Merge branch 'master' of github.com:apache/spark into yarn-jars

Author: andrewor14

README.md

@@ -39,17 +39,22 @@ And run the following command, which should also return 1000:
 ## Example Programs
 
 Spark also comes with several sample programs in the `examples` directory.
-To run one of them, use `./bin/run-example <class> <params>`. For example:
+To run one of them, use `./bin/run-example <class> [params]`. For example:
 
-    ./bin/run-example org.apache.spark.examples.SparkLR local[2]
+    ./bin/run-example org.apache.spark.examples.SparkLR
 
-will run the Logistic Regression example locally on 2 CPUs.
+will run the Logistic Regression example locally.
 
-Each of the example programs prints usage help if no params are given.
+You can set the MASTER environment variable when running examples to submit
+examples to a cluster. This can be a mesos:// or spark:// URL, 
+"yarn-cluster" or "yarn-client" to run on YARN, and "local" to run 
+locally with one thread, or "local[N]" to run locally with N threads. You 
+can also use an abbreviated class name if the class is in the `examples`
+package. For instance:
 
-All of the Spark samples take a `<master>` parameter that is the cluster URL
-to connect to. This can be a mesos:// or spark:// URL, or "local" to run
-locally with one thread, or "local[N]" to run locally with N threads.
+    MASTER=spark://host:7077 ./bin/run-example SparkPi
+
+Many of the example programs print usage help if no params are given.
 
 ## Running Tests
 

bin/pyspark

@@ -31,7 +31,7 @@ if [ ! -f "$FWDIR/RELEASE" ]; then
   ls "$FWDIR"/assembly/target/scala-$SCALA_VERSION/spark-assembly*hadoop*.jar >& /dev/null
   if [[ $? != 0 ]]; then
     echo "Failed to find Spark assembly in $FWDIR/assembly/target" >&2
-    echo "You need to build Spark with sbt/sbt assembly before running this program" >&2
+    echo "You need to build Spark before running this program" >&2
     exit 1
   fi
 fi

bin/run-example

@@ -17,28 +17,10 @@
 # limitations under the License.
 #
 
-cygwin=false
-case "`uname`" in
-    CYGWIN*) cygwin=true;;
-esac
-
 SCALA_VERSION=2.10
 
-# Figure out where the Scala framework is installed
 FWDIR="$(cd `dirname $0`/..; pwd)"
-
-# Export this as SPARK_HOME
 export SPARK_HOME="$FWDIR"
-
-. $FWDIR/bin/load-spark-env.sh
-
-if [ -z "$1" ]; then
-  echo "Usage: run-example <example-class> [<args>]" >&2
-  exit 1
-fi
-
-# Figure out the JAR file that our examples were packaged into. This includes a bit of a hack
-# to avoid the -sources and -doc packages that are built by publish-local.
 EXAMPLES_DIR="$FWDIR"/examples
 
 if [ -f "$FWDIR/RELEASE" ]; then
@@ -49,46 +31,29 @@ fi
 
 if [[ -z $SPARK_EXAMPLES_JAR ]]; then
   echo "Failed to find Spark examples assembly in $FWDIR/lib or $FWDIR/examples/target" >&2
-  echo "You need to build Spark with sbt/sbt assembly before running this program" >&2
+  echo "You need to build Spark before running this program" >&2
   exit 1
 fi
 
+EXAMPLE_MASTER=${MASTER:-"local[*]"}
 
-# Since the examples JAR ideally shouldn't include spark-core (that dependency should be
-# "provided"), also add our standard Spark classpath, built using compute-classpath.sh.
-CLASSPATH=`$FWDIR/bin/compute-classpath.sh`
-CLASSPATH="$SPARK_EXAMPLES_JAR:$CLASSPATH"
-
-if $cygwin; then
-    CLASSPATH=`cygpath -wp $CLASSPATH`
-    export SPARK_EXAMPLES_JAR=`cygpath -w $SPARK_EXAMPLES_JAR`
-fi
-
-# Find java binary
-if [ -n "${JAVA_HOME}" ]; then
-  RUNNER="${JAVA_HOME}/bin/java"
-else
-  if [ `command -v java` ]; then
-    RUNNER="java"
-  else
-    echo "JAVA_HOME is not set" >&2
-    exit 1
-  fi
-fi
-
-# Set JAVA_OPTS to be able to load native libraries and to set heap size
-JAVA_OPTS="$SPARK_JAVA_OPTS"
-# Load extra JAVA_OPTS from conf/java-opts, if it exists
-if [ -e "$FWDIR/conf/java-opts" ] ; then
-  JAVA_OPTS="$JAVA_OPTS `cat $FWDIR/conf/java-opts`"
+if [ -n "$1" ]; then
+  EXAMPLE_CLASS="$1"
+  shift
+else 
+  echo "usage: ./bin/run-example <example-class> [example-args]" 
+  echo "  - set MASTER=XX to use a specific master"
+  echo "  - can use abbreviated example class name (e.g. SparkPi, mllib.MovieLensALS)"
+  echo
+  exit -1
 fi
-export JAVA_OPTS
 
