如何利用spark向HDFS的目录中追加数据?

       我们在利用spark处理存储在HDFS上的数据时经常会有这样的需求，需要不断的向同一个目录写入数据（比如，每个小时将kafka中的数据落到HDFS的同一个目录），当然这种需求有很多解决方案可以使用，今天我们所探讨的就是如何通过修改spark 数据输出组件来实现这个功能，

1.1 SparkContext 这个类中、提供了多种读取HDFS上文件的API，如下代码所示: 

/**
   * Read a text file from HDFS, a local file system (available on all nodes), or any
   * Hadoop-supported file system URI, and return it as an RDD of Strings.
   * @param path path to the text file on a supported file system
   * @param minPartitions suggested minimum number of partitions for the resulting RDD
   * @return RDD of lines of the text file
   */
  def textFile(
      path: String,
      minPartitions: Int = defaultMinPartitions): RDD[String] = withScope {
    assertNotStopped()
    hadoopFile(path, classOf[TextInputFormat], classOf[LongWritable], classOf[Text],
      minPartitions).map(pair => pair._2.toString).setName(path)
  }

  /**
   * Read a directory of text files from HDFS, a local file system (available on all nodes), or any
   * Hadoop-supported file system URI. Each file is read as a single record and returned in a
   * key-value pair, where the key is the path of each file, the value is the content of each file.
   *
   * <p> For example, if you have the following files:
   * {
  {
  {
   *   hdfs://a-hdfs-path/part-00000
   *   hdfs://a-hdfs-path/part-00001
   *   ...
   *   hdfs://a-hdfs-path/part-nnnnn
   * }}}
   *
   * Do `val rdd = sparkContext.wholeTextFile("hdfs://a-hdfs-path")`,
   *
   * <p> then `rdd` contains
   * {
  {
  {
   *   (a-hdfs-path/part-00000, its content)
   *   (a-hdfs-path/part-00001, its content)
   *   ...
   *   (a-hdfs-path/part-nnnnn, its content)
   * }}}
   *
   * @note Small files are preferred, large file is also allowable, but may cause bad performance.
   * @note On some filesystems, `.../path/&#42;` can be a more efficient way to read all files
   *       in a directory rather than `.../path/` or `.../path`
   * @note Partitioning is determined by data locality. This may result in too few partitions
   *       by default.
   *
   * @param path Directory to the input data files, the path can be comma separated paths as the
   *             list of inputs.
   * @param minPartitions A suggestion value of the minimal splitting number for input data.
   * @return RDD representing tuples of file path and the corresponding file content
   */
  def wholeTextFiles(
      path: String,
      minPartitions: Int = defaultMinPartitions): RDD[(String, String)] = withScope {
    assertNotStopped()
    val job = NewHadoopJob.getInstance(hadoopConfiguration)
    // Use setInputPaths so that wholeTextFiles aligns with hadoopFile/textFile in taking
    // comma separated files as input. (see SPARK-7155)
    NewFileInputFormat.setInputPaths(job, path)
    val updateConf = job.getConfiguration
    new WholeTextFileRDD(
      this,
      classOf[WholeTextFileInputFormat],
      classOf[Text],
      classOf[Text],
      updateConf,
      minPartitions).map(record => (record._1.toString, record._2.toString)).setName(path)
  }

  /**
   * Get an RDD for a Hadoop-readable dataset as PortableDataStream for each file
   * (useful for binary data)
   *
   * For example, if you have the following files:
   * {
  {
  {
   *   hdfs://a-hdfs-path/part-00000
   *   hdfs://a-hdfs-path/part-00001
   *   ...
   *   hdfs://a-hdfs-path/part-nnnnn
   * }}}
   *
   * Do
   * `val rdd = sparkContext.binaryFiles("hdfs://a-hdfs-path")`,
   *
   * then `rdd` contains
   * {
  {
  {
   *   (a-hdfs-path/part-00000, its content)
   *   (a-hdfs-path/part-00001, its content)
   *   ...
   *   (a-hdfs-path/part-nnnnn, its content)
   * }}}
   *
   * @note Small files are preferred; very large files may cause bad performance.
   * @note On some filesystems, `.../path/&#42;` can be a more efficient way to read all files
   *       in a directory rather than `.../path/` or `.../path`
   * @note Partitioning is determined by data locality. This may result in too few partitions
   *       by default.
   *
   * @param path Directory to the input data files, the path can be comma separated paths as the
   *             list of inputs.
   * @param minPartitions A suggestion value of the minimal splitting number for input data.
   * @return RDD representing tuples of file path and corresponding file content
   */
  def binaryFiles(
      path: String,
      minPartitions: Int = defaultMinPartitions): RDD[(String, PortableDataStream)] = withScope {
    assertNotStopped()
    val job = NewHadoopJob.getInstance(hadoopConfiguration)
    // Use setInputPaths so that binaryFiles aligns with hadoopFile/textFile in taking
    // comma separated files as input. (see SPARK-7155)
    NewFileInputFormat.setInputPaths(job, path)
    val updateConf = job.getConfiguration
    new BinaryFileRDD(
      this,
      classOf[StreamInputFormat],
      classOf[String],
      classOf[PortableDataStream],
      updateConf,
      minPartitions).setName(path)
  }

