diff options
Diffstat (limited to 'bin')
-rw-r--r-- | bin/compute-classpath.cmd | 124 | ||||
-rwxr-xr-x | bin/compute-classpath.sh | 161 | ||||
-rw-r--r-- | bin/load-spark-env.sh | 8 | ||||
-rwxr-xr-x | bin/pyspark | 59 | ||||
-rw-r--r-- | bin/pyspark2.cmd | 57 | ||||
-rwxr-xr-x | bin/run-example | 2 | ||||
-rwxr-xr-x | bin/spark-class | 180 | ||||
-rw-r--r-- | bin/spark-class2.cmd | 141 | ||||
-rwxr-xr-x | bin/spark-shell | 23 | ||||
-rw-r--r-- | bin/spark-shell2.cmd | 27 | ||||
-rwxr-xr-x | bin/spark-sql | 20 | ||||
-rwxr-xr-x | bin/spark-submit | 66 | ||||
-rw-r--r-- | bin/spark-submit2.cmd | 71 | ||||
-rwxr-xr-x | bin/utils.sh | 60 | ||||
-rw-r--r-- | bin/windows-utils.cmd | 60 |
15 files changed, 162 insertions, 897 deletions
diff --git a/bin/compute-classpath.cmd b/bin/compute-classpath.cmd deleted file mode 100644 index 088f993954..0000000000 --- a/bin/compute-classpath.cmd +++ /dev/null @@ -1,124 +0,0 @@ -@echo off - -rem -rem Licensed to the Apache Software Foundation (ASF) under one or more -rem contributor license agreements. See the NOTICE file distributed with -rem this work for additional information regarding copyright ownership. -rem The ASF licenses this file to You under the Apache License, Version 2.0 -rem (the "License"); you may not use this file except in compliance with -rem the License. You may obtain a copy of the License at -rem -rem http://www.apache.org/licenses/LICENSE-2.0 -rem -rem Unless required by applicable law or agreed to in writing, software -rem distributed under the License is distributed on an "AS IS" BASIS, -rem WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -rem See the License for the specific language governing permissions and -rem limitations under the License. -rem - -rem This script computes Spark's classpath and prints it to stdout; it's used by both the "run" -rem script and the ExecutorRunner in standalone cluster mode. - -rem If we're called from spark-class2.cmd, it already set enabledelayedexpansion and setting -rem it here would stop us from affecting its copy of the CLASSPATH variable; otherwise we -rem need to set it here because we use !datanucleus_jars! below. -if "%DONT_PRINT_CLASSPATH%"=="1" goto skip_delayed_expansion -setlocal enabledelayedexpansion -:skip_delayed_expansion - -set SCALA_VERSION=2.10 - -rem Figure out where the Spark framework is installed -set FWDIR=%~dp0..\ - -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% - -if not "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 ( - set ASSEMBLY_JAR=%%d - ) -) else ( - for %%d in ("%FWDIR%assembly\target\scala-%SCALA_VERSION%\spark-assembly*hadoop*.jar") do ( - set ASSEMBLY_JAR=%%d - ) -) - -set CLASSPATH=%CLASSPATH%;%ASSEMBLY_JAR% - -rem When Hive support is needed, Datanucleus jars must be included on the classpath. -rem Datanucleus jars do not work if only included in the uber jar as plugin.xml metadata is lost. -rem Both sbt and maven will populate "lib_managed/jars/" with the datanucleus jars when Spark is -rem built with Hive, so look for them there. -if exist "%FWDIR%RELEASE" ( - set datanucleus_dir=%FWDIR%lib -) else ( - set datanucleus_dir=%FWDIR%lib_managed\jars -) -set "datanucleus_jars=" -for %%d in ("%datanucleus_dir%\datanucleus-*.jar") do ( - set datanucleus_jars=!datanucleus_jars!;%%d -) -set CLASSPATH=%CLASSPATH%;%datanucleus_jars% - -set SPARK_CLASSES=%FWDIR%core\target\scala-%SCALA_VERSION%\classes -set SPARK_CLASSES=%SPARK_CLASSES%;%FWDIR%repl\target\scala-%SCALA_VERSION%\classes -set SPARK_CLASSES=%SPARK_CLASSES%;%FWDIR%mllib\target\scala-%SCALA_VERSION%\classes -set SPARK_CLASSES=%SPARK_CLASSES%;%FWDIR%bagel\target\scala-%SCALA_VERSION%\classes -set SPARK_CLASSES=%SPARK_CLASSES%;%FWDIR%graphx\target\scala-%SCALA_VERSION%\classes -set SPARK_CLASSES=%SPARK_CLASSES%;%FWDIR%streaming\target\scala-%SCALA_VERSION%\classes -set SPARK_CLASSES=%SPARK_CLASSES%;%FWDIR%tools\target\scala-%SCALA_VERSION%\classes -set SPARK_CLASSES=%SPARK_CLASSES%;%FWDIR%sql\catalyst\target\scala-%SCALA_VERSION%\classes -set SPARK_CLASSES=%SPARK_CLASSES%;%FWDIR%sql\core\target\scala-%SCALA_VERSION%\classes -set SPARK_CLASSES=%SPARK_CLASSES%;%FWDIR%sql\hive\target\scala-%SCALA_VERSION%\classes - -set