Merge branch 'master' of https://github.com/apache/spark

Author: zhzhan

assembly/pom.xml

@@ -349,5 +349,15 @@
         </plugins>
       </build>
     </profile>
+    <profile>
+      <id>kinesis-asl</id>
+      <dependencies>
+        <dependency>
+          <groupId>org.apache.httpcomponents</groupId>
+          <artifactId>httpclient</artifactId>
+          <version>${commons.httpclient.version}</version>
+        </dependency>
+      </dependencies>
+    </profile>
   </profiles>
 </project>

bin/compute-classpath.cmd

@@ -36,7 +36,13 @@ rem Load environment variables from conf\spark-env.cmd, if it exists
 if exist "%FWDIR%conf\spark-env.cmd" call "%FWDIR%conf\spark-env.cmd"
 
 rem Build up classpath
-set CLASSPATH=%SPARK_CLASSPATH%;%SPARK_SUBMIT_CLASSPATH%;%FWDIR%conf
+set CLASSPATH=%SPARK_CLASSPATH%;%SPARK_SUBMIT_CLASSPATH%
+
+if "x%SPARK_CONF_DIR%"!="x" (
+  set CLASSPATH=%CLASSPATH%;%SPARK_CONF_DIR%
+) else (
+  set CLASSPATH=%CLASSPATH%;%FWDIR%conf
+)
 
 if exist "%FWDIR%RELEASE" (
   for %%d in ("%FWDIR%lib\spark-assembly*.jar") do (

bin/compute-classpath.sh

@@ -27,8 +27,14 @@ FWDIR="$(cd "`dirname "$0"`"/..; pwd)"
 
 . "$FWDIR"/bin/load-spark-env.sh
 
+CLASSPATH="$SPARK_CLASSPATH:$SPARK_SUBMIT_CLASSPATH"
+
 # Build up classpath
-CLASSPATH="$SPARK_CLASSPATH:$SPARK_SUBMIT_CLASSPATH:$FWDIR/conf"
+if [ -n "$SPARK_CONF_DIR" ]; then
+  CLASSPATH="$CLASSPATH:$SPARK_CONF_DIR"
+else
+  CLASSPATH="$CLASSPATH:$FWDIR/conf"
+fi
 
 ASSEMBLY_DIR="$FWDIR/assembly/target/scala-$SCALA_VERSION"
 

bin/pyspark

@@ -52,10 +52,20 @@ fi
 
 # Figure out which Python executable to use
 if [[ -z "$PYSPARK_PYTHON" ]]; then
-  PYSPARK_PYTHON="python"
+  if [[ "$IPYTHON" = "1" || -n "$IPYTHON_OPTS" ]]; then
+    # for backward compatibility
+    PYSPARK_PYTHON="ipython"
+  else
+    PYSPARK_PYTHON="python"
+  fi
 fi
 export PYSPARK_PYTHON
 
+if [[ -z "$PYSPARK_PYTHON_OPTS" && -n "$IPYTHON_OPTS" ]]; then
+  # for backward compatibility
+  PYSPARK_PYTHON_OPTS="$IPYTHON_OPTS"
+fi
+
 # Add the PySpark classes to the Python path:
 export PYTHONPATH="$SPARK_HOME/python/:$PYTHONPATH"
 export PYTHONPATH="$SPARK_HOME/python/lib/py4j-0.8.2.1-src.zip:$PYTHONPATH"
@@ -64,11 +74,6 @@ export PYTHONPATH="$SPARK_HOME/python/lib/py4j-0.8.2.1-src.zip:$PYTHONPATH"
 export OLD_PYTHONSTARTUP="$PYTHONSTARTUP"
 export PYTHONSTARTUP="$FWDIR/python/pyspark/shell.py"
 
-# If IPython options are specified, assume user wants to run IPython
-if [[ -n "$IPYTHON_OPTS" ]]; then
-  IPYTHON=1
-fi
-
 # Build up arguments list manually to preserve quotes and backslashes.
 # We export Spark submit arguments as an environment variable because shell.py must run as a
 # PYTHONSTARTUP script, which does not take in arguments. This is required for IPython notebooks.
@@ -106,10 +111,5 @@ if [[ "$1" =~ \.py$ ]]; then
 else
   # PySpark shell requires special handling downstream
   export PYSPARK_SHELL=1
-  # Only use ipython if no command line arguments were provided [SPARK-1134]
-  if [[ "$IPYTHON" = "1" ]]; then
-    exec ${PYSPARK_PYTHON:-ipython} $IPYTHON_OPTS
-  else
-    exec "$PYSPARK_PYTHON"
-  fi
+  exec "$PYSPARK_PYTHON" $PYSPARK_PYTHON_OPTS
 fi

