List of usage examples for org.apache.hadoop.fs FileSystem mkdirs
public boolean mkdirs(Path f) throws IOException
From source file:com.blackberry.logtools.logmultisearch.java
License:Apache License
@SuppressWarnings("static-access") public int run(String[] argv) throws Exception { //Configuring configuration and filesystem to work on HDFS final Configuration conf = getConf(); //Configuration processed by ToolRunner FileSystem fs = FileSystem.get(conf); //Initiate tools used for running search LogTools tools = new LogTools(); //Other options String date_format = "RFC5424"; String field_separator = " "; ArrayList<String> D_options = new ArrayList<String>(); boolean quiet = true; boolean silent = false; boolean log = false; boolean forcelocal = false; boolean forceremote = false; //The arguments are // - strings//from w w w . j ava 2 s . c o m // - dc number // - service // - component // - startTime (Something 'date' can parse, or just a time in ms from epoch) // - endTime (Same as start) // - outputDir //Indexing for arguments to be passed for Mapreduce int stringsNum = 0; int dcNum = 1; int svcNum = 2; int compNum = 3; int startNum = 4; int endNum = 5; int outNum = 6; //Parsing through user arguments String[] args = new String[7]; int count = 0; //Count created to track the parse of all arguments int argcount = 0; //Count created to track number of arguments to be passed on while (count < argv.length) { String arg = argv[count]; count++; if (arg.equals("--")) { break; } else if (arg.startsWith("-")) { if (arg.equals("--v")) { quiet = tools.parseV(silent); } else if (arg.equals("--i")) { conf.set("logdriver.search.case.insensitive", "true"); } else if (arg.equals("--a")) { LogTools.logConsole(quiet, silent, warn, "AND searching selected"); conf.set("logdriver.search.and", "true"); } else if (arg.startsWith("--dateFormat=")) { arg = arg.replace("--dateFormat=", ""); date_format = arg; } else if (arg.startsWith("--fieldSeparator=")) { arg = arg.replace("--fieldSeparator=", ""); field_separator = arg; } else if (arg.startsWith("-strings=")) { arg = arg.replace("-strings=", ""); args[stringsNum] = arg; argcount++; } else if (arg.startsWith("-dc=")) { arg = arg.replace("-dc=", ""); args[dcNum] = arg; argcount++; } else if (arg.startsWith("-svc=")) { arg = arg.replace("-svc=", ""); args[svcNum] = arg; argcount++; } else if (arg.startsWith("-comp=")) { arg = arg.replace("-comp=", ""); args[compNum] = arg; argcount++; } else if (arg.startsWith("-start=")) { arg = arg.replace("-start=", ""); args[startNum] = arg; argcount++; } else if (arg.startsWith("-end=")) { arg = arg.replace("-end=", ""); args[endNum] = arg; argcount++; } //User inputs output directory that is to be created //Check to see if parent directory exists && output directory does not exist else if (arg.startsWith("--out=")) { args[outNum] = tools.parseOut(arg, fs); argcount++; } else if (arg.startsWith("-D")) { D_options.add(arg); } else if (arg.equals("--silent")) { silent = tools.parseSilent(quiet); } else if (arg.equals("--log")) { log = true; } else if (arg.equals("--l")) { forcelocal = tools.parsePigMode(forceremote); } else if (arg.equals("--r")) { forceremote = tools.parsePigMode(forcelocal); } else { LogTools.logConsole(quiet, silent, error, "Unrecognized option: " + arg); System.exit(1); } } else { LogTools.logConsole(quiet, silent, error, "Unrecognized option: " + arg); System.exit(1); } } //Default output should be stdout represented by "-" if (args[outNum] == null) { args[outNum] = "-"; argcount++; LogTools.logConsole(quiet, silent, info, "Output set to default stdout."); } if (argcount < 7) { System.err.println(";****************************************" + "\n\t\t\t