SPARK_TEST_CLASSES=%FWDIR%core\target\scala-%SCALA_VERSION%\test-classes -set SPARK_TEST_CLASSES=%SPARK_TEST_CLASSES%;%FWDIR%repl\target\scala-%SCALA_VERSION%\test-classes -set SPARK_TEST_CLASSES=%SPARK_TEST_CLASSES%;%FWDIR%mllib\target\scala-%SCALA_VERSION%\test-classes -set SPARK_TEST_CLASSES=%SPARK_TEST_CLASSES%;%FWDIR%bagel\target\scala-%SCALA_VERSION%\test-classes -set SPARK_TEST_CLASSES=%SPARK_TEST_CLASSES%;%FWDIR%graphx\target\scala-%SCALA_VERSION%\test-classes -set SPARK_TEST_CLASSES=%SPARK_TEST_CLASSES%;%FWDIR%streaming\target\scala-%SCALA_VERSION%\test-classes -set SPARK_TEST_CLASSES=%SPARK_TEST_CLASSES%;%FWDIR%sql\catalyst\target\scala-%SCALA_VERSION%\test-classes -set SPARK_TEST_CLASSES=%SPARK_TEST_CLASSES%;%FWDIR%sql\core\target\scala-%SCALA_VERSION%\test-classes -set SPARK_TEST_CLASSES=%SPARK_TEST_CLASSES%;%FWDIR%sql\hive\target\scala-%SCALA_VERSION%\test-classes - -if "x%SPARK_TESTING%"=="x1" ( - rem Add test clases to path - note, add SPARK_CLASSES and SPARK_TEST_CLASSES before CLASSPATH - rem so that local compilation takes precedence over assembled jar - set CLASSPATH=%SPARK_CLASSES%;%SPARK_TEST_CLASSES%;%CLASSPATH% -) - -rem Add hadoop conf dir - else FileSystem.*, etc fail -rem Note, this assumes that there is either a HADOOP_CONF_DIR or YARN_CONF_DIR which hosts -rem the configurtion files. -if "x%HADOOP_CONF_DIR%"=="x" goto no_hadoop_conf_dir - set CLASSPATH=%CLASSPATH%;%HADOOP_CONF_DIR% -:no_hadoop_conf_dir - -if "x%YARN_CONF_DIR%"=="x" goto no_yarn_conf_dir - set CLASSPATH=%CLASSPATH%;%YARN_CONF_DIR% -:no_yarn_conf_dir - -rem To allow for distributions to append needed libraries to the classpath (e.g. when -rem using the "hadoop-provided" profile to build Spark), check SPARK_DIST_CLASSPATH and -rem append it to tbe final classpath. -if not "x%$SPARK_DIST_CLASSPATH%"=="x" ( - set CLASSPATH=%CLASSPATH%;%SPARK_DIST_CLASSPATH% -) - -rem A bit of a hack to allow calling this script within run2.cmd without seeing output -if "%DONT_PRINT_CLASSPATH%"=="1" goto exit - -echo %CLASSPATH% - -:exit diff --git a/bin/compute-classpath.sh b/bin/compute-classpath.sh deleted file mode 100755 index f4f6b7b909..0000000000 --- a/bin/compute-classpath.sh +++ /dev/null @@ -1,161 +0,0 @@ -#!/usr/bin/env bash - -# -# Licensed to the Apache Software Foundation (ASF) under one or more -# contributor license agreements. See the NOTICE file distributed with -# this work for additional information regarding copyright ownership. -# The ASF licenses this file to You under the Apache License, Version 2.0 -# (the "License"); you may not use this file except in compliance with -# the License. You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. -# - -# This script computes Spark's classpath and prints it to stdout; it's used by both the "run" -# script and the ExecutorRunner in standalone cluster mode. - -# Figure out where Spark is installed -FWDIR="$(cd "`dirname "$0"`"/..; pwd)" - -. "$FWDIR"/bin/load-spark-env.sh - -if [ -n "$SPARK_CLASSPATH" ]; then - CLASSPATH="$SPARK_CLASSPATH:$SPARK_SUBMIT_CLASSPATH" -else - CLASSPATH="$SPARK_SUBMIT_CLASSPATH" -fi - -# Build up classpath -if [ -n "$SPARK_CONF_DIR" ]; then - CLASSPATH="$CLASSPATH:$SPARK_CONF_DIR" -else - CLASSPATH="$CLASSPATH:$FWDIR/conf" -fi - -ASSEMBLY_DIR="$FWDIR/assembly/target/scala-$SPARK_SCALA_VERSION" - -if [ -n "$JAVA_HOME" ]; then - JAR_CMD="$JAVA_HOME/bin/jar" -else - JAR_CMD="jar" -fi - -# A developer option to prepend more recently compiled Spark classes -if [ -n "$SPARK_PREPEND_CLASSES" ]; then - echo "NOTE: SPARK_PREPEND_CLASSES is set, placing locally compiled Spark"\ - "classes ahead of assembly." >&2 - # Spark classes - CLASSPATH="$CLASSPATH:$FWDIR/core/target/scala-$SPARK_SCALA_VERSION/classes" - CLASSPATH="$CLASSPATH:$FWDIR/repl/target/scala-$SPARK_SCALA_VERSION/classes" - CLASSPATH="$CLASSPATH:$FWDIR/mllib/target/scala-$SPARK_SCALA_VERSION/classes" - CLASSPATH="$CLASSPATH:$FWDIR/bagel/target/scala-$SPARK_SCALA_VERSION/classes" - CLASSPATH="$CLASSPATH:$FWDIR/graphx/target/scala-$SPARK_SCALA_VERSION/classes" - CLASSPATH="$CLASSPATH:$FWDIR/streaming/target/scala-$SPARK_SCALA_VERSION/classes" - CLASSPATH="$CLASSPATH:$FWDIR/tools/target/scala-$SPARK_SCALA_VERSION/classes" - CLASSPATH="$CLASSPATH:$FWDIR/sql/catalyst/target/scala-$SPARK_SCALA_VERSION/classes" - CLASSPATH="$CLASSPATH:$FWDIR/sql/core/target/scala-$SPARK_SCALA_VERSION/classes" - CLASSPATH="$CLASSPATH:$FWDIR/sql/hive/target/scala-$SPARK_SCALA_VERSION/classes" - CLASSPATH="$CLASSPATH:$FWDIR/sql/hive-thriftserver/target/scala-$SPARK_SCALA_VERSION/classes" - CLASSPATH="$CLASSPATH:$FWDIR/yarn/stable/target/scala-$SPARK_SCALA_VERSION/classes" - # Jars