersList(JobContext context) {
String[] servers = null;
try {
JobClient jc = new JobClient((JobConf) context.getConfiguration());
ClusterStatus status = jc.getClusterStatus(true);
Collection<String> atc = status.getActiveTrackerNames();
servers = new String[atc.size()];
int s = 0;
for (String serverInfo : atc) {
// System.out.println("serverInfo:" + serverInfo);
StringTokenizer st = new StringTokenizer(serverInfo, ":");
String trackerName = st.nextToken();
// System.out.println("trackerName:" + trackerName);
StringTokenizer st1 = new StringTokenizer(trackerName, "_");
st1.nextToken();
servers[s++] = st1.nextToken();
}
} catch (IOException e) {
e.printStackTrace();
}
return servers;
}
private List<InputSplit> getSplits(JobContext cx, List<Path> dirs)
throws FileNotFoundException, IOException {
List<InputSplit> splits = Lists.newArrayList();
List<Path> subdirs = Lists.newArrayList();
long totalFileCount = 0;
FileSystem fs = FileSystem.get(cx.getConfiguration());
for (Path input : dirs) {
long count = fs.getContentSummary(input).getFileCount();
subdirs.add(input);
if (totalFileCount + count < GET_SPLIT_NUM_FILES_TRHESHOLD) {
totalFileCount += count;
} else {
addAvroFilesInSubdirsToSplits(splits, subdirs, fs, cx);
subdirs.clear();
totalFileCount = 0;
}
}
if (totalFileCount > 0) {
addAvroFilesInSubdirsToSplits(splits, subdirs, fs, cx);
}
return splits;
}
/**
* This functions modifies the splits list in place to retain only a random fraction of the input
* splits. The fraction is expected in "starfish.profiler.sampling.fraction" as a number between 0
* and 1. The default value is 0.1.
*
* @param job The job context
* @param splits The list of input splits to modify
*/
public static void sampleInputSplits(JobContext job, List<InputSplit> splits) {
// Get the sampling fraction
Configuration conf = job.getConfiguration();
double fraction = conf.getFloat(Profiler.PROFILER_SAMPLING_FRACTION, 0.1f);
if (fraction < 0 || fraction > 1)
throw new RuntimeException("ERROR: Invalid sampling fraction: " + fraction);
// Handle corner cases
if (fraction == 0 || splits.size() == 0) {
splits.clear();
return;
}
if (fraction == 1) return;
// Calculate the number of samples
int numSplits = splits.size();
int sampleSize = (int) Math.round(numSplits * fraction);
if (sampleSize == 0) sampleSize = 1;
// Shuffle the splits
Collections.shuffle(splits);
// Retain only a sampleSize number of splits
for (int i = splits.size() - 1; i >= sampleSize; --i) {
splits.remove(i);
}
nf.setMaximumFractionDigits(2);
LOG.info("Executing only " + nf.format(fraction * 100) + "% of the map tasks");
}
@Override
public List<InputSplit> getSplits(JobContext context) throws IOException, InterruptedException {
Configuration conf = context.getConfiguration();
final String overlordUrl = conf.get(CONF_DRUID_OVERLORD_HOSTPORT);
final String storageDir = conf.get(CONF_DRUID_STORAGE_STORAGE_DIR);
String dataSource = conf.get(CONF_DRUID_DATASOURCE);
String intervalStr = conf.get(CONF_DRUID_INTERVAL);
logger.info("druid overlord url = " + overlordUrl);
logger.info("druid storage dir = " + storageDir);
logger.info("druid datasource = " + dataSource);
logger.info("druid datasource interval = " + intervalStr);
// TODO: currently we are creating 1 split per segment which is not really
// necessary, we can use some configuration to combine multiple segments into
// one input split
List<InputSplit> splits =
Lists.transform(
druid.getSegmentPathsToLoad(
dataSource, new Interval(intervalStr), storageDir, overlordUrl),
new Function<String, InputSplit>() {
@Override
public InputSplit apply(String input) {
return new DruidInputSplit(input);
}
});
logger.info("Number of splits = " + splits.size());
return splits;
}
/**
* Splitter used by both Vertex and Edge Input Format.
*
* @param context The job context
* @param estimation Number of estimated objects
* @return splits to be generated to read the input
*/
public static List<InputSplit> getSplits(JobContext context, long estimation)
throws IOException, InterruptedException {
int chunks = context.getConfiguration().getInt("mapred.map.tasks", 1);
long chunkSize = estimation / chunks;
List<InputSplit> splits = new ArrayList<InputSplit>();
if (LOG.isDebugEnabled()) {
LOG.debug(String.format("Estimated objects: %d", estimation));
LOG.debug(String.format("Number of chunks: %d", chunks));
}
for (int i = 0; i < chunks; ++i) {
long start = i * chunkSize;
long end = ((i + 1) == chunks) ? Long.MAX_VALUE : (i * chunkSize) + chunkSize;
RexsterInputSplit split = new RexsterInputSplit(start, end);
splits.add(split);
if (LOG.isDebugEnabled()) {
LOG.debug(String.format("Chunk: start %d; end %d;", start, end));
LOG.debug(String.format("Chunk: size %d;", chunkSize));
LOG.debug(split);
}
}
return splits;
}
/**
* Get a PathFilter instance of the filter set for the input paths.