NOT ENOUGH ARGUMENTS\n" + "\n\tUSAGE: logmultisearch [REQUIRED ARGUMENTS] [OPTIONS] (Order does not matter)" + "\n\tREQUIRED ARGUMENTS:" + "\n\t\t-strings=[STRINGS_DIR|STRINGS_FILE|STRING] String/file/directory of strings to search." + "\n\t\t-dc=[DATACENTER] Data Center." + "\n\t\t-svc=[SERVICE] Service." + "\n\t\t-comp=[COMPONENT] Component." + "\n\t\t-start=[START] Start time." + "\n\t\t-end=[END] End time." + "\n\tOptions:" + "\n\t\t--out=[DIRECTORY] Desired output directory. If not defined, output to stdout." + "\n\t\t--v Verbose output." + "\n\t\t--r Force remote sort." + "\n\t\t--l Force local sort." + "\n\t\t--dateFormat=[FORMAT] Valid formats are RFC822, RFC3164 (zero padded day)," + "\n\t RFC5424 (default), or any valid format string for FastDateFormat." + "\n\t\t--fieldSeparator=X The separator to use to separate fields in intermediate" + "\n\t files. Defaults to 'INFORMATION SEPARATOR ONE' (U+001F)." + "\n\t\t--silent Output only the data." + "\n\t\t--i Make search case insensitive." + "\n\t\t--a Enable AND searching." + "\n\t\t--log Save all the logs.\n" + ";****************************************"); System.exit(1); } //Parse time inputs for start and end of search args[startNum] = tools.parseDate(args[startNum]); args[endNum] = tools.parseDate(args[endNum]); tools.checkTime(args[startNum], args[endNum]); //Retrieve 'strings' argument to be able to pass search strings to HDFS //Retrieve 'out' argument to determine where output of results should be sent String strings = args[stringsNum]; String out = args[outNum]; //Generate files to temporarily store output of mapreduce jobs and pig logs locally File local_output = File.createTempFile("tmp.", RandomStringUtils.randomAlphanumeric(10)); if (log != true) { local_output.deleteOnExit(); } File pig_tmp = File.createTempFile("tmp.", RandomStringUtils.randomAlphanumeric(10)); if (log != true) { pig_tmp.deleteOnExit(); } //Name the temp directory for storing results in HDFS String tmp = "tmp/logmultisearch-" + RandomStringUtils.randomAlphanumeric(10); //Set args[stringsNum] to be location of search strings to be used for the Multisearch args[stringsNum] = (StringEscapeUtils.escapeJava(tmp) + "/strings"); //Set args[outNum] to be temp output directory to be passed onto MultiSearchByTime instead of UserInput argument args[outNum] = (StringEscapeUtils.escapeJava(tmp) + "/rawlines"); //Managing console output - deal with --v/--silent Logger LOG = LoggerFactory.getLogger(logmultisearch.class); tools.setConsoleOutput(local_output, quiet, silent); //Create temp directory in HDFS to store logsearch logs before sorting tools.tmpDirHDFS(quiet, silent, fs, conf, tmp, log); //If the strings argument is the path of a file, copy the file to HDFS. //If the strings argument is the path of a directory, copy all files in the directory to HDFS. //If the strings argument is not a path to a file/directory, write to a newly created file in HDFS. try { File f = new File(strings); if (f.isFile()) { LogTools.logConsole(quiet, silent, warn, "Strings input is a File..."); //dos2unix file conversion File dos2unix = File.createTempFile("tmp.", RandomStringUtils.randomAlphanumeric(10)); dos2unix.deleteOnExit(); tools.dosTounix(f, dos2unix); //Copy over temp directory into a new directory in HDFS to be used for logmultisearch fs.copyFromLocalFile(new Path(dos2unix.getAbsolutePath()), new Path(tmp + "/strings")); } else if (f.isDirectory()) { LogTools.logConsole(quiet, silent, warn, "Strings input is a Directory..."); //Get list of all files in directory to convert from dos2unix String[] fileList = f.list(); //Create temp directory to store