for shaded deps in their original form (copied here during build) - CLASSPATH="$CLASSPATH:$FWDIR/core/target/jars/*" -fi - -# Use spark-assembly jar from either RELEASE or assembly directory -if [ -f "$FWDIR/RELEASE" ]; then - assembly_folder="$FWDIR"/lib -else - assembly_folder="$ASSEMBLY_DIR" -fi - -num_jars=0 - -for f in "${assembly_folder}"/spark-assembly*hadoop*.jar; do - if [[ ! -e "$f" ]]; then - echo "Failed to find Spark assembly in $assembly_folder" 1>&2 - echo "You need to build Spark before running this program." 1>&2 - exit 1 - fi - ASSEMBLY_JAR="$f" - num_jars=$((num_jars+1)) -done - -if [ "$num_jars" -gt "1" ]; then - echo "Found multiple Spark assembly jars in $assembly_folder:" 1>&2 - ls "${assembly_folder}"/spark-assembly*hadoop*.jar 1>&2 - echo "Please remove all but one jar." 1>&2 - exit 1 -fi - -# Verify that versions of java used to build the jars and run Spark are compatible -jar_error_check=$("$JAR_CMD" -tf "$ASSEMBLY_JAR" nonexistent/class/path 2>&1) -if [[ "$jar_error_check" =~ "invalid CEN header" ]]; then - echo "Loading Spark jar with '$JAR_CMD' failed. " 1>&2 - echo "This is likely because Spark was compiled with Java 7 and run " 1>&2 - echo "with Java 6. (see SPARK-1703). Please use Java 7 to run Spark " 1>&2 - echo "or build Spark with Java 6." 1>&2 - exit 1 -fi - -CLASSPATH="$CLASSPATH:$ASSEMBLY_JAR" - -# When Hive support is needed, Datanucleus jars must be included on the classpath. -# Datanucleus jars do not work if only included in the uber jar as plugin.xml metadata is lost. -# Both sbt and maven will populate "lib_managed/jars/" with the datanucleus jars when Spark is -# built with Hive, so first check if the datanucleus jars exist, and then ensure the current Spark -# assembly is built for Hive, before actually populating the CLASSPATH with the jars. -# Note that this check order is faster (by up to half a second) in the case where Hive is not used. -if [ -f "$FWDIR/RELEASE" ]; then - datanucleus_dir="$FWDIR"/lib -else - datanucleus_dir="$FWDIR"/lib_managed/jars -fi - -datanucleus_jars="$(find "$datanucleus_dir" 2>/dev/null | grep "datanucleus-.*\\.jar$")" -datanucleus_jars="$(echo "$datanucleus_jars" | tr "\n" : | sed s/:$//g)" - -if [ -n "$datanucleus_jars" ]; then - hive_files=$("$JAR_CMD" -tf "$ASSEMBLY_JAR" org/apache/hadoop/hive/ql/exec 2>/dev/null) - if [ -n "$hive_files" ]; then - echo "Spark assembly has been built with Hive, including Datanucleus jars on classpath" 1>&2 - CLASSPATH="$CLASSPATH:$datanucleus_jars" - fi -fi - -# Add test classes if we're running from SBT or Maven with SPARK_TESTING set to 1 -if [[ $SPARK_TESTING == 1 ]]; then - CLASSPATH="$CLASSPATH:$FWDIR/core/target/scala-$SPARK_SCALA_VERSION/test-classes" - CLASSPATH="$CLASSPATH:$FWDIR/repl/target/scala-$SPARK_SCALA_VERSION/test-classes" - CLASSPATH="$CLASSPATH:$FWDIR/mllib/target/scala-$SPARK_SCALA_VERSION/test-classes" - CLASSPATH="$CLASSPATH:$FWDIR/bagel/target/scala-$SPARK_SCALA_VERSION/test-classes" - CLASSPATH="$CLASSPATH:$FWDIR/graphx/target/scala-$SPARK_SCALA_VERSION/test-classes" - CLASSPATH="$CLASSPATH:$FWDIR/streaming/target/scala-$SPARK_SCALA_VERSION/test-classes" - CLASSPATH="$CLASSPATH:$FWDIR/sql/catalyst/target/scala-$SPARK_SCALA_VERSION/test-classes" - CLASSPATH="$CLASSPATH:$FWDIR/sql/core/target/scala-$SPARK_SCALA_VERSION/test-classes" - CLASSPATH="$CLASSPATH:$FWDIR/sql/hive/target/scala-$SPARK_SCALA_VERSION/test-classes" -fi - -# Add hadoop conf dir if given -- otherwise FileSystem.