*
* @return the PathFilter instance set for the job, NULL if none has been set.
*/
public static PathFilter getInputPathFilter(JobContext context) {
Configuration conf = context.getConfiguration();
Class<?> filterClass = conf.getClass(PATHFILTER_CLASS, null, PathFilter.class);
return (filterClass != null)
? (PathFilter) ReflectionUtils.newInstance(filterClass, conf)
: null;
}
@Override
public List<LuceneSegmentInputSplit> getSplits(JobContext context)
throws IOException, InterruptedException {
Configuration configuration = context.getConfiguration();
LuceneStorageConfiguration lucene2SeqConfiguration =
new LuceneStorageConfiguration(configuration);
List<LuceneSegmentInputSplit> inputSplits = new ArrayList<>();
List<Path> indexPaths = lucene2SeqConfiguration.getIndexPaths();
for (Path indexPath : indexPaths) {
ReadOnlyFileSystemDirectory directory =
new ReadOnlyFileSystemDirectory(
FileSystem.get(configuration), indexPath, false, configuration);
SegmentInfos segmentInfos = new SegmentInfos();
segmentInfos.read(directory);
for (SegmentCommitInfo segmentInfo : segmentInfos) {
LuceneSegmentInputSplit inputSplit =
new LuceneSegmentInputSplit(
indexPath, segmentInfo.info.name, segmentInfo.sizeInBytes());
inputSplits.add(inputSplit);
LOG.info(
"Created {} byte input split for index '{}' segment {}",
segmentInfo.sizeInBytes(),
indexPath.toUri(),
segmentInfo.info.name);
}
}
return inputSplits;
}
private void addAvroFilesInSubdirsToSplits(
List<InputSplit> splits, List<Path> subdirs, FileSystem fs, JobContext cx)
throws FileNotFoundException, IOException {
List<Path> files = findAvroFilesInDirs(subdirs, fs);
Job helperJob = Job.getInstance(cx.getConfiguration());
setInputPaths(helperJob, files.toArray(new Path[files.size()]));
splits.addAll(super.getSplits(helperJob));
}
@Override
protected boolean isSplitable(JobContext context, Path filename) {
CompressionCodec codec =
new CompressionCodecFactory(context.getConfiguration()).getCodec(filename);
return codec == null;
}
@Override
protected boolean isSplitable(JobContext context, Path file) {
CompressionCodec codec = new CompressionCodecFactory(context.getConfiguration()).getCodec(file);
if (null == codec) {
return true;
}
return codec instanceof SplittableCompressionCodec;
}
public List<InputSplit> getSplits(JobContext jobContext) {
List<InputSplit> ret = new ArrayList<InputSplit>();
int numSplits = jobContext.getConfiguration().getInt(MRJobConfig.NUM_MAPS, 1);
for (int i = 0; i < numSplits; ++i) {
ret.add(new EmptySplit());
}
return ret;
}
/** Create the desired number of splits, dividing the number of rows between the mappers. */
@Override
public List<InputSplit> getSplits(JobContext job) {
long totalRows = job.getConfiguration().getLong(NUMROWS, 0);
int numSplits = job.getConfiguration().getInt(NUMSPLITS, 1);
long rowsPerSplit = totalRows / numSplits;
System.out.println(
"Generating " + totalRows + " using " + numSplits + " maps with step of " + rowsPerSplit);
ArrayList<InputSplit> splits = new ArrayList<>(numSplits);
long currentRow = 0;
for (int split = 0; split < numSplits - 1; ++split) {
splits.add(new RangeInputSplit(currentRow, rowsPerSplit));
currentRow += rowsPerSplit;
}
splits.add(new RangeInputSplit(currentRow, totalRows - currentRow));
System.out.println("Done Generating.");
return splits;
}
/**
* Get the list of input {@link Path}s for the map-reduce job.
*
* @param context The job
* @return the list of input {@link Path}s for the map-reduce job.
*/
public static Path[] getInputPaths(JobContext context) {
String dirs = context.getConfiguration().get(INPUT_DIR, "");
String[] list = StringUtils.split(dirs);
Path[] result = new Path[list.length];
for (int i = 0; i < list.length; i++) {
result[i] = new Path(StringUtils.unEscapeString(list[i]));
}
return result;
}
/**
* Sets the access mode for stage resources in the job.
*
* @param context the current job context
* @param mode the access mode
* @since 0.7.1
*/
public static void setAccessMode(JobContext context, AccessMode mode) {
if (context == null) {
throw new IllegalArgumentException("context must not be null"); // $NON-NLS-1$
}
if (mode == null) {
throw new IllegalArgumentException("mode must not be null"); // $NON-NLS-1$
}
context.getConfiguration().set(KEY_ACCESS_MODE, mode.encode());
}
/** @inheritDoc */
@Override
protected boolean isSplitable(JobContext context, Path file) {
CompressionCodec codec = new CompressionCodecFactory(context.getConfiguration()).getCodec(file);
if (codec != null) {
return false;
}
return true;
}
