Package com.m6d.filecrush.crush

Examples of com.m6d.filecrush.crush.Crush

98
99
100
101
102
103
104
105
106
107
108
    createFile(subsubdir, "skipped-4", 4, 25);
    createFile(subsubdir, "skipped-5", 5, 25);

    File out = new File(tmp.getRoot(), "out");

    ToolRunner.run(job, new Crush(), new String[] {
        subdir.getAbsolutePath(), out.getAbsolutePath()
    });

    /*
     * Make sure the original files are still there.
View Full Code Here
130
131
132
133
134
135
136
137
138
139
140
  public void noFiles() throws Exception {
    File in = tmp.newFolder("in");

    File out = new File(tmp.getRoot(), "out");

    ToolRunner.run(job, new Crush(), new String[] {
        in.getAbsolutePath(), out.getAbsolutePath()
    });

    assertThat(out.exists(), is(false));
  }
View Full Code Here
101
102
103
104
105
106
107
  public void restoreJavaIoTmpDir() {
    System.setProperty("java.io.tmpdir", javaIoTmpDir);
  }

  private void run(String... args) throws Exception {
    ToolRunner.run(job, new Crush(), args);
  }
View Full Code Here
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
    expectedBucketFiles.add(format("%s  %s", dir2_4_2.getAbsolutePath() + "-0", new File(dir2_4_2, "file1").getAbsolutePath()));
    expectedBucketFiles.add(format("%s  %s", dir2_4_2.getAbsolutePath() + "-0", new File(dir2_4_2, "file2").getAbsolutePath()));
    expectedBucketFiles.add(format("%s  %s", dir2_4_2.getAbsolutePath() + "-0", new File(dir2_4_2, "file3").getAbsolutePath()));


    Crush crush = new Crush();

    crush.setConf(job);
    crush.setFileSystem(fileSystem);

    /*
     * Call these in the same order that run() does.
     */
    crush.createJobConfAndParseArgs("--compress=none", "--max-file-blocks=1", in.getAbsolutePath(), new File(tmp.getRoot(), "out").getAbsolutePath(), "20101124171730");
    crush.writeDirs();


    /*
     * Verify bucket contents.
     */

    List<String> actualBucketFiles = new ArrayList<String>();

    Text key = new Text();
    Text value = new Text();

    Reader reader = new Reader(FileSystem.get(job), crush.getBucketFiles(), job);

    while(reader.next(key, value)) {
      actualBucketFiles.add(format("%s\t%s", key, value));
    }

    reader.close();

    Collections.sort(expectedBucketFiles);
    Collections.sort(actualBucketFiles);

    assertThat(actualBucketFiles, equalTo(expectedBucketFiles));

    /*
     * Verify the partition map.
     */
    Reader partitionMapReader = new Reader(FileSystem.get(job), crush.getPartitionMap(), job);

    IntWritable partNum = new IntWritable();

    Map<String, Integer> actualPartitions = new HashMap<String, Integer>();

    while (partitionMapReader.next(key, partNum)) {
      actualPartitions.put(key.toString(), partNum.get());
    }

    partitionMapReader.close();

    /*
     * These crush files need to allocated into 5 partitions:
     *
     * in/2-1            55 bytes
     * in/1/1.1-0        45 bytes
     * in/1/1.1-2        40 bytes
     * in/1/1.1-1        40 bytes
     * in/1/1.2-0        30 bytes
     * in/2/2.2-1        40 bytes
     * in/2/2.4/2.4.2-0  50 bytes
     *
     *   0              1                      2                3                4
     *   in/2-1 55      in/2/2.4/2.4.2-0  50  in/1/1.1-0  45  in/1/1.1-2  40  in/1/1.1-1  40
     *                                                       in/2/2.2-1  40  in/1/1.2-0  39
     */
    Map<String, Integer> expectedPartitions = new HashMap<String, Integer>();

    //TODO: this may not be deterministic due to jvm/hashmap/filesystem
    expectedPartitions.put(dir2.getAbsolutePath() + "-1",      0);
    expectedPartitions.put(dir2_4_2.getAbsolutePath() + "-0",  1);
    expectedPartitions.put(dir1_1.getAbsolutePath() + "-0",    2);
    expectedPartitions.put(dir1_1.getAbsolutePath() + "-2",    4);
    expectedPartitions.put(dir2_2.getAbsolutePath() + "-1",    3);
    expectedPartitions.put(dir1_1.getAbsolutePath() + "-1",    3);
    expectedPartitions.put(dir1_2.getAbsolutePath() + "-0",    4);

    assertThat(actualPartitions, equalTo(expectedPartitions));


