Not really.  In practice I write everything out to HDFS and that is working fine.  But I write lots of unit tests and example scripts and it is convenient to be able to test a Spark application (or sequence of spark functions) in a very local way such that it doesn't depend on any outside infrastructure (e.g. an HDFS server.)  So, it is convenient to write out a small amount of data locally and manually inspect the results - esp. as I'm building up a unit or regression test.

So, ultimately writing results out to a local file isn't that important to me.  However, I was just trying to run a simple example script that worked before and is now not working. 


On 1/2/2014 10:28 AM, Andrew Ash wrote:
You want to write it to a local file on the machine?  Try using "file:///path/to/target/mydir/" instead

I'm not sure what behavior would be if you did this on a multi-machine cluster though -- you may get a bit of data on each machine in that local directory.

On Thu, Jan 2, 2014 at 12:22 PM, Philip Ogren <> wrote:
I have a very simple Spark application that looks like the following:

var myRdd: RDD[Array[String]] = initMyRdd()
println(myRdd.first.mkString(", "))


The println statements work as expected.  The first saveAsTextFile statement also works as expected.  The second saveAsTextFile statement does not (even if the first is commented out.)  I get the exception pasted below.  If I inspect "target/mydir" I see that there is a directory called _temporary/0/_temporary/attempt_201401020953_0000_m_000000_1 which contains an empty part-00000 file.  It's curious because this code worked before with Spark 0.8.0 and now I am running on Spark 0.8.1. I happen to be running this on Windows in "local" mode at the moment.  Perhaps I should try running it on my linux box.


Exception in thread "main" org.apache.spark.SparkException: Job aborted: Task 2.0:0 failed more than 0 times; aborting job java.lang.NullPointerException
    at org.apache.spark.scheduler.DAGScheduler$$anonfun$abortStage$1.apply(DAGScheduler.scala:827)
    at org.apache.spark.scheduler.DAGScheduler$$anonfun$abortStage$1.apply(DAGScheduler.scala:825)
    at scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:60)
    at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:47)
    at org.apache.spark.scheduler.DAGScheduler.abortStage(DAGScheduler.scala:825)
    at org.apache.spark.scheduler.DAGScheduler.processEvent(DAGScheduler.scala:440)
    at org.apache.spark.scheduler.DAGScheduler$$anon$