org.apache.avro.mapreduce.AvroKeyInputFormat Java Examples

The following examples show how to use org.apache.avro.mapreduce.AvroKeyInputFormat. You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may check out the related API usage on the sidebar.
Example #1
Source File: HiveIncrPullSource.java    From hudi with Apache License 2.0 6 votes vote down vote up 
@Override
protected InputBatch<JavaRDD<GenericRecord>> fetchNewData(Option<String> lastCheckpointStr, long sourceLimit) {
  try {
    // find the source commit to pull
    Option<String> commitToPull = findCommitToPull(lastCheckpointStr);

    if (!commitToPull.isPresent()) {
      return new InputBatch<>(Option.empty(), lastCheckpointStr.isPresent() ? lastCheckpointStr.get() : "");
    }

    // read the files out.
    List<FileStatus> commitDeltaFiles = Arrays.asList(fs.listStatus(new Path(incrPullRootPath, commitToPull.get())));
    String pathStr = commitDeltaFiles.stream().map(f -> f.getPath().toString()).collect(Collectors.joining(","));
    JavaPairRDD<AvroKey, NullWritable> avroRDD = sparkContext.newAPIHadoopFile(pathStr, AvroKeyInputFormat.class,
        AvroKey.class, NullWritable.class, sparkContext.hadoopConfiguration());
    return new InputBatch<>(Option.of(avroRDD.keys().map(r -> ((GenericRecord) r.datum()))),
        String.valueOf(commitToPull.get()));
  } catch (IOException ioe) {
    throw new HoodieIOException("Unable to read from source from checkpoint: " + lastCheckpointStr, ioe);
  }
}
Example #2
Source File: AvroHdfsFileSource.java    From components with Apache License 2.0 5 votes vote down vote up 
private AvroHdfsFileSource(UgiDoAs doAs, String filepattern, LazyAvroCoder<?> lac, ExtraHadoopConfiguration extraConfig,
        SerializableSplit serializableSplit) {
    super(doAs, filepattern, (Class) AvroKeyInputFormat.class, AvroKey.class, NullWritable.class, extraConfig,
            serializableSplit);
    this.lac = lac;
    setDefaultCoder(LazyAvroKeyWrapper.of(lac), WritableCoder.of(NullWritable.class));
}
Example #3
Source File: AvroDFSSource.java    From hudi with Apache License 2.0 4 votes vote down vote up 
private JavaRDD<GenericRecord> fromFiles(String pathStr) {
  JavaPairRDD<AvroKey, NullWritable> avroRDD = sparkContext.newAPIHadoopFile(pathStr, AvroKeyInputFormat.class,
      AvroKey.class, NullWritable.class, sparkContext.hadoopConfiguration());
  return avroRDD.keys().map(r -> ((GenericRecord) r.datum()));
}
Example #4
Source File: CombineAvroKeyInputFormat.java    From incubator-pinot with Apache License 2.0 4 votes vote down vote up 
public AvroKeyRecordReaderWrapper(CombineFileSplit split, TaskAttemptContext context, Integer index)
    throws IOException, InterruptedException {
  super(new AvroKeyInputFormat<>(), split, context, index);
}
Example #5
Source File: AggregationPhaseJob.java    From incubator-pinot with Apache License 2.0 4 votes vote down vote up 
public Job run() throws Exception {
  Job job = Job.getInstance(getConf());
  job.setJobName(name);
  job.setJarByClass(AggregationPhaseJob.class);

  FileSystem fs = FileSystem.get(getConf());
  Configuration configuration = job.getConfiguration();

  // Properties
  LOGGER.info("Properties {}", props);

   // Input Path
  String inputPathDir = getAndSetConfiguration(configuration, AGG_PHASE_INPUT_PATH);
  LOGGER.info("Input path dir: " + inputPathDir);
  for (String inputPath : inputPathDir.split(ThirdEyeConstants.FIELD_SEPARATOR)) {
    LOGGER.info("Adding input:" + inputPath);
    Path input = new Path(inputPath);
    FileInputFormat.addInputPath(job, input);
  }

  // Output path
  Path outputPath = new Path(getAndSetConfiguration(configuration, AGG_PHASE_OUTPUT_PATH));
  LOGGER.info("Output path dir: " + outputPath.toString());
  if (fs.exists(outputPath)) {
    fs.delete(outputPath, true);
  }
  FileOutputFormat.setOutputPath(job, outputPath);