    /*
     * Verify counters.
     */
    Counters actualCounters = new Counters();

    DataInputStream countersStream = FileSystem.get(job).open(crush.getCounters());

    actualCounters.readFields(countersStream);

    countersStream.close();

View Full Code Here
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
  @Test
  public void execute() throws Exception {
    writeFiles(true, true, true);

    Crush crush = new Crush();

    ToolRunner.run(job, crush, new String [] {
      "--threshold=0.015",
      "--max-file-blocks=1",
      "--verbose",

      "--regex=.+/other",
      "--replacement=${crush.timestamp}-${crush.task.num}-middle-${crush.file.num}-tail",
      "--input-format=" + SequenceFileInputFormat.class.getName(),
      "--output-format=" + TextOutputFormat.class.getName(),

      "--regex=.+/dir",
      "--replacement=secondregex-${crush.timestamp}-${crush.task.num}-${crush.file.num}",
      "--input-format=" + TextInputFormat.class.getName(),
      "--output-format=" + SequenceFileOutputFormat.class.getName(),

      "--regex=.+/dir/([^/]+/)*(.+)",
      "--replacement=thirdregex-$2-${crush.timestamp}-${crush.task.num}-${crush.file.num}",
      "--input-format=" + SequenceFileInputFormat.class.getName(),
      "--output-format=" + SequenceFileOutputFormat.class.getName(),

      "--regex=.+/text",
      "--replacement=fourthregex-${crush.task.num}-${crush.timestamp}-${crush.file.num}",
      "--input-format=" + TextInputFormat.class.getName(),
      "--output-format=" + TextOutputFormat.class.getName(),

      /*
       * This is the default regex and replacement, which we add last so we can exercise the default logic.
       */
      "--regex=.+",
      "--replacement=crushed_file-${crush.timestamp}-${crush.task.num}-${crush.file.num}",
      "--input-format=" + SequenceFileInputFormat.class.getName(),
      "--output-format=" + TextOutputFormat.class.getName(),

      "--compress=" + CustomCompressionCodec.class.getName(),

      "in", "out", "20101116153015"
    });


    /*
     * Crushed files.
     */
    verifyOutput(homeDir + "/out/dir",                  "secondregex-20101116153015-*-*",            Format.TEXT,      Format.SEQUENCE,  customCodec, "file10", "file11", "file12", "file13");
    verifyOutput(homeDir + "/out/dir/subdir",            "thirdregex-subdir-20101116153015-*-*",      Format.SEQUENCE,  Format.SEQUENCE,  customCodec, "file20", "file21", "file22", "file23", "file24");
    verifyOutput(homeDir + "/out/dir/subdir/subsubdir",  "thirdregex-subsubdir-20101116153015-*-*",  Format.SEQUENCE,  Format.SEQUENCE,  customCodec, "file30", "file31", "file32", "file33", "file34");
    verifyOutput(homeDir + "/out/dir/subdir/other",      "20101116153015-*-middle-*-tail",            Format.SEQUENCE,  Format.TEXT,       customCodec, "file40", "file41", "file42", "file43");
    verifyOutput(homeDir + "/out/dir/other",            "20101116153015-*-middle-*-tail",            Format.SEQUENCE,  Format.TEXT,      customCodec, "file50", "file51", "file52", "file53", "file54", "file55");
    verifyOutput(homeDir + "/out/text",                  "fourthregex-*-20101116153015-*",            Format.TEXT,      Format.TEXT,      customCodec, "file60", "file61", "file62", "file63");
    verifyOutput(homeDir + "/out",                      "crushed_file-20101116153015-*-*",          Format.SEQUENCE,  Format.TEXT,      customCodec, "file70", "file71", "file72");


    /*
     * Skipped files should have been moved to the output dir.
     */
    verifyOutput(homeDir + "/out/dir/skipped", "file80", Format.SEQUENCE, Format.SEQUENCE, defaultCodec, "file80");
    verifyHugeFile(homeDir + "/out/huge", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);


