Passed

org.apache.spark.rdd.PairRDDFunctionsSuite.saveAsHadoopDataset should respect empty output directory when there are no files to be committed to an absolute output location

Took 40 ms.