  // Schema
  Schema avroSchema = ThirdeyeAvroUtils.getSchema(inputPathDir);
  LOGGER.info("Schema : {}", avroSchema.toString(true));
  job.getConfiguration().set(AGG_PHASE_AVRO_SCHEMA.toString(), avroSchema.toString());

  // ThirdEyeConfig
  String dimensionTypesProperty = ThirdeyeAvroUtils.getDimensionTypesProperty(
      props.getProperty(ThirdEyeConfigProperties.THIRDEYE_DIMENSION_NAMES.toString()), avroSchema);
  props.setProperty(ThirdEyeConfigProperties.THIRDEYE_DIMENSION_TYPES.toString(), dimensionTypesProperty);
  String metricTypesProperty = ThirdeyeAvroUtils.getMetricTypesProperty(
      props.getProperty(ThirdEyeConfigProperties.THIRDEYE_METRIC_NAMES.toString()),
      props.getProperty(ThirdEyeConfigProperties.THIRDEYE_METRIC_TYPES.toString()), avroSchema);
  props.setProperty(ThirdEyeConfigProperties.THIRDEYE_METRIC_TYPES.toString(), metricTypesProperty);
  ThirdEyeConfig thirdeyeConfig = ThirdEyeConfig.fromProperties(props);
  LOGGER.info("Thirdeye Config {}", thirdeyeConfig.encode());
  job.getConfiguration().set(AGG_PHASE_THIRDEYE_CONFIG.toString(), OBJECT_MAPPER.writeValueAsString(thirdeyeConfig));

  // Map config
  job.setMapperClass(AggregationMapper.class);
  job.setInputFormatClass(AvroKeyInputFormat.class);
  job.setMapOutputKeyClass(BytesWritable.class);
  job.setMapOutputValueClass(BytesWritable.class);

  // Reduce config
  job.setReducerClass(AggregationReducer.class);
  job.setOutputKeyClass(AvroKey.class);
  job.setOutputValueClass(NullWritable.class);
  AvroJob.setOutputKeySchema(job, avroSchema);
  job.setOutputFormatClass(AvroKeyOutputFormat.class);
  String numReducers = props.getProperty(ThirdEyeJobProperties.THIRDEYE_NUM_REDUCERS.getName());
  LOGGER.info("Num Reducers : {}", numReducers);
  if (StringUtils.isNotBlank(numReducers)) {
    job.setNumReduceTasks(Integer.valueOf(numReducers));
    LOGGER.info("Setting num reducers {}", job.getNumReduceTasks());
  }

  job.waitForCompletion(true);

  Counter counter = job.getCounters().findCounter(AggregationCounter.NUMBER_OF_RECORDS);
  LOGGER.info(counter.getDisplayName() + " : " + counter.getValue());
  if (counter.getValue() == 0) {
    throw new IllegalStateException("No input records in " + inputPathDir);
  }
  counter = job.getCounters().findCounter(AggregationCounter.NUMBER_OF_RECORDS_FLATTENED);
  LOGGER.info(counter.getDisplayName() + " : " + counter.getValue());

  for (String metric : thirdeyeConfig.getMetricNames()) {
    counter = job.getCounters().findCounter(thirdeyeConfig.getCollection(), metric);
    LOGGER.info(counter.getDisplayName() + " : " + counter.getValue());
  }

  return job;
}
Example #6
Source File: TopKPhaseJob.java    From incubator-pinot with Apache License 2.0 4 votes vote down vote up 
public Job run() throws Exception {
  Job job = Job.getInstance(getConf());
  job.setJobName(name);
  job.setJarByClass(TopKPhaseJob.class);

  Configuration configuration = job.getConfiguration();
  FileSystem fs = FileSystem.get(configuration);

  // Properties
  LOGGER.info("Properties {}", props);

   // Input Path
  String inputPathDir = getAndSetConfiguration(configuration, TOPK_PHASE_INPUT_PATH);
  LOGGER.info("Input path dir: " + inputPathDir);
  for (String inputPath : inputPathDir.split(ThirdEyeConstants.FIELD_SEPARATOR)) {
    LOGGER.info("Adding input:" + inputPath);
    Path input = new Path(inputPath);
    FileInputFormat.addInputPath(job, input);
  }

  // Output path
  Path outputPath = new Path(getAndSetConfiguration(configuration, TOPK_PHASE_OUTPUT_PATH));
  LOGGER.info("Output path dir: " + outputPath.toString());
  if (fs.exists(outputPath)) {
    fs.delete(outputPath, true);
  }
  FileOutputFormat.setOutputPath(job, outputPath);