-if [ "$SPARK_PRINT_LAUNCH_COMMAND" == "1" ]; then
-  echo -n "Spark Command: "
-  echo "$RUNNER" -cp "$CLASSPATH" $JAVA_OPTS "$@"
-  echo "========================================"
-  echo
+if [[ ! $EXAMPLE_CLASS == org.apache.spark.examples* ]]; then
+  EXAMPLE_CLASS="org.apache.spark.examples.$EXAMPLE_CLASS"
 fi
 
-exec "$RUNNER" -cp "$CLASSPATH" $JAVA_OPTS "$@"
+./bin/spark-submit \
+  --master $EXAMPLE_MASTER \
+  --class $EXAMPLE_CLASS \
+  $SPARK_EXAMPLES_JAR \
+  "$@"

bin/spark-class

@@ -114,7 +114,7 @@ if [ ! -f "$FWDIR/RELEASE" ]; then
   jars_list=$(ls "$FWDIR"/assembly/target/scala-$SCALA_VERSION/ | grep "spark-assembly.*hadoop.*.jar")
   if [ "$num_jars" -eq "0" ]; then
     echo "Failed to find Spark assembly in $FWDIR/assembly/target/scala-$SCALA_VERSION/" >&2
-    echo "You need to build Spark with 'sbt/sbt assembly' before running this program." >&2
+    echo "You need to build Spark before running this program." >&2
     exit 1
   fi
   if [ "$num_jars" -gt "1" ]; then

core/src/main/scala/org/apache/spark/rdd/RDD.scala

@@ -330,9 +330,9 @@ abstract class RDD[T: ClassTag](
     if (shuffle) {
       // include a shuffle step so that our upstream tasks are still distributed
       new CoalescedRDD(
-        new ShuffledRDD[T, Null, (T, Null)](map(x => (x, null)),
+        new ShuffledRDD[Int, T, (Int, T)](map(x => (Utils.random.nextInt(), x)),
         new HashPartitioner(numPartitions)),
-        numPartitions).keys
+        numPartitions).values
     } else {
       new CoalescedRDD(this, numPartitions)
     }

core/src/main/scala/org/apache/spark/scheduler/ShuffleMapTask.scala

@@ -57,15 +57,12 @@ private[spark] object ShuffleMapTask {
   }
 
   def deserializeInfo(stageId: Int, bytes: Array[Byte]): (RDD[_], ShuffleDependency[_,_]) = {
-    synchronized {
-      val loader = Thread.currentThread.getContextClassLoader
-      val in = new GZIPInputStream(new ByteArrayInputStream(bytes))
-      val ser = SparkEnv.get.closureSerializer.newInstance()
-      val objIn = ser.deserializeStream(in)
-      val rdd = objIn.readObject().asInstanceOf[RDD[_]]
-      val dep = objIn.readObject().asInstanceOf[ShuffleDependency[_,_]]
-      (rdd, dep)
-    }
+    val in = new GZIPInputStream(new ByteArrayInputStream(bytes))
+    val ser = SparkEnv.get.closureSerializer.newInstance()
+    val objIn = ser.deserializeStream(in)
+    val rdd = objIn.readObject().asInstanceOf[RDD[_]]
+    val dep = objIn.readObject().asInstanceOf[ShuffleDependency[_,_]]
+    (rdd, dep)
   }
 
   // Since both the JarSet and FileSet have the same format this is used for both.

core/src/main/scala/org/apache/spark/storage/BlockManagerMasterActor.scala

@@ -250,7 +250,7 @@ class BlockManagerMasterActor(val isLocal: Boolean, conf: SparkConf, listenerBus
           // Remove the block from the slave's BlockManager.
           // Doesn't actually wait for a confirmation and the message might get lost.
           // If message loss becomes frequent, we should add retry logic here.
-          blockManager.get.slaveActor ! RemoveBlock(blockId)
+          blockManager.get.slaveActor.ask(RemoveBlock(blockId))(akkaTimeout)
         }
       }
     }

docs/building-with-maven.md

@@ -96,7 +96,7 @@ Tests are run by default via the [ScalaTest Maven plugin](http://www.scalatest.o
 
 The ScalaTest plugin also supports running only a specific test suite as follows:
 
-    $ mvn -Dhadoop.version=... -Dsuites=org.apache.spark.repl.ReplSuite test
+    $ mvn -Dhadoop.version=... -DwildcardSuites=org.apache.spark.repl.ReplSuite test
 
 
 ## Continuous Compilation ##

docs/running-on-yarn.md

@@ -53,7 +53,7 @@ For example:
         --driver-memory 4g \
         --executor-memory 2g \
         --executor-cores 1
-        examples/target/scala-{{site.SCALA_BINARY_VERSION}}/spark-examples-assembly-{{site.SPARK_VERSION}}.jar \
+        lib/spark-examples*.jar \
         yarn-cluster 5
 
 The above starts a YARN client program which starts the default Application Master. Then SparkPi will be run as a child thread of Application Master. The client will periodically poll the Application Master for status updates and display them in the console. The client will exit once your application has finished running.  Refer to the "Viewing Logs" section below for how to see driver and executor logs.

make-distribution.sh

@@ -40,6 +40,8 @@
 #
 
 set -o pipefail
+set -e
+
 # Figure out where the Spark framework is installed
 FWDIR="$(cd `dirname $0`; pwd)"
 DISTDIR="$FWDIR/dist"