diff --git a/src/scala/spark/HdfsFile.scala b/src/scala/spark/HdfsFile.scala index 595386fceb07a179fc7149ac43793ef2cff75017..886272a8edd14131f096b3ef2a233a901876db1c 100644 --- a/src/scala/spark/HdfsFile.scala +++ b/src/scala/spark/HdfsFile.scala @@ -14,6 +14,7 @@ import org.apache.hadoop.mapred.Reporter @serializable class HdfsSplit(@transient s: InputSplit) extends Split { val inputSplit = new SerializableWritable[InputSplit](s) + override def toString = inputSplit.toString } class HdfsTextFile(sc: SparkContext, path: String) diff --git a/src/scala/spark/RDD.scala b/src/scala/spark/RDD.scala index 4d0c8c671160763127f7fab9a2f52a6f2bb2d712..181f7e8b030fe08eed5757a51e48088cb0f813f1 100644 --- a/src/scala/spark/RDD.scala +++ b/src/scala/spark/RDD.scala @@ -198,6 +198,7 @@ extends RDD[T](prev.sparkContext) with Logging { override def iterator(split: Split): Iterator[T] = { val key = id + "::" + split.toString + logInfo("CachedRDD split key is " + key) val cache = CachedRDD.cache val loading = CachedRDD.loading val cachedVal = cache.get(key)