  // Schema
  Schema avroSchema = ThirdeyeAvroUtils.getSchema(inputPathDir);
  LOGGER.info("Schema : {}", avroSchema.toString(true));

  // ThirdEyeConfig
  String dimensionTypesProperty = ThirdeyeAvroUtils.getDimensionTypesProperty(
      props.getProperty(ThirdEyeConfigProperties.THIRDEYE_DIMENSION_NAMES.toString()), avroSchema);
  props.setProperty(ThirdEyeConfigProperties.THIRDEYE_DIMENSION_TYPES.toString(), dimensionTypesProperty);
  String metricTypesProperty = ThirdeyeAvroUtils.getMetricTypesProperty(
      props.getProperty(ThirdEyeConfigProperties.THIRDEYE_METRIC_NAMES.toString()),
      props.getProperty(ThirdEyeConfigProperties.THIRDEYE_METRIC_TYPES.toString()), avroSchema);
  props.setProperty(ThirdEyeConfigProperties.THIRDEYE_METRIC_TYPES.toString(), metricTypesProperty);
  ThirdEyeConfig thirdeyeConfig = ThirdEyeConfig.fromProperties(props);
  LOGGER.info("Thirdeye Config {}", thirdeyeConfig.encode());
  job.getConfiguration().set(TOPK_PHASE_THIRDEYE_CONFIG.toString(), OBJECT_MAPPER.writeValueAsString(thirdeyeConfig));

  // Map config
  job.setMapperClass(TopKPhaseMapper.class);
  job.setInputFormatClass(AvroKeyInputFormat.class);
  job.setMapOutputKeyClass(BytesWritable.class);
  job.setMapOutputValueClass(BytesWritable.class);

  // Combiner
  job.setCombinerClass(TopKPhaseCombiner.class);

   // Reduce config
  job.setReducerClass(TopKPhaseReducer.class);
  job.setOutputKeyClass(NullWritable.class);
  job.setOutputValueClass(NullWritable.class);
  job.setNumReduceTasks(1);

  job.waitForCompletion(true);

  return job;
}
Example #7
Source File: DerivedColumnTransformationPhaseJob.java    From incubator-pinot with Apache License 2.0 4 votes vote down vote up 
public Job run() throws Exception {
  Job job = Job.getInstance(getConf());
  job.setJobName(name);
  job.setJarByClass(DerivedColumnTransformationPhaseJob.class);

  Configuration configuration = job.getConfiguration();
  FileSystem fs = FileSystem.get(configuration);

  // Input Path
  String inputPathDir = getAndSetConfiguration(configuration, DERIVED_COLUMN_TRANSFORMATION_PHASE_INPUT_PATH);
  LOGGER.info("Input path dir: " + inputPathDir);
  for (String inputPath : inputPathDir.split(",")) {
    LOGGER.info("Adding input:" + inputPath);
    Path input = new Path(inputPath);
    FileInputFormat.addInputPath(job, input);
  }

  // Topk path
  String topkPath = getAndSetConfiguration(configuration, DERIVED_COLUMN_TRANSFORMATION_PHASE_TOPK_PATH);
  LOGGER.info("Topk path : " + topkPath);

  // Output path
  Path outputPath = new Path(getAndSetConfiguration(configuration, DERIVED_COLUMN_TRANSFORMATION_PHASE_OUTPUT_PATH));
  LOGGER.info("Output path dir: " + outputPath.toString());
  if (fs.exists(outputPath)) {
    fs.delete(outputPath, true);
  }
  FileOutputFormat.setOutputPath(job, outputPath);

  // Schema
  Schema avroSchema = ThirdeyeAvroUtils.getSchema(inputPathDir);
  LOGGER.info("Schema : {}", avroSchema.toString(true));