all converted files File tempDir = Files.createTempDir(); tempDir.deleteOnExit(); //Convert all files from dos2unix and write to temp directory for (int i = 0; i < fileList.length; i++) { File dos2unix = File.createTempFile("unix", fileList[i], tempDir); dos2unix.deleteOnExit(); tools.dosTounix(new File(f.getAbsolutePath() + "/" + fileList[i]), dos2unix); } //Copy over temp directory into a new directory in HDFS to be used for logmultisearch fs.copyFromLocalFile(new Path(tempDir.getAbsolutePath()), new Path(tmp + "/strings")); } else { LogTools.logConsole(quiet, silent, warn, "Strings input is a search string..."); //Make directory and file for strings fs.mkdirs(new Path(tmp + "/strings")); fs.createNewFile(new Path(tmp + "/strings/strings")); //Write search strings to file FSDataOutputStream hdfsOut = fs.create(new Path(tmp + "/strings/strings")); hdfsOut.writeUTF(strings); hdfsOut.close(); } } catch (Exception e) { e.printStackTrace(); System.exit(1); } LogTools.logConsole(quiet, silent, warn, "Searching..."); LogTools.logConsole(quiet, silent, warn, "Passing Arguments: SearchStrings=" + strings + " DC=" + args[dcNum] + " Service=" + args[svcNum] + " Component=" + args[compNum] + " StartTime=" + args[startNum] + " EndTime=" + args[endNum] + " Output=" + out); //Set standard configuration for running Mapreduce and PIG String queue_name = "logsearch"; //Start Mapreduce job tools.runMRJob(quiet, silent, conf, D_options, out, LOG, field_separator, queue_name, args, "MultiSearchByTime", new MultiSearchByTime()); //Before sorting, determine the number of records and size of the results found long foundresults = tools.getResults(local_output); long size = tools.getSize(foundresults, tmp, fs); //Run PIG job if results found tools.runPig(silent, quiet, foundresults, size, tmp, out, D_options, queue_name, date_format, field_separator, pig_tmp, fs, conf, forcelocal, forceremote); //Display location of tmp files if log enabled tools.logs(log, local_output, pig_tmp, tmp); return 0; }
From source file:com.blackberry.logtools.LogTools.java
License:Apache License
public void tmpDirHDFS(boolean quiet, boolean silent, FileSystem fs, Configuration conf, String tmp, boolean log) { logConsole(quiet, silent, info, "Creating new Temp Directory in HDFS: " + tmp); try {/*from w ww . j a v a2 s . c om*/ Path path = new Path(tmp); if (!(fs.exists(path))) { //Create directory fs.mkdirs(path); if (log != true) { fs.deleteOnExit(path); } } } catch (IOException e) { if (e.toString().contains("Failed to find any Kerberos")) { logConsole(true, true, error, "No/bad Kerberos ticket - please authenticate."); System.exit(1); } else if (e.toString().contains("quota") && e.toString().contains("exceeded")) { logConsole(true, true, error, "Disk quota Exceeded."); System.exit(1); } e.printStackTrace(); System.exit(1); } }
From source file:com.chinacache.robin.format.action.SimpleMoveFileAction.java
License:Apache License
@Override public void execute(FileSystem fileSystem, Path filePath) throws IOException { String newDestPath = destination; newDestPath += "/BILog/src/" + dfs.format(new Date()) + "/" + userName + "/" + channelId + "/"; Path destPath = new Path(newDestPath); LOG.info("Moving file {} to {}", filePath, destPath); fileSystem.mkdirs(destPath); fileSystem.rename(filePath, destPath); }
From source file:com.ci.backports.hadoop.hbase.ZHFileOutputFormat.java
License:Apache License
