https://blog.csdn.net/ZMC921/article/details/74948786 由于工作需要,我用scala实现在已将有的目录下面继续写入文件。需要重写MultipleTextOutputFormat这个类,具体的请看下面代码,需要交流可以联系我 import java.text.SimpleDateFormat import java.util.Date import org.apache.hadoop.fs.{FileSystem, Path} import org.apache.hadoop.mapred.lib.MultipleTextOutputFormat import org.apache.hadoop.mapred.{InvalidJobConfException, JobConf} import org.apache.hadoop.mapreduce.security.TokenCache import org.apache.spark.{HashPartitioner, SparkConf, SparkContext} import org.apache.spark.rdd.RDD; /** * 在HDFS目录下继续追加文件,不会覆盖以前的文件 * Created by csw on 2017/6/23. */ object MultipleTextOutput { def main(args: Array[String]) { val filePath = "hdfs://master:9000/csw/tmp/data"; val savePath = "hdfs://master:9000/hxzj/mydata/tatol"; val conf = new SparkConf().setAppName("Spark shell") val sc = new SparkContext(conf) //读取文件后,不进行split操作,直接将整行内容看作key, val rdd: RDD[(String, String)] = sc.textFile(filePath).map(x => (x, "")) //rdd必须是(key,value)形式的 RDD.rddToPairRDDFunctions(rdd).partitionBy(new HashPartitioner(1)).saveAsHadoopFile(savePath, classOf[String], classOf[String], classOf[RDDMultipleTextOutputFormat]) sc.stop() } /** * 自定义一个输出文件类 */ case class RDDMultipleTextOutputFormat() extends MultipleTextOutputFormat[Any, Any] { val currentTime: Date = new Date() val formatter = new SimpleDateFormat("yyyy-MM-dd-HHmmss"); val dateString = formatter.format(currentTime); //自定义保存文件名 override def generateFileNameForKeyValue(key: Any, value: Any, name: String): String = { //key 和 value就是rdd中的(key,value),name是part-00000默认的文件名 //保存的文件名称,这里用字符串拼接系统生成的时间戳来区分文件名,可以自己定义 "HTLXYFY" + dateString } override def checkOutputSpecs(ignored: FileSystem, job: JobConf): Unit = { val name: String = job.get(org.apache.hadoop.mapreduce.lib.output.FileOutputFormat.OUTDIR) var outDir: Path = if (name == null) null else new Path(name) //当输出任务不等于0 且输出的路径为空,则抛出异常 if (outDir == null && job.getNumReduceTasks != 0) { throw new InvalidJobConfException("Output directory not set in JobConf.") } //当有输出任务和输出路径不为null时 if (outDir != null) { val fs: FileSystem = outDir.getFileSystem(job) outDir = fs.makeQualified(outDir) outDir = new Path(job.getWorkingDirectory, outDir) job.set(org.apache.hadoop.mapreduce.lib.output.FileOutputFormat.OUTDIR, outDir.toString) TokenCache.obtainTokensForNamenodes(job.getCredentials, Array[Path](outDir), job) //下面的注释掉,就不会出现这个目录已经存在的提示了 /* if (fs.exists(outDir)) { throw new FileAlreadyExistsException("Outputdirectory" + outDir + "alreadyexists"); } }*/ } } } } --------------------- 原文链接:https://blog.csdn.net/ZMC921/article/details/74948786 |
|
来自: jasonbetter > 《Spark_Flink》