  /**
   * Load data from a flat binary file, assuming the length of each record is constant.
   *
   * @note We ensure that the byte array for each record in the resulting RDD
   * has the provided record length.
   *
   * @param path Directory to the input data files, the path can be comma separated paths as the
   *             list of inputs.
   * @param recordLength The length at which to split the records
   * @param conf Configuration for setting up the dataset.
   *
   * @return An RDD of data with values, represented as byte arrays
   */
  def binaryRecords(
      path: String,
      recordLength: Int,
      conf: Configuration = hadoopConfiguration): RDD[Array[Byte]] = withScope {
    assertNotStopped()
    conf.setInt(FixedLengthBinaryInputFormat.RECORD_LENGTH_PROPERTY, recordLength)
    val br = newAPIHadoopFile[LongWritable, BytesWritable, FixedLengthBinaryInputFormat](path,
      classOf[FixedLengthBinaryInputFormat],
      classOf[LongWritable],
      classOf[BytesWritable],
      conf = conf)
    br.map { case (k, v) =>
      val bytes = v.copyBytes()
      assert(bytes.length == recordLength, "Byte array does not have correct length")
      bytes
    }
  }

  /**
   * Get an RDD for a Hadoop-readable dataset from a Hadoop JobConf given its InputFormat and other
   * necessary info (e.g. file name for a filesystem-based dataset, table name for HyperTable),
   * using the older MapReduce API (`org.apache.hadoop.mapred`).
   *
   * @param conf JobConf for setting up the dataset. Note: This will be put into a Broadcast.
   *             Therefore if you plan to reuse this conf to create multiple RDDs, you need to make
   *             sure you won't modify the conf. A safe approach is always creating a new conf for
   *             a new RDD.
   * @param inputFormatClass storage format of the data to be read
   * @param keyClass `Class` of the key associated with the `inputFormatClass` parameter
   * @param valueClass `Class` of the value associated with the `inputFormatClass` parameter
   * @param minPartitions Minimum number of Hadoop Splits to generate.
   * @return RDD of tuples of key and corresponding value
   *
   * @note Because Hadoop's RecordReader class re-uses the same Writable object for each
   * record, directly caching the returned RDD or directly passing it to an aggregation or shuffle
   * operation will create many references to the same object.
   * If you plan to directly cache, sort, or aggregate Hadoop writable objects, you should first
   * copy them using a `map` function.
   */
  def hadoopRDD[K, V](
      conf: JobConf,
      inputFormatClass: Class[_ <: InputFormat[K, V]],
      keyClass: Class[K],
      valueClass: Class[V],
      minPartitions: Int = defaultMinPartitions): RDD[(K, V)] = withScope {
    assertNotStopped()

    // This is a hack to enforce loading hdfs-site.xml.
    // See SPARK-11227 for details.
    FileSystem.getLocal(conf)

    // Add necessary security credentials to the JobConf before broadcasting it.
    SparkHadoopUtil.get.addCredentials(conf)
    new HadoopRDD(this, conf, inputFormatClass, keyClass, valueClass, minPartitions)
  }

  /** Get an RDD for a Hadoop file with an arbitrary InputFormat
   *
   * @note Because Hadoop's RecordReader class re-uses the same Writable object for each
   * record, directly caching the returned RDD or directly passing it to an aggregation or shuffle
   * operation will create many references to the same object.
   * If you plan to directly cache, sort, or aggregate Hadoop writable objects, you should first
   * copy them using a `map` function.
   * @param path directory to the input data files, the path can be comma separated paths
   * as a list of inputs
   * @param inputFormatClass storage format of the data to be read
   * @param keyClass `Class` of the key associated with the `inputFormatClass` parameter
   * @param valueClass `Class` of the value associated with the `inputFormatClass` parameter
   * @param minPartitions suggested minimum number of partitions for the resulting RDD
   * @return RDD of tuples of key and corresponding value
   */
  def hadoopFile[K, V](
      path: String,
      inputFormatClass: Class[_ <: InputFormat[K, V]],
      keyClass: Class[K],
      valueClass: Class[V],
      minPartitions: Int = defaultMinPartitions): RDD[(K, V)] = withScope {
    assertNotStopped()

    // This is a hack to enforce loading hdfs-site.xml.
    // See SPARK-11227 for details.
    FileSystem.getLocal(hadoopConfiguration)

    // A Hadoop configuration can be about 10 KB, which is p