*, etc fail ! -# Note, this assumes that there is either a HADOOP_CONF_DIR or YARN_CONF_DIR which hosts -# the configurtion files. -if [ -n "$HADOOP_CONF_DIR" ]; then - CLASSPATH="$CLASSPATH:$HADOOP_CONF_DIR" -fi -if [ -n "$YARN_CONF_DIR" ]; then - CLASSPATH="$CLASSPATH:$YARN_CONF_DIR" -fi - -# To allow for distributions to append needed libraries to the classpath (e.g. when -# using the "hadoop-provided" profile to build Spark), check SPARK_DIST_CLASSPATH and -# append it to tbe final classpath. -if [ -n "$SPARK_DIST_CLASSPATH" ]; then - CLASSPATH="$CLASSPATH:$SPARK_DIST_CLASSPATH" -fi - -echo "$CLASSPATH" diff --git a/bin/load-spark-env.sh b/bin/load-spark-env.sh index 356b3d49b2..2d7070c25d 100644 --- a/bin/load-spark-env.sh +++ b/bin/load-spark-env.sh @@ -41,9 +41,9 @@ fi if [ -z "$SPARK_SCALA_VERSION" ]; then - ASSEMBLY_DIR2="$FWDIR/assembly/target/scala-2.11" - ASSEMBLY_DIR1="$FWDIR/assembly/target/scala-2.10" - + ASSEMBLY_DIR2="$SPARK_HOME/assembly/target/scala-2.11" + ASSEMBLY_DIR1="$SPARK_HOME/assembly/target/scala-2.10" + if [[ -d "$ASSEMBLY_DIR2" && -d "$ASSEMBLY_DIR1" ]]; then echo -e "Presence of build for both scala versions(SCALA 2.10 and SCALA 2.11) detected." 1>&2 echo -e 'Either clean one of them or, export SPARK_SCALA_VERSION=2.11 in spark-env.sh.' 1>&2 @@ -54,5 +54,5 @@ if [ -z "$SPARK_SCALA_VERSION" ]; then export SPARK_SCALA_VERSION="2.11" else export SPARK_SCALA_VERSION="2.10" - fi + fi fi diff --git a/bin/pyspark b/bin/pyspark index 0b4f695dd0..e7f6a1a072 100755 --- a/bin/pyspark +++ b/bin/pyspark @@ -18,36 +18,24 @@ # # Figure out where Spark is installed -FWDIR="$(cd "`dirname "$0"`"/..; pwd)" +export SPARK_HOME="$(cd "`dirname "$0"`"/..; pwd)" -# Export this as SPARK_HOME -export SPARK_HOME="$FWDIR" - -source "$FWDIR/bin/utils.sh" - -source "$FWDIR"/bin/load-spark-env.sh +source "$SPARK_HOME"/bin/load-spark-env.sh function usage() { + if [ -n "$1" ]; then + echo $1 + fi echo "Usage: ./bin/pyspark [options]" 1>&2 - "$FWDIR"/bin/spark-submit --help 2>&1 | grep -v Usage 1>&2 - exit 0 + "$SPARK_HOME"/bin/spark-submit --help 2>&1 | grep -v Usage 1>&2 + exit $2 } +export -f usage if [[ "$@" = *--help ]] || [[ "$@" = *-h ]]; then usage fi -# Exit if the user hasn't compiled Spark -if [ ! -f "$FWDIR/RELEASE" ]; then - # Exit if the user hasn't compiled Spark - ls "$FWDIR"/assembly/target/scala-$SPARK_SCALA_VERSION/spark-assembly*hadoop*.jar >& /dev/null - if [[ $? != 0 ]]; then - echo "Failed to find Spark assembly in $FWDIR/assembly/target" 1>&2 - echo "You need to build Spark before running this program" 1>&2 - exit 1 - fi -fi - # In Spark <= 1.1, setting IPYTHON=1 would cause the driver to be launched using the `ipython` # executable, while the worker would still be launched using PYSPARK_PYTHON. # @@ -95,26 +83,13 @@ export PYTHONPATH="$SPARK_HOME/python/lib/py4j-0.8.2.1-src.zip:$PYTHONPATH" # Load the PySpark shell.py script when ./pyspark is used interactively: export OLD_PYTHONSTARTUP="$PYTHONSTARTUP" -export PYTHONSTARTUP="$FWDIR/python/pyspark/shell.py" - -# Build up arguments list manually to preserve quotes and backslashes. -# We export Spark submit arguments as an environment variable because shell.py must run as a -# PYTHONSTARTUP script, which does not take in arguments. This is required for IPython notebooks. -SUBMIT_USAGE_FUNCTION=usage -gatherSparkSubmitOpts "$@" -PYSPARK_SUBMIT_ARGS="" -whitespace="[[:space:]]" -for i in "${SUBMISSION_OPTS[@]}"; do - if [[ $i =~ \" ]]; then i=$(echo $i | sed 's/\"/\\\"/g'); fi - if [[ $i =~ $whitespace ]]; then i=\"$i\"; fi - PYSPARK_SUBMIT_ARGS="$PYSPARK_SUBMIT_ARGS $i" -done -export PYSPARK_SUBMIT_ARGS +export PYTHONSTARTUP="$SPARK_HOME/python/pyspark/shell.py" # For pyspark tests if [[ -n "$SPARK_TESTING" ]]; then unset YARN_CONF_DIR unset HADOOP_CONF_DIR + export PYSPARK_SUBMIT_ARGS=pyspark-shell if [[ -n "$PYSPARK_DOC_TEST" ]]; then exec "$PYSPARK_DRIVER_PYTHON" -m doctest $1 else @@ -123,14 +98,6 @@ if [[ -n "$SPARK_TESTING" ]]; then exit fi -# If a python file is provided, directly run spark-submit. -if [[ "$1" =~ \.py$ ]]; then - echo -e "\nWARNING: Running python applications through ./bin/pyspark is deprecated as of Spark 1.0." 