    /*
     * Crush input files should remain in the input dir.
     */
    for (String file : new String[] { "file10", "file11", "file12", "file13" }) {
      verifyOutput(homeDir + "/in/dir", file, Format.TEXT, Format.TEXT, null, file);
    }

    for (String file : new String[] { "file20", "file21", "file22", "file23", "file24" }) {
      verifyOutput(homeDir + "/in/dir/subdir", file, Format.SEQUENCE, Format.SEQUENCE, defaultCodec, file);
    }

    for (String file : new String[] { "file30", "file31", "file32", "file33", "file34" }) {
      verifyOutput(homeDir + "/in/dir/subdir/subsubdir", file, Format.SEQUENCE, Format.SEQUENCE, defaultCodec, file);
    }

    for (String file : new String[] { "file40", "file41", "file42", "file43" }) {
      verifyOutput(homeDir + "/in/dir/subdir/other", file, Format.SEQUENCE, Format.SEQUENCE, defaultCodec, file);
    }

    for (String file : new String[] { "file50", "file51", "file52", "file53", "file54", "file55" }) {
      verifyOutput(homeDir + "/in/dir/other", file, Format.SEQUENCE, Format.SEQUENCE, defaultCodec, file);
    }

    for (String file : new String[] { "file60", "file61", "file62", "file63" }) {
      verifyOutput(homeDir + "/in/text", file, Format.TEXT, Format.TEXT, null, file);
    }

    for (String file : new String[] { "file70", "file71", "file72" }) {
      verifyOutput(homeDir + "/in", file, Format.SEQUENCE, Format.SEQUENCE, defaultCodec, file);
    }


    Counters jobCounters = crush.getJobCounters();

    assertThat(jobCounters.getCounter(MapperCounter.DIRS_FOUND),      equalTo( 8L));
    assertThat(jobCounters.getCounter(MapperCounter.DIRS_ELIGIBLE),    equalTo( 7L));
    assertThat(jobCounters.getCounter(MapperCounter.DIRS_SKIPPED),    equalTo( 1L));
    assertThat(jobCounters.getCounter(MapperCounter.FILES_FOUND),      equalTo(33L));
View Full Code Here
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
  public void executeHugeFilesOnly() throws Exception {
    writeHugeFile("in/huge", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);
    writeHugeFile("in/foo/huge1", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);
    writeHugeFile("in/foo/huge2", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);

    Crush crush = new Crush();

    ToolRunner.run(job, crush, new String [] {
      "--threshold=0.015",
      "--max-file-blocks=1",
      "--verbose",
      "--compress=" + CustomCompressionCodec.class.getName(),

      "in", "out", "20101116153015"
    });


    /*
     * Skipped files should have been moved to the output dir.
     */
    verifyHugeFile(homeDir + "/out/huge", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);
    verifyHugeFile(homeDir + "/out/foo/huge1", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);
    verifyHugeFile(homeDir + "/out/foo/huge2", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);

    Counters jobCounters = crush.getJobCounters();

    assertThat(jobCounters.getCounter(MapperCounter.DIRS_FOUND),      equalTo(2L));
    assertThat(jobCounters.getCounter(MapperCounter.DIRS_ELIGIBLE),    equalTo(0L));
    assertThat(jobCounters.getCounter(MapperCounter.DIRS_SKIPPED),    equalTo(2L));
    assertThat(jobCounters.getCounter(MapperCounter.FILES_FOUND),      equalTo(3L));
View Full Code Here
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
  @Test
  public void executeNoFiles() throws Exception {
    FileSystem.get(job).mkdirs(new Path("in/foo"));
    FileSystem.get(job).mkdirs(new Path("in/hello/world"));

    Crush crush = new Crush();

    ToolRunner.run(job, crush, new String [] {
      "--threshold=0.015",
      "--max-file-blocks=1",
      "--verbose",
      "--compress=" + CustomCompressionCodec.class.getName(),

      "in", "out", "20101116153015"
    });

    Counters jobCounters = crush.getJobCounters();

    assertThat(jobCounters.getCounter(MapperCounter.DIRS_FOUND),      equalTo(4L));
    assertThat(jobCounters.getCounter(MapperCounter.DIRS_ELIGIBLE),    equalTo(0L));
    assertThat(jobCounters.getCounter(MapperCounter.DIRS_SKIPPED),    equalTo(4L));
    assertThat(jobCounters.getCounter(MapperCounter.FILES_FOUND),      equalTo(0L));
View Full Code Here
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
  @Test
  public void executeClone() throws Exception {
    writeFiles(true, true, true);

    Crush crush = new Crush();