public RecordWriter<ImmutableBytesWritable, KeyValue> getRecordWriter(final TaskAttemptContext context) throws IOException, InterruptedException { // Get the path of the temporary output file final Path outputPath = FileOutputFormat.getOutputPath(context); final Path outputdir = new FileOutputCommitter(outputPath, context).getWorkPath(); Configuration conf = context.getConfiguration(); final FileSystem fs = outputdir.getFileSystem(conf); // These configs. are from hbase-*.xml final long maxsize = conf.getLong("hbase.hregion.max.filesize", HConstants.DEFAULT_MAX_FILE_SIZE); final int blocksize = conf.getInt("hfile.min.blocksize.size", HFile.DEFAULT_BLOCKSIZE); // Invented config. Add to hbase-*.xml if other than default compression. final String compression = conf.get("hfile.compression", Compression.Algorithm.NONE.getName()); return new RecordWriter<ImmutableBytesWritable, KeyValue>() { // Map of families to writers and how much has been output on the writer. private final Map<byte[], WriterLength> writers = new TreeMap<byte[], WriterLength>( Bytes.BYTES_COMPARATOR); private byte[] previousRow = HConstants.EMPTY_BYTE_ARRAY; private final byte[] now = Bytes.toBytes(System.currentTimeMillis()); private boolean rollRequested = false; public void write(ImmutableBytesWritable row, KeyValue kv) throws IOException { // null input == user explicitly wants to flush if (row == null && kv == null) { rollWriters();//w ww . j a v a 2s .co m return; } byte[] rowKey = kv.getRow(); long length = kv.getLength(); byte[] family = kv.getFamily(); WriterLength wl = this.writers.get(family); // If this is a new column family, verify that the directory exists if (wl == null) { fs.mkdirs(new Path(outputdir, Bytes.toString(family))); } // If any of the HFiles for the column families has reached // maxsize, we need to roll all the writers if (wl != null && wl.written + length >= maxsize) { this.rollRequested = true; } // This can only happen once a row is finished though if (rollRequested && Bytes.compareTo(this.previousRow, rowKey) != 0) { rollWriters(); } // create a new HLog writer, if necessary if (wl == null || wl.writer == null) { wl = getNewWriter(family); } // we now have the proper HLog writer. full steam ahead kv.updateLatestStamp(this.now); wl.writer.append(kv); wl.written += length; // Copy the row so we know when a row transition. this.previousRow = rowKey; } private void rollWriters() throws IOException { for (WriterLength wl : this.writers.values()) { if (wl.writer != null) { LOG.info("Writer=" + wl.writer.getPath() + ((wl.written == 0) ? "" : ", wrote=" + wl.written)); close(wl.writer); } wl.writer = null; wl.written = 0; } this.rollRequested = false; } /* Create a new HFile.Writer. * @param family * @return A WriterLength, containing a new HFile.Writer. * @throws IOException */ private WriterLength getNewWriter(byte[] family) throws IOException { WriterLength wl = new WriterLength(); Path familydir = new Path(outputdir, Bytes.toString(family)); wl.writer = new HFile.Writer(fs, StoreFile.getUniqueFile(fs, familydir), blocksize, compression, KeyValue.KEY_COMPARATOR); this.writers.put(family, wl); return wl; } private void close(final HFile.Writer w) throws IOException { if (w != null) { w.appendFileInfo(StoreFile.BULKLOAD_TIME_KEY, Bytes.toBytes(System.currentTimeMillis())); w.appendFileInfo(StoreFile.BULKLOAD_TASK_KEY, Bytes.toBytes(context.getTaskAttemptID().toString())); w.appendFileInfo(StoreFile.MAJOR_COMPACTION_KEY, Bytes.toBytes(true)); w.close(); } } public void close(TaskAttemptContext c) throws IOException, InterruptedException { for (WriterLength wl : this.writers.values()) { close(wl.writer); } } }; }
From source file:com.citic.zxyjs.zwlscx.mapreduce.lib.input.HFileOutputFormatBase.java
License:Apache License