1>&2 - echo -e "Use ./bin/spark-submit <python file>\n" 1>&2 - primary="$1" - shift - gatherSparkSubmitOpts "$@" - exec "$FWDIR"/bin/spark-submit "${SUBMISSION_OPTS[@]}" "$primary" "${APPLICATION_OPTS[@]}" -else - exec "$PYSPARK_DRIVER_PYTHON" $PYSPARK_DRIVER_PYTHON_OPTS -fi +export PYSPARK_DRIVER_PYTHON +export PYSPARK_DRIVER_PYTHON_OPTS +exec "$SPARK_HOME"/bin/spark-submit pyspark-shell-main "$@" diff --git a/bin/pyspark2.cmd b/bin/pyspark2.cmd index a542ec80b4..4f5eb5e206 100644 --- a/bin/pyspark2.cmd +++ b/bin/pyspark2.cmd @@ -17,59 +17,22 @@ rem See the License for the specific language governing permissions and rem limitations under the License. rem -set SCALA_VERSION=2.10 - rem Figure out where the Spark framework is installed -set FWDIR=%~dp0..\ - -rem Export this as SPARK_HOME -set SPARK_HOME=%FWDIR% - -rem Test whether the user has built Spark -if exist "%FWDIR%RELEASE" goto skip_build_test -set FOUND_JAR=0 -for %%d in ("%FWDIR%assembly\target\scala-%SCALA_VERSION%\spark-assembly*hadoop*.jar") do ( - set FOUND_JAR=1 -) -if [%FOUND_JAR%] == [0] ( - echo Failed to find Spark assembly JAR. - echo You need to build Spark before running this program. - goto exit -) -:skip_build_test +set SPARK_HOME=%~dp0.. 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" +if exist "%SPARK_HOME%\conf\spark-env.cmd" call "%SPARK_HOME%\conf\spark-env.cmd" rem Figure out which Python to use. -if [%PYSPARK_PYTHON%] == [] set PYSPARK_PYTHON=python +if "x%PYSPARK_DRIVER_PYTHON%"=="x" ( + set PYSPARK_DRIVER_PYTHON=python + if not [%PYSPARK_PYTHON%] == [] set PYSPARK_DRIVER_PYTHON=%PYSPARK_PYTHON% +) -set PYTHONPATH=%FWDIR%python;%PYTHONPATH% -set PYTHONPATH=%FWDIR%python\lib\py4j-0.8.2.1-src.zip;%PYTHONPATH% +set PYTHONPATH=%SPARK_HOME%\python;%PYTHONPATH% +set PYTHONPATH=%SPARK_HOME%\python\lib\py4j-0.8.2.1-src.zip;%PYTHONPATH% set OLD_PYTHONSTARTUP=%PYTHONSTARTUP% -set PYTHONSTARTUP=%FWDIR%python\pyspark\shell.py -set PYSPARK_SUBMIT_ARGS=%* - -echo Running %PYSPARK_PYTHON% with PYTHONPATH=%PYTHONPATH% - -rem Check whether the argument is a file -for /f %%i in ('echo %1^| findstr /R "\.py"') do ( - set PYTHON_FILE=%%i -) - -if [%PYTHON_FILE%] == [] ( - if [%IPYTHON%] == [1] ( - ipython %IPYTHON_OPTS% - ) else ( - %PYSPARK_PYTHON% - ) -) else ( - echo. - echo WARNING: Running python applications through ./bin/pyspark.cmd is deprecated as of Spark 1.0. - echo Use ./bin/spark-submit ^<python file^> - echo. - "%FWDIR%\bin\spark-submit.cmd" %PYSPARK_SUBMIT_ARGS% -) +set PYTHONSTARTUP=%SPARK_HOME%\python\pyspark\shell.py -:exit +call %SPARK_HOME%\bin\spark-submit2.cmd pyspark-shell-main %* diff --git a/bin/run-example b/bin/run-example index a106411392..798e2caeb8 100755 --- a/bin/run-example +++ b/bin/run-example @@ -67,7 +67,7 @@ if [[ ! $EXAMPLE_CLASS == org.apache.spark.examples* ]]; then EXAMPLE_CLASS="org.apache.spark.examples.$EXAMPLE_CLASS" fi -"$FWDIR"/bin/spark-submit \ +exec "$FWDIR"/bin/spark-submit \ --master $EXAMPLE_MASTER \ --class $EXAMPLE_CLASS \ "$SPARK_EXAMPLES_JAR" \ diff --git a/bin/spark-class b/bin/spark-class index 2f0441bb3c..e29b234afa 100755 --- a/bin/spark-class +++ b/bin/spark-class @@ -16,89 +16,18 @@ # See the License for the specific language governing permissions and # limitations under the License. # - -# NOTE: Any changes to this file must be reflected in SparkSubmitDriverBootstrapper.scala! - -cygwin=false -case "`uname`" in - CYGWIN*) cygwin=true;; -esac +set -e # Figure out where Spark is installed -FWDIR="$(cd "`dirname "$0"`"/..; pwd)" +export SPARK_HOME="$(cd "`dirname "$0"`"/..; pwd)" -# Export this as SPARK_HOME -export SPARK_HOME="$FWDIR" -export SPARK_CONF_DIR="${SPARK_CONF_DIR:-"$SPARK_HOME/conf"}" - -. "$FWDIR"/bin/load-spark-env.sh +. "$SPARK_HOME"/bin/load-spark-env.sh if [ -z "$1" ]; then echo "Usage: spark-class <class> [<args>]" 1>&2 exit 1 fi -if [ -n "$SPARK_MEM" ]; then - echo -e "Warning: SPARK_MEM is deprecated, please use a more specific config option" 1>&2 - echo -e "(e.g., spark.executor.memory or spark.driver.memory)." 