    ToolRunner.run(job, crush, new String [] {
      "--threshold=0.015",
      "--max-file-blocks=1",
      "--verbose",

      "--regex=.+/other",
      "--replacement=${crush.timestamp}-${crush.task.num}-middle-${crush.file.num}-tail",
      "--input-format=sequence",
      "--output-format=text",

      "--regex=.+/dir",
      "--replacement=secondregex-${crush.timestamp}-${crush.task.num}-${crush.file.num}",
      "--input-format=text",
      "--output-format=" + SequenceFileOutputFormat.class.getName(),

      "--regex=.+/dir/([^/]+/)*(.+)",
      "--replacement=thirdregex-$2-${crush.timestamp}-${crush.task.num}-${crush.file.num}",
      "--input-format=" + SequenceFileInputFormat.class.getName(),
      "--output-format=sequence",

      "--regex=.+/text",
      "--replacement=fourthregex-${crush.task.num}-${crush.timestamp}-${crush.file.num}",
      "--input-format=" + TextInputFormat.class.getName(),
      "--output-format=" + TextOutputFormat.class.getName(),

      /*
       * This is the default regex and replacement, which we add last so we can exercise the default logic.
       */
      "--regex=.+",
      "--replacement=crushed_file-${crush.timestamp}-${crush.task.num}-${crush.file.num}",
      "--input-format=" + SequenceFileInputFormat.class.getName(),
      "--output-format=" + TextOutputFormat.class.getName(),

      "--compress=" + CustomCompressionCodec.class.getName(),

      "--clone",

      "in", "out", "20101116153015"
    });


    /*
     * Crushed files.
     */
    verifyOutput(homeDir + "/in/dir",                    "secondregex-20101116153015-*-*",            Format.TEXT,      Format.SEQUENCE,  customCodec, "file10", "file11", "file12", "file13");
    verifyOutput(homeDir + "/in/dir/subdir",            "thirdregex-subdir-20101116153015-*-*",      Format.SEQUENCE,  Format.SEQUENCE,  customCodec, "file20", "file21", "file22", "file23", "file24");
    verifyOutput(homeDir + "/in/dir/subdir/subsubdir",  "thirdregex-subsubdir-20101116153015-*-*",  Format.SEQUENCE,  Format.SEQUENCE,  customCodec, "file30", "file31", "file32", "file33", "file34");
    verifyOutput(homeDir + "/in/dir/subdir/other",      "20101116153015-*-middle-*-tail",            Format.SEQUENCE,  Format.TEXT,       customCodec, "file40", "file41", "file42", "file43");
    verifyOutput(homeDir + "/in/dir/other",              "20101116153015-*-middle-*-tail",            Format.SEQUENCE,  Format.TEXT,      customCodec, "file50", "file51", "file52", "file53", "file54", "file55");
    verifyOutput(homeDir + "/in/text",                  "fourthregex-*-20101116153015-*",            Format.TEXT,      Format.TEXT,      customCodec, "file60", "file61", "file62", "file63");
    verifyOutput(homeDir + "/in",                        "crushed_file-20101116153015-*-*",          Format.SEQUENCE,  Format.TEXT,      customCodec, "file70", "file71", "file72");


    /*
     * Skipped files should remain in the input dir.
     */
    verifyOutput(homeDir + "/in/dir/skipped", "file80", Format.SEQUENCE, Format.SEQUENCE, defaultCodec, "file80");
    verifyHugeFile(homeDir + "/in/huge", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);


    /*
     * Crush input files should be moved to the clone dir.
     */
    for (String file : new String[] { "file10", "file11", "file12", "file13" }) {
      verifyOutput(homeDir + "/out" + homeDir + "/in/dir", file, Format.TEXT, Format.TEXT, null, file);
    }

    for (String file : new String[] { "file20", "file21", "file22", "file23", "file24" }) {
      verifyOutput(homeDir + "/out" + homeDir + "/in/dir/subdir", file, Format.SEQUENCE, Format.SEQUENCE, defaultCodec, file);
    }

    for (String file : new String[] { "file30", "file31", "file32", "file33", "file34" }) {
      verifyOutput(homeDir + "/out" + homeDir + "/in/dir/subdir/subsubdir", file, Format.SEQUENCE, Format.SEQUENCE, defaultCodec, file);
    }

    for (String file : new String[] { "file40", "file41", "file42", "file43" }) {
      verifyOutput(homeDir + "/out" + homeDir + "/in/dir/subdir/other", file, Format.SEQUENCE, Format.SEQUENCE, defaultCodec, file);
    }