bin/pyspark2.cmd

@@ -33,7 +33,7 @@ for %%d in ("%FWDIR%assembly\target\scala-%SCALA_VERSION%\spark-assembly*hadoop*
 )
 if [%FOUND_JAR%] == [0] (
   echo Failed to find Spark assembly JAR.
-  echo You need to build Spark with sbt\sbt assembly before running this program.
+  echo You need to build Spark before running this program.
   goto exit
 )
 :skip_build_test

bin/run-example2.cmd

@@ -52,7 +52,7 @@ if exist "%FWDIR%RELEASE" (
 )
 if "x%SPARK_EXAMPLES_JAR%"=="x" (
   echo Failed to find Spark examples assembly JAR.
-  echo You need to build Spark with sbt\sbt assembly before running this program.
+  echo You need to build Spark before running this program.
   goto exit
 )
 

bin/spark-class

@@ -146,7 +146,7 @@ fi
 if [[ "$1" =~ org.apache.spark.tools.* ]]; then
   if test -z "$SPARK_TOOLS_JAR"; then
     echo "Failed to find Spark Tools Jar in $FWDIR/tools/target/scala-$SCALA_VERSION/" 1>&2
-    echo "You need to build spark before running $1." 1>&2
+    echo "You need to build Spark before running $1." 1>&2
     exit 1
   fi
   CLASSPATH="$CLASSPATH:$SPARK_TOOLS_JAR"

bin/spark-class2.cmd

@@ -104,7 +104,7 @@ for %%d in ("%FWDIR%assembly\target\scala-%SCALA_VERSION%\spark-assembly*hadoop*
 )
 if "%FOUND_JAR%"=="0" (
   echo Failed to find Spark assembly JAR.
-  echo You need to build Spark with sbt\sbt assembly before running this program.
+  echo You need to build Spark before running this program.
   goto exit
 )
 :skip_build_test

bin/utils.sh

@@ -17,7 +17,7 @@
 # limitations under the License.
 #
 
-# Gather all all spark-submit options into SUBMISSION_OPTS
+# Gather all spark-submit options into SUBMISSION_OPTS
 function gatherSparkSubmitOpts() {
 
   if [ -z "$SUBMIT_USAGE_FUNCTION" ]; then

core/src/main/scala/org/apache/spark/FutureAction.scala

@@ -83,6 +83,15 @@ trait FutureAction[T] extends Future[T] {
    */
   @throws(classOf[Exception])
   def get(): T = Await.result(this, Duration.Inf)
+
+  /**
+   * Returns the job IDs run by the underlying async operation.
+   *
+   * This returns the current snapshot of the job list. Certain operations may run multiple
+   * jobs, so multiple calls to this method may return different lists.
+   */
+  def jobIds: Seq[Int]
+
 }
 
 
@@ -150,8 +159,7 @@ class SimpleFutureAction[T] private[spark](jobWaiter: JobWaiter[_], resultFunc:
     }
   }
 
-  /** Get the corresponding job id for this action. */
-  def jobId = jobWaiter.jobId
+  def jobIds = Seq(jobWaiter.jobId)
 }
 
 
@@ -171,6 +179,8 @@ class ComplexFutureAction[T] extends FutureAction[T] {
   // is cancelled before the action was even run (and thus we have no thread to interrupt).
   @volatile private var _cancelled: Boolean = false
 
+  @volatile private var jobs: Seq[Int] = Nil
+
   // A promise used to signal the future.
   private val p = promise[T]()
 
@@ -219,6 +229,8 @@ class ComplexFutureAction[T] extends FutureAction[T] {
       }
     }
 
+    this.jobs = jobs ++ job.jobIds
+
     // Wait for the job to complete. If the action is cancelled (with an interrupt),
     // cancel the job and stop the execution. This is not in a synchronized block because
     // Await.ready eventually waits on the monitor in FutureJob.jobWaiter.
@@ -255,4 +267,7 @@ class ComplexFutureAction[T] extends FutureAction[T] {
   override def isCompleted: Boolean = p.isCompleted
 
   override def value: Option[Try[T]] = p.future.value
+
+  def jobIds = jobs
+
 }