1>&2 -fi - -# Use SPARK_MEM or 512m as the default memory, to be overridden by specific options -DEFAULT_MEM=${SPARK_MEM:-512m} - -SPARK_DAEMON_JAVA_OPTS="$SPARK_DAEMON_JAVA_OPTS -Dspark.akka.logLifecycleEvents=true" - -# Add java opts and memory settings for master, worker, history server, executors, and repl. -case "$1" in - # Master, Worker, and HistoryServer use SPARK_DAEMON_JAVA_OPTS (and specific opts) + SPARK_DAEMON_MEMORY. - 'org.apache.spark.deploy.master.Master') - OUR_JAVA_OPTS="$SPARK_DAEMON_JAVA_OPTS $SPARK_MASTER_OPTS" - OUR_JAVA_MEM=${SPARK_DAEMON_MEMORY:-$DEFAULT_MEM} - ;; - 'org.apache.spark.deploy.worker.Worker') - OUR_JAVA_OPTS="$SPARK_DAEMON_JAVA_OPTS $SPARK_WORKER_OPTS" - OUR_JAVA_MEM=${SPARK_DAEMON_MEMORY:-$DEFAULT_MEM} - ;; - 'org.apache.spark.deploy.history.HistoryServer') - OUR_JAVA_OPTS="$SPARK_DAEMON_JAVA_OPTS $SPARK_HISTORY_OPTS" - OUR_JAVA_MEM=${SPARK_DAEMON_MEMORY:-$DEFAULT_MEM} - ;; - - # Executors use SPARK_JAVA_OPTS + SPARK_EXECUTOR_MEMORY. - 'org.apache.spark.executor.CoarseGrainedExecutorBackend') - OUR_JAVA_OPTS="$SPARK_JAVA_OPTS $SPARK_EXECUTOR_OPTS" - OUR_JAVA_MEM=${SPARK_EXECUTOR_MEMORY:-$DEFAULT_MEM} - ;; - 'org.apache.spark.executor.MesosExecutorBackend') - OUR_JAVA_OPTS="$SPARK_JAVA_OPTS $SPARK_EXECUTOR_OPTS" - OUR_JAVA_MEM=${SPARK_EXECUTOR_MEMORY:-$DEFAULT_MEM} - export PYTHONPATH="$FWDIR/python:$PYTHONPATH" - export PYTHONPATH="$FWDIR/python/lib/py4j-0.8.2.1-src.zip:$PYTHONPATH" - ;; - - # Spark submit uses SPARK_JAVA_OPTS + SPARK_SUBMIT_OPTS + - # SPARK_DRIVER_MEMORY + SPARK_SUBMIT_DRIVER_MEMORY. - 'org.apache.spark.deploy.SparkSubmit') - OUR_JAVA_OPTS="$SPARK_JAVA_OPTS $SPARK_SUBMIT_OPTS" - OUR_JAVA_MEM=${SPARK_DRIVER_MEMORY:-$DEFAULT_MEM} - if [ -n "$SPARK_SUBMIT_LIBRARY_PATH" ]; then - if [[ $OSTYPE == darwin* ]]; then - export DYLD_LIBRARY_PATH="$SPARK_SUBMIT_LIBRARY_PATH:$DYLD_LIBRARY_PATH" - else - export LD_LIBRARY_PATH="$SPARK_SUBMIT_LIBRARY_PATH:$LD_LIBRARY_PATH" - fi - fi - if [ -n "$SPARK_SUBMIT_DRIVER_MEMORY" ]; then - OUR_JAVA_MEM="$SPARK_SUBMIT_DRIVER_MEMORY" - fi - ;; - - *) - OUR_JAVA_OPTS="$SPARK_JAVA_OPTS" - OUR_JAVA_MEM=${SPARK_DRIVER_MEMORY:-$DEFAULT_MEM} - ;; -esac - # Find the java binary if [ -n "${JAVA_HOME}" ]; then RUNNER="${JAVA_HOME}/bin/java" @@ -110,83 +39,48 @@ else exit 1 fi fi -JAVA_VERSION=$("$RUNNER" -version 2>&1 | grep 'version' | sed 's/.* version "\(.*\)\.\(.*\)\..*"/\1\2/; 1q') - -# Set JAVA_OPTS to be able to load native libraries and to set heap size -if [ "$JAVA_VERSION" -ge 18 ]; then - JAVA_OPTS="$OUR_JAVA_OPTS" -else - JAVA_OPTS="-XX:MaxPermSize=128m $OUR_JAVA_OPTS" -fi -JAVA_OPTS="$JAVA_OPTS -Xms$OUR_JAVA_MEM -Xmx$OUR_JAVA_MEM" - -# Load extra JAVA_OPTS from conf/java-opts, if it exists -if [ -e "$SPARK_CONF_DIR/java-opts" ] ; then - JAVA_OPTS="$JAVA_OPTS `cat "$SPARK_CONF_DIR"/java-opts`" -fi - -# Attention: when changing the way the JAVA_OPTS are assembled, the change must be reflected in CommandUtils.scala! - -TOOLS_DIR="$FWDIR"/tools -SPARK_TOOLS_JAR="" -if [ -e "$TOOLS_DIR"/target/scala-$SPARK_SCALA_VERSION/spark-tools*[0-9Tg].jar ]; then - # Use the JAR from the SBT build - export SPARK_TOOLS_JAR="`ls "$TOOLS_DIR"/target/scala-$SPARK_SCALA_VERSION/spark-tools*[0-9Tg].jar`" -fi -if [ -e "$TOOLS_DIR"/target/spark-tools*[0-9Tg].jar ]; then - # Use the JAR from the Maven build - # TODO: this also needs to become an assembly! - export SPARK_TOOLS_JAR="`ls "$TOOLS_DIR"/target/spark-tools*[0-9Tg].jar`" -fi -# Compute classpath using external script -classpath_output=$("$FWDIR"/bin/compute-classpath.sh) -if [[ "$?" != "0" ]]; then - echo "$classpath_output" - exit 1 -else - CLASSPATH="$classpath_output" -fi +# Look for the launcher. In non-release mode, add the compiled classes directly to the classpath +# instead of looking for a jar file. +SPARK_LAUNCHER_CP= +if [ -f $SPARK_HOME/RELEASE ]; then + LAUNCHER_DIR="$SPARK_HOME/lib" + num_jars="$(ls -1 "$LAUNCHER_DIR" | grep "^spark-launcher.*\.jar$" | wc -l)" + if [ "$num_jars" -eq "0" -a -z "$SPARK_LAUNCHER_CP" ]; then + echo "Failed to find Spark launcher in $LAUNCHER_DIR." 1>&2 + echo "You need to build Spark before running this program." 1>&2 + exit 1 + fi -if [[ "$1" =~ org.apache.spark.tools.* ]]; then - if test -z "$SPARK_TOOLS_JAR"; then - echo "Failed to find Spark Tools Jar in $FWDIR/tools/target/scala-$SPARK_SCALA_VERSION/" 1>&2 - echo "You need to run \"build/sbt tools/package\" before running $1." 1>&2 + LAUNCHER_JARS="$(ls -1 "$LAUNCHER_DIR" | grep "^spark-launcher.