  // ThirdEyeConfig
  String dimensionTypesProperty = ThirdeyeAvroUtils.getDimensionTypesProperty(
      props.getProperty(ThirdEyeConfigProperties.THIRDEYE_DIMENSION_NAMES.toString()), avroSchema);
  props.setProperty(ThirdEyeConfigProperties.THIRDEYE_DIMENSION_TYPES.toString(), dimensionTypesProperty);
  String metricTypesProperty = ThirdeyeAvroUtils.getMetricTypesProperty(
      props.getProperty(ThirdEyeConfigProperties.THIRDEYE_METRIC_NAMES.toString()),
      props.getProperty(ThirdEyeConfigProperties.THIRDEYE_METRIC_TYPES.toString()), avroSchema);
  props.setProperty(ThirdEyeConfigProperties.THIRDEYE_METRIC_TYPES.toString(), metricTypesProperty);
  ThirdEyeConfig thirdeyeConfig = ThirdEyeConfig.fromProperties(props);
  job.getConfiguration().set(DERIVED_COLUMN_TRANSFORMATION_PHASE_THIRDEYE_CONFIG.toString(),
      OBJECT_MAPPER.writeValueAsString(thirdeyeConfig));
  LOGGER.info("ThirdEyeConfig {}", thirdeyeConfig.encode());

  // New schema
  Schema outputSchema = newSchema(thirdeyeConfig);
  job.getConfiguration().set(DERIVED_COLUMN_TRANSFORMATION_PHASE_OUTPUT_SCHEMA.toString(), outputSchema.toString());

  // Map config
  job.setMapperClass(DerivedColumnTransformationPhaseMapper.class);
  job.setInputFormatClass(AvroKeyInputFormat.class);
  job.setMapOutputKeyClass(AvroKey.class);
  job.setMapOutputValueClass(NullWritable.class);
  AvroJob.setOutputKeySchema(job, outputSchema);
  LazyOutputFormat.setOutputFormatClass(job, AvroKeyOutputFormat.class);
  AvroMultipleOutputs.addNamedOutput(job, "avro", AvroKeyOutputFormat.class, outputSchema);

  job.setNumReduceTasks(0);

  job.waitForCompletion(true);

  return job;
}
Example #8
Source File: OSMRunner.java    From geowave with Apache License 2.0 4 votes vote down vote up 
@Override
public int run(final String[] args) throws Exception {

  final Configuration conf = getConf();
  conf.set("tableName", ingestOptions.getQualifiedTableName());
  conf.set("osmVisibility", ingestOptions.getVisibilityOptions().getVisibility());

  // job settings
  final Job job = Job.getInstance(conf, ingestOptions.getJobName());
  job.setJarByClass(OSMRunner.class);

  switch (ingestOptions.getMapperType()) {
    case "NODE": {
      configureSchema(AvroNode.getClassSchema());
      inputAvroFile = ingestOptions.getNodesBasePath();
      job.setMapperClass(OSMNodeMapper.class);
      break;
    }
    case "WAY": {
      configureSchema(AvroWay.getClassSchema());
      inputAvroFile = ingestOptions.getWaysBasePath();
      job.setMapperClass(OSMWayMapper.class);
      break;
    }
    case "RELATION": {
      configureSchema(AvroRelation.getClassSchema());
      inputAvroFile = ingestOptions.getRelationsBasePath();
      job.setMapperClass(OSMRelationMapper.class);
      break;
    }
    default:
      break;
  }
  if ((avroSchema == null) || (inputAvroFile == null)) {
    throw new MissingArgumentException(
        "argument for mapper type must be one of: NODE, WAY, or RELATION");
  }

  enableLocalityGroups(ingestOptions);

  // input format
  job.setInputFormatClass(AvroKeyInputFormat.class);
  FileInputFormat.setInputPaths(job, inputAvroFile);
  AvroJob.setInputKeySchema(job, avroSchema);

  // mappper

  job.setOutputKeyClass(Text.class);
  job.setOutputValueClass(Mutation.class);
  job.setOutputFormatClass(AccumuloOutputFormat.class);
  AccumuloOutputFormat.setConnectorInfo(
      job,
      accumuloOptions.getUser(),
      new PasswordToken(accumuloOptions.getPassword()));
  AccumuloOutputFormat.setCreateTables(job, true);
  AccumuloOutputFormat.setDefaultTableName(job, ingestOptions.getQualifiedTableName());
  AccumuloOutputFormat.setZooKeeperInstance(
      job,
      new ClientConfiguration().withInstance(accumuloOptions.getInstance()).withZkHosts(
          accumuloOptions.getZookeeper()));

  // reducer
  job.setNumReduceTasks(0);

  return job.waitForCompletion(true) ? 0 : -1;
}
Example #9
Source File: FileSystemViewKeyInputFormat.java    From kite with Apache License 2.0 4 votes vote down vote up 
@Override
FileInputFormat<AvroKey<E>, NullWritable> getInputFormat() {
  return new AvroKeyInputFormat<E>();
}