spark中操作hdfs

1 获取路径

val output = new Path("hdfs://master:9000/output/");  
val hdfs = org.apache.hadoop.fs.FileSystem.get(  
      new java.net.URI("hdfs://master:9000"), new org.apache.hadoop.conf.Configuration())  

2 相关操作

// 删除输出目录  
  if (hdfs.exists(output)) hdfs.delete(output, true)  
//遍历目录
 val fs= hdfs.listStatus(output)
 19  val listPath = FileUtil.stat2Paths(fs)
 20  import org.apache.hadoop.fs.FileUtil
 21  val listPath = FileUtil.stat2Paths(fs)
 22  for(p<-listPath) println(p)

转载于:https://www.cnblogs.com/mayidudu/p/5794786.html