*\.jar$" || true)" + if [ "$num_jars" -gt "1" ]; then + echo "Found multiple Spark launcher jars in $LAUNCHER_DIR:" 1>&2 + echo "$LAUNCHER_JARS" 1>&2 + echo "Please remove all but one jar." 1>&2 exit 1 fi - CLASSPATH="$CLASSPATH:$SPARK_TOOLS_JAR" -fi -if $cygwin; then - CLASSPATH="`cygpath -wp "$CLASSPATH"`" - if [ "$1" == "org.apache.spark.tools.JavaAPICompletenessChecker" ]; then - export SPARK_TOOLS_JAR="`cygpath -w "$SPARK_TOOLS_JAR"`" + SPARK_LAUNCHER_CP="${LAUNCHER_DIR}/${LAUNCHER_JARS}" +else + LAUNCHER_DIR="$SPARK_HOME/launcher/target/scala-$SPARK_SCALA_VERSION" + if [ ! -d "$LAUNCHER_DIR/classes" ]; then + echo "Failed to find Spark launcher classes in $LAUNCHER_DIR." 1>&2 + echo "You need to build Spark before running this program." 1>&2 + exit 1 fi + SPARK_LAUNCHER_CP="$LAUNCHER_DIR/classes" fi -export CLASSPATH -# In Spark submit client mode, the driver is launched in the same JVM as Spark submit itself. -# Here we must parse the properties file for relevant "spark.driver.*" configs before launching -# the driver JVM itself. Instead of handling this complexity in Bash, we launch a separate JVM -# to prepare the launch environment of this driver JVM. +# The launcher library will print arguments separated by a NULL character, to allow arguments with +# characters that would be otherwise interpreted by the shell. Read that in a while loop, populating +# an array that will be used to exec the final command. +CMD=() +while IFS= read -d '' -r ARG; do + CMD+=("$ARG") +done < <("$RUNNER" -cp "$SPARK_LAUNCHER_CP" org.apache.spark.launcher.Main "$@") -if [ -n "$SPARK_SUBMIT_BOOTSTRAP_DRIVER" ]; then - # This is used only if the properties file actually contains these special configs - # Export the environment variables needed by SparkSubmitDriverBootstrapper - export RUNNER - export CLASSPATH - export JAVA_OPTS - export OUR_JAVA_MEM - export SPARK_CLASS=1 - shift # Ignore main class (org.apache.spark.deploy.SparkSubmit) and use our own - exec "$RUNNER" org.apache.spark.deploy.SparkSubmitDriverBootstrapper "$@" +if [ "${CMD[0]}" = "usage" ]; then + "${CMD[@]}" else - # Note: The format of this command is closely echoed in SparkSubmitDriverBootstrapper.scala - if [ -n "$SPARK_PRINT_LAUNCH_COMMAND" ]; then - echo -n "Spark Command: " 1>&2 - echo "$RUNNER" -cp "$CLASSPATH" $JAVA_OPTS "$@" 1>&2 - echo -e "========================================\n" 1>&2 - fi - exec "$RUNNER" -cp "$CLASSPATH" $JAVA_OPTS "$@" + exec "${CMD[@]}" fi - diff --git a/bin/spark-class2.cmd b/bin/spark-class2.cmd index da46543647..37d22215a0 100644 --- a/bin/spark-class2.cmd +++ b/bin/spark-class2.cmd @@ -17,135 +17,54 @@ rem See the License for the specific language governing permissions and rem limitations under the License. rem -rem Any changes to this file must be reflected in SparkSubmitDriverBootstrapper.scala! - -setlocal enabledelayedexpansion - -set SCALA_VERSION=2.10 - rem Figure out where the Spark framework is installed -set FWDIR=%~dp0..\ - -rem Export this as SPARK_HOME -set SPARK_HOME=%FWDIR% +set SPARK_HOME=%~dp0.. 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" +if exist "%SPARK_HOME%\conf\spark-env.cmd" call "%SPARK_HOME%\conf\spark-env.cmd" rem Test that an argument was given -if not "x%1"=="x" goto arg_given +if "x%1"=="x" ( echo Usage: spark-class ^<class^> [^<args^>] - goto exit -:arg_given - -if not "x%SPARK_MEM%"=="x" ( - echo Warning: SPARK_MEM is deprecated, please use a more specific config option - echo e.g., spark.executor.memory or spark.driver.memory. + exit /b 1 ) -rem Use SPARK_MEM or 512m as the default memory, to be overridden by specific options -set OUR_JAVA_MEM=%SPARK_MEM% -if "x%OUR_JAVA_MEM%"=="x" set OUR_JAVA_MEM=512m - -set SPARK_DAEMON_JAVA_OPTS=%SPARK_DAEMON_JAVA_OPTS% -Dspark.akka.logLifecycleEvents=true - -rem Add java opts and memory settings for master, worker, history server, executors, and repl. -rem Master, Worker and HistoryServer use SPARK_DAEMON_JAVA_OPTS (and specific opts) + SPARK_DAEMON_MEMORY. -if "%1"=="org.apache.spark.deploy.master.Master" ( - set OUR_JAVA_OPTS=%SPARK_DAEMON_JAVA_OPTS% %SPARK_MASTER_OPTS% - if not "x%SPARK_DAEMON_MEMORY%"=="x" set OUR_JAVA_MEM=%SPARK_DAEMON_MEMORY% -) else if "%1"=="org.apache.spark.deploy.worker.Worker" ( - set OUR_JAVA_OPTS=%SPARK_DAEMON_JAVA_OPTS% %SPARK_WORKER_OPTS% - if not "x%SPARK_DAEMON_MEMORY%"=="x" set OUR_JAVA_MEM=%SPARK_DAEMON_MEMORY% -) else if "%1"=="org.apache.spark.deploy.history.HistoryServer" ( - set OUR_JAVA_OPTS=%SPARK_DAEMON_JAVA_OPTS% %SPARK_HISTORY_OPTS% - if not "x%SPARK_DAEMON_MEMORY%"=="x" set OUR_JAVA_MEM=%SPARK_DAEMON_MEMORY% - -rem Executors use SPARK_JAVA_OPTS + SPARK_EXECUTOR_MEMORY. -) else if "%1"=="org.apache.spark.executor.CoarseGrainedExecutorBackend" ( - set OUR_JAVA_OPTS=%SPARK_JAVA_OPTS% %SPARK_EXECUTOR_OPTS% - if not "x%SPARK_EXECUTOR_MEMORY%"=="x" set OUR_JAVA_MEM=%SPARK_EXECUTOR_MEMORY% -) else if "%1"=="org.apache.spark.executor.MesosExecutorBackend" ( - set OUR_JAVA_OPTS=%SPARK_JAVA_OPTS% %SPARK_EXECUTOR_OPTS% - if not "x%SPARK_EXECUTOR_MEMORY%"=="x" set OUR_JAVA_MEM=%SPARK_EXECUTOR_MEMORY% +set LAUNCHER_CP=0 +if exist %SPARK_HOME%\RELEASE goto find_release_launcher -rem Spark submit uses SPARK_JAVA_OPTS + SPARK_SUBMIT_OPTS + -rem SPARK_DRIVER_MEMORY + SPARK_SUBMIT_DRIVER_MEMORY. -rem The repl also uses SPARK_REPL_OPTS. -) else if "%1"=="org.apache.spark.deploy.SparkSubmit" ( - set OUR_JAVA_OPTS=%SPARK_JAVA_OPTS% %SPARK_SUBMIT_OPTS% %SPARK_REPL_OPTS% - if not "x%SPARK_SUBMIT_LIBRARY_PATH%"=="x" ( - set OUR_JAVA_OPTS=!OUR_JAVA_OPTS! -Djava.library.path=%SPARK_SUBMIT_LIBRARY_PATH% - ) else if not "x%SPARK_LIBRARY_PATH%"=="x" ( - set OUR_JAVA_OPTS=!OUR_JAVA_OPTS! -Djava.library.path=%SPARK_LIBRARY_PATH% - ) - if not "x%SPARK_DRIVER_MEMORY%"=="x" set OUR_JAVA_MEM=%SPARK_DRIVER_MEMORY% - if not "x%SPARK_SUBMIT_DRIVER_MEMORY%"=="x" set OUR_JAVA_MEM=%SPARK_SUBMIT_DRIVER_MEMORY% -) else ( - set OUR_JAVA_OPTS=%SPARK_JAVA_OPTS% - if not "x%SPARK_DRIVER_MEMORY%"=="x" set OUR_JAVA_MEM=%SPARK_DRIVER_MEMORY% +rem Look for the Spark launcher in both Scala build directories. The launcher doesn't use Scala so +rem it doesn't really matter which one is picked up. Add the compiled classes directly to the +rem classpath instead of looking for a jar file, since it's very common for people using sbt to use +rem the "assembly" target instead of "package". +set LAUNCHER_CLASSES=%SPARK_HOME%\launcher\target\scala-2.10\classes +if exist %LAUNCHER_CLASSES% ( + set LAUNCHER_CP=%LAUNCHER_CLASSES% ) - -rem Set JAVA_OPTS to be able to load native libraries and to set heap size -for /f "tokens=3" %%i in ('java -version 2^>^&1 ^| find "version"') do set jversion=%%i -for /f "tokens=1 delims=_" %%i in ("%jversion:~1,-1%") do set jversion=%%i -if "%jversion%" geq "1.8.0" ( - set JAVA_OPTS=%OUR_JAVA_OPTS% -Xms%OUR_JAVA_MEM% -Xmx%OUR_JAVA_MEM% -) else ( - set JAVA_OPTS=-XX:MaxPermSize=128m %OUR_JAVA_OPTS% -Xms%OUR_JAVA_MEM% -Xmx%OUR_JAVA_MEM% +set LAUNCHER_CLASSES=%SPARK_HOME%\launcher\target\scala-2.11\classes +if exist %LAUNCHER_CLASSES% ( + set LAUNCHER_CP=%LAUNCHER_CLASSES% ) -rem Attention: when changing the way the JAVA_OPTS are assembled, the change must be reflected in CommandUtils.scala! +goto check_launcher -rem Test whether the user has built Spark -if exist "%FWDIR%RELEASE" goto skip_build_test -set FOUND_JAR=0 -for %%d in ("%FWDIR%assembly\target\scala-%SCALA_VERSION%\spark-assembly*hadoop*.jar") do ( - set FOUND_JAR=1 -) -if "%FOUND_JAR%"=="0" ( - echo Failed to find Spark assembly JAR. - echo You need to build Spark before running this program. - goto exit +:find_release_launcher +for %%d in (%SPARK_HOME%\lib\spark-launcher*.jar) do ( + set LAUNCHER_CP=%%d ) -:skip_build_test -set TOOLS_DIR=%FWDIR%tools -set SPARK_TOOLS_JAR= -for %%d in ("%TOOLS_DIR%\target\scala-%SCALA_VERSION%\spark-tools*assembly*.jar") do ( - set SPARK_TOOLS_JAR=%%d +:check_launcher +if "%LAUNCHER_CP%"=="0" ( + echo Failed to find Spark launcher JAR. + echo You need to build Spark before running this program. + exit /b 1 ) -rem Compute classpath using external script -set DONT_PRINT_CLASSPATH=1 -call "%FWDIR%bin\compute-classpath.cmd" -set DONT_PRINT_CLASSPATH=0 -set CLASSPATH=%CLASSPATH%;%SPARK_TOOLS_JAR% - rem Figure out where java is. set RUNNER=java if not "x%JAVA_HOME%"=="x" set RUNNER=%JAVA_HOME%\bin\java -rem In Spark submit client mode, the driver is launched in the same JVM as Spark submit itself. -rem Here we must parse the properties file for relevant "spark.driver.*" configs before launching -rem the driver JVM itself. Instead of handling this complexity here, we launch a separate JVM -rem to prepare the launch environment of this driver JVM. - -rem In this case, leave out