They are two different RDDs. Spark doesn't guarantee that the first
partition of RDD1 and the first partition of RDD2 will stay in the
same worker node. If that is the case, if you have 1000
single-partition RDDs the first worker will have very heavy load.
-Xiangrui
On Thu, Aug 7, 2014 at 2:20 AM, losmi83 <milos.nikoli...@gmail.com> wrote:
> Hi guys,
>
> the latest Spark version 1.0.2 exhibits a very strange behavior when it
> comes to deciding on which node a given partition should reside. The
> following example was tested in the standalone Spark mode.
>
> val partitioner = new HashPartitioner(10)
>
> val dummyJob1 = sc.parallelize(0 until 10).map(x =>
> (x,x)).partitionBy(partitioner)
> dummyJob1.foreach { case (id, x) => println("Dummy1 -> Id = " + id) }
>
> val dummyJob2 = sc.parallelize(0 until 10).map(x =>
> (x,x)).partitionBy(partitioner)
> dummyJob2.foreach { case (id, x) => println("Dummy2 -> Id = " + id) }
>
> On one node I get something like:
> Dummy1 -> Id = 2
> Dummy2 -> Id = 7
>
> This is a very strange behavior... One would expect that partitions with the
> same ID are placed on the same node -- that is exactly happening with Spark
> 0.9.2 (and below), but not the latest versions (tested with 1.0.1 and
> 1.0.2). Can anyone explain this?
>
> Thanks in advance,
> Milos
>
>
>
>
> --
> View this message in context:
> http://apache-spark-user-list.1001560.n3.nabble.com/Where-do-my-partitions-go-tp11635.html
> Sent from the Apache Spark User List mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscr...@spark.apache.org
> For additional commands, e-mail: user-h...@spark.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscr...@spark.apache.org
For additional commands, e-mail: user-h...@spark.apache.org