public RecordWriter<ImmutableBytesWritable, KeyValue> getRecordWriter(final TaskAttemptContext context) throws IOException, InterruptedException { // Get the path of the temporary output file final Path outputPath = FileOutputFormat.getOutputPath(context); final Path outputdir = new FileOutputCommitter(outputPath, context).getWorkPath(); final Path ignoreOutputPath = new Path(outputPath + "_ignore"); final Configuration conf = context.getConfiguration(); final FileSystem fs = outputdir.getFileSystem(conf); // These configs. are from hbase-*.xml final long maxsize = conf.getLong(HConstants.HREGION_MAX_FILESIZE, HConstants.DEFAULT_MAX_FILE_SIZE); // Invented config. Add to hbase-*.xml if other than default // compression. final String defaultCompression = conf.get("hfile.compression", Compression.Algorithm.NONE.getName()); final boolean compactionExclude = conf.getBoolean("hbase.mapreduce.hfileoutputformat.compaction.exclude", false);//from ww w. j a va 2 s .c o m if (fs.exists(ignoreOutputPath)) { LOG.info("Deleted " + ignoreOutputPath.toString() + " success."); fs.delete(ignoreOutputPath, true); } // create a map from column family to the compression algorithm final Map<byte[], String> compressionMap = createFamilyCompressionMap(conf); final Map<byte[], String> bloomTypeMap = createFamilyBloomMap(conf); final Map<byte[], String> blockSizeMap = createFamilyBlockSizeMap(conf); String dataBlockEncodingStr = conf.get(DATABLOCK_ENCODING_CONF_KEY); final HFileDataBlockEncoder encoder; if (dataBlockEncodingStr == null) { encoder = NoOpDataBlockEncoder.INSTANCE; } else { try { encoder = new HFileDataBlockEncoderImpl(DataBlockEncoding.valueOf(dataBlockEncodingStr)); } catch (IllegalArgumentException ex) { throw new RuntimeException("Invalid data block encoding type configured for the param " + DATABLOCK_ENCODING_CONF_KEY + " : " + dataBlockEncodingStr); } } return new RecordWriter<ImmutableBytesWritable, KeyValue>() { // Map of families to writers and how much has been output on the // writer. private final Map<byte[], WriterLength> writers = new TreeMap<byte[], WriterLength>( Bytes.BYTES_COMPARATOR); private final FSDataOutputStream dos = fs.create(ignoreOutputPath); private byte[] previousRow = HConstants.EMPTY_BYTE_ARRAY; private final byte[] now = Bytes.toBytes(System.currentTimeMillis()); private boolean rollRequested = false; public void write(ImmutableBytesWritable row, KeyValue kv) throws IOException { // null input == user explicitly wants to flush if (row == null && kv == null) { rollWriters(); return; } byte[] rowKey = kv.getRow(); long length = kv.getLength(); byte[] family = kv.getFamily(); if (ignore(kv)) { byte[] readBuf = rowKey; dos.write(readBuf, 0, readBuf.length); dos.write(Bytes.toBytes("\n")); return; } WriterLength wl = this.writers.get(family); // If this is a new column family, verify that the directory // exists if (wl == null) { Path path = null; path = new Path(outputdir, Bytes.toString(family)); fs.mkdirs(path); } // If any of the HFiles for the column families has reached // maxsize, we need to roll all the writers if (wl != null && wl.written + length >= maxsize) { this.rollRequested = true; } // This can only happen once a row is finished though if (rollRequested && Bytes.compareTo(this.previousRow, rowKey) != 0) { rollWriters(); } // create a new HLog writer, if necessary if (wl == null || wl.writer == null) { wl = getNewWriter(family, conf); } // we now have the proper HLog writer. full steam ahead kv.updateLatestStamp(this.now); wl.writer.append(kv); wl.written += length; // Copy the row so we know when a row transition. this.previousRow = rowKey; } private void rollWriters() throws IOException { for (WriterLength wl : this.writers.values()) { if (wl.writer != null) { LOG.info("Writer=" + wl.writer.getPath() + ((wl.written == 0) ? "" : ", wrote=" + wl.written)); close(wl.writer); } wl.writer = null; wl.written = 0; } this.rollRequested = false; } /* * Create a new StoreFile.Writer. * @param family * @return A WriterLength, containing a new StoreFile.Writer. * @throws IOException */ private WriterLength getNewWriter(byte[] family, Configuration conf) throws IOException { WriterLength wl = new WriterLength(); Path familydir = new Path(outputdir, Bytes.toString(family)); String compression = compressionMap.get(family); compression = compression == null ? defaultCompression : compression; String bloomTypeStr = bloomTypeMap.get(family); BloomType bloomType = BloomType.NONE; if (bloomTypeStr != null) { bloomType = BloomType.valueOf(bloomTypeStr); } String blockSizeString = blockSizeMap.get(family); int blockSize = blockSizeString == null ? HConstants.DEFAULT_BLOCKSIZE : Integer.parseInt(blockSizeString); Configuration tempConf = new Configuration(conf); tempConf.setFloat(HConstants.HFILE_BLOCK_CACHE_SIZE_KEY, 0.0f); wl.writer = new StoreFile.WriterBuilder(conf, new CacheConfig(tempConf), fs, blockSize) .withOutputDir(familydir) .withCompression(AbstractHFileWriter.compressionByName(compression)) .withBloomType(bloomType).withComparator(KeyValue.COMPARATOR).withDataBlockEncoder(encoder) .withChecksumType(HStore.getChecksumType(conf)) .withBytesPerChecksum(HStore.getBytesPerChecksum(conf)).build(); this.writers.put(family, wl); return wl; } private void close(final StoreFile.Writer w) throws IOException { if (w != null) { w.appendFileInfo(StoreFile.BULKLOAD_TIME_KEY, Bytes.toBytes(System.currentTimeMillis())); w.appendFileInfo(StoreFile.BULKLOAD_TASK_KEY, Bytes.toBytes(context.getTaskAttemptID().toString())); w.appendFileInfo(StoreFile.MAJOR_COMPACTION_KEY, Bytes.toBytes(true)); w.appendFileInfo(StoreFile.EXCLUDE_FROM_MINOR_COMPACTION_KEY, Bytes.toBytes(compactionExclude)); w.appendTrackedTimestampsToMetadata(); w.close(); } } public void close(TaskAttemptContext c) throws IOException, InterruptedException { dos.flush(); dos.close(); for (WriterLength wl : this.writers.values()) { close(wl.writer); } } }; }
From source file:com.cloudera.beeswax.Server.java
License:Apache License
/** * Hive won't work unless /tmp and /user/hive/warehouse are usable, * so we create them for the user.// w w w .j a v a2 s .c o m */ private static void createDirectoriesAsNecessary() { try { LOG.debug("Classpath: " + System.getProperty("java.class.path")); HiveConf conf = new HiveConf(Driver.class); FileSystem fs = FileSystem.get(conf); Path tmpDir = new Path("/tmp"); Path metaDir = new Path(conf.get(HiveConf.ConfVars.METASTOREWAREHOUSE.varname)); for (Path dir : new Path[] { tmpDir, metaDir }) { if (!fs.exists(dir)) { if (fs.mkdirs(dir)) { fs.setPermission(dir, new FsPermission((short) 0777)); LOG.info("Created " + dir + " with world-writable permissions."); } else { LOG.error("Could not create " + dir); } } } } catch (IOException e) { HiveConf conf = new HiveConf(Driver.class); LOG.error("Error while trying to check/create /tmp and warehouse directory " + conf.get(HiveConf.ConfVars.METASTOREWAREHOUSE.varname), e); } }
From source file:com.cloudera.CacheTool.java
License:Apache License
public static void createFile(FileSystem fs, Path fileName, long fileLen) throws IOException { int bufferLen = 1024; assert bufferLen > 0; if (!fs.mkdirs(fileName.getParent())) { throw new IOException("Mkdirs failed to create " + fileName.getParent().toString()); }//from ww w. j a va 2 s . c o m FSDataOutputStream out = null; try { out = fs.create(fileName, true, fs.getConf().getInt(CommonConfigurationKeys.IO_FILE_BUFFER_SIZE_KEY, 4096), (short) 1, fs.getDefaultBlockSize(fileName)); if (fileLen > 0) { byte[] toWrite = new byte[bufferLen]; Random rb = new Random(0); long bytesToWrite = fileLen; while (bytesToWrite > 0) { rb.nextBytes(toWrite); int bytesToWriteNext = (bufferLen < bytesToWrite) ? bufferLen : (int) bytesToWrite; out.write(toWrite, 0, bytesToWriteNext); bytesToWrite -= bytesToWriteNext; } } } finally { if (out != null) { out.close(); } } }