    for (String file : new String[] { "file50", "file51", "file52", "file53", "file54", "file55" }) {
      verifyOutput(homeDir + "/out" + homeDir + "/in/dir/other", file, Format.SEQUENCE, Format.SEQUENCE, defaultCodec, file);
    }

    for (String file : new String[] { "file60", "file61", "file62", "file63" }) {
      verifyOutput(homeDir + "/out" + homeDir + "/in/text", file, Format.TEXT, Format.TEXT, null, file);
    }

    for (String file : new String[] { "file70", "file71", "file72" }) {
      verifyOutput(homeDir + "/out" + homeDir + "/in", file, Format.SEQUENCE, Format.SEQUENCE, defaultCodec, file);
    }


    Counters jobCounters = crush.getJobCounters();

    assertThat(jobCounters.getCounter(MapperCounter.DIRS_FOUND),      equalTo( 8L));
    assertThat(jobCounters.getCounter(MapperCounter.DIRS_ELIGIBLE),    equalTo( 7L));
    assertThat(jobCounters.getCounter(MapperCounter.DIRS_SKIPPED),    equalTo( 1L));
    assertThat(jobCounters.getCounter(MapperCounter.FILES_FOUND),      equalTo(33L));
View Full Code Here
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
  public void executeCloneHugeFilesOnly() throws Exception {
    writeHugeFile("in/huge", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);
    writeHugeFile("in/foo/huge1", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);
    writeHugeFile("in/foo/huge2", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);

    Crush crush = new Crush();

    ToolRunner.run(job, crush, new String [] {
      "--threshold=0.015",
      "--max-file-blocks=1",
      "--verbose",
      "--compress=" + CustomCompressionCodec.class.getName(),
      "--clone",

      "in", "out", "20101116153015"
    });


    /*
     * Skipped files should remain in the input dir.
     */
    verifyHugeFile(homeDir + "/in/huge", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);
    verifyHugeFile(homeDir + "/in/foo/huge1", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);
    verifyHugeFile(homeDir + "/in/foo/huge2", (long) (((float) 0.015) * 1024 * 1024 * 64) + 1);

    Counters jobCounters = crush.getJobCounters();

    assertThat(jobCounters.getCounter(MapperCounter.DIRS_FOUND),      equalTo(2L));
    assertThat(jobCounters.getCounter(MapperCounter.DIRS_ELIGIBLE),    equalTo(0L));
    assertThat(jobCounters.getCounter(MapperCounter.DIRS_SKIPPED),    equalTo(2L));
    assertThat(jobCounters.getCounter(MapperCounter.FILES_FOUND),      equalTo(3L));
View Full Code Here
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
  @Test
  public void executeCloneNoFiles() throws Exception {
    FileSystem.get(job).mkdirs(new Path("in/foo"));
    FileSystem.get(job).mkdirs(new Path("in/hello/world"));

    Crush crush = new Crush();

    ToolRunner.run(job, crush, new String [] {
      "--threshold=0.015",
      "--max-file-blocks=1",
      "--verbose",
      "--compress=" + CustomCompressionCodec.class.getName(),
      "--clone",

      "in", "out", "20101116153015"
    });

    Counters jobCounters = crush.getJobCounters();

    assertThat(jobCounters.getCounter(MapperCounter.DIRS_FOUND),      equalTo(4L));
    assertThat(jobCounters.getCounter(MapperCounter.DIRS_ELIGIBLE),    equalTo(0L));
    assertThat(jobCounters.getCounter(MapperCounter.DIRS_SKIPPED),    equalTo(4L));
    assertThat(jobCounters.getCounter(MapperCounter.FILES_FOUND),      equalTo(0L));
View Full Code Here
TOP

Related Classes of com.m6d.filecrush.crush.Crush

  • com.m6d.filecrush.crush.CrushOptionParsingTest
  • com.m6d.filecrush.crush.CrushStandAloneSequenceFileTest
  • com.m6d.filecrush.crush.CrushStandAloneTextTest
  • com.m6d.filecrush.crush.CrushTest
  • com.m6d.filecrush.crush.integration.CrushMapReduceTest
  • org.apache.commons.cli.CommandLine
  • org.apache.commons.cli.GnuParser
  • org.apache.commons.cli.Option
  • org.apache.commons.cli.Options
  • org.apache.hadoop.fs.FileStatus
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.