From source file:com.cloudera.cdk.data.filesystem.FileSystemDatasetRepository.java
License:Apache License
/** * Creates, if necessary, the given the location for {@code descriptor}. * * @param conf A Configuration//from w ww. j a v a2s . c om * @param descriptor A DatasetDescriptor */ static void ensureExists(DatasetDescriptor descriptor, Configuration conf) { Preconditions.checkArgument(descriptor.getLocation() != null, "Cannot get FileSystem for a descriptor with no location"); final Path dataPath = new Path(descriptor.getLocation()); final FileSystem fs = fsForPath(dataPath, conf); try { if (!fs.exists(dataPath)) { fs.mkdirs(dataPath); } } catch (IOException ex) { throw new DatasetRepositoryException("Cannot access data location", ex); } }
From source file:com.cloudera.cdk.data.filesystem.FileSystemWriters.java
License:Apache License
@SuppressWarnings("unchecked") // See https://github.com/Parquet/parquet-mr/issues/106 public static <E> DatasetWriter<E> newFileWriter(FileSystem fs, Path path, DatasetDescriptor descriptor) { // ensure the path exists try {/*from ww w . j av a2 s . co m*/ fs.mkdirs(path); } catch (IOException ex) { throw new DatasetWriterException("Could not create path:" + path, ex); } final Format format = descriptor.getFormat(); final Path file = new Path(path, uniqueFilename(descriptor.getFormat())); if (Formats.PARQUET.equals(format)) { return new ParquetFileSystemDatasetWriter(fs, file, descriptor.getSchema()); } else if (Formats.AVRO.equals(format)) { return new FileSystemDatasetWriter.Builder().fileSystem(fs).path(file).schema(descriptor.getSchema()) .build(); } else { throw new UnknownFormatException("Unknown format:" + format); } }
From source file:com.cloudera.circus.test.TestXTest.java
License:Open Source License
@Test @TestHadoop/*from www. j ava 2 s . c o m*/ public void testHadoopMapReduce() throws Exception { JobConf conf = getHadoopConf(); FileSystem fs = FileSystem.get(conf); JobClient jobClient = new JobClient(conf); try { Path inputDir = new Path(getHadoopTestDir(), "input"); Path outputDir = new Path(getHadoopTestDir(), "output"); fs.mkdirs(inputDir); Writer writer = new OutputStreamWriter(fs.create(new Path(inputDir, "data.txt"))); writer.write("a\n"); writer.write("b\n"); writer.write("c\n"); writer.close(); JobConf jobConf = getHadoopConf(); jobConf.setInt("mapred.map.tasks", 1); jobConf.setInt("mapred.map.max.attempts", 1); jobConf.setInt("mapred.reduce.max.attempts", 1); jobConf.set("mapred.input.dir", inputDir.toString()); jobConf.set("mapred.output.dir", outputDir.toString()); final RunningJob runningJob = jobClient.submitJob(jobConf); waitFor(60 * 1000, true, new Predicate() { @Override public boolean evaluate() throws Exception { return runningJob.isComplete(); } }); Assert.assertTrue(runningJob.isSuccessful()); Assert.assertTrue(fs.exists(new Path(outputDir, "part-00000"))); BufferedReader reader = new BufferedReader( new InputStreamReader(fs.open(new Path(outputDir, "part-00000")))); Assert.assertTrue(reader.readLine().trim().endsWith("a")); Assert.assertTrue(reader.readLine().trim().endsWith("b")); Assert.assertTrue(reader.readLine().trim().endsWith("c")); Assert.assertNull(reader.readLine()); reader.close(); } finally { fs.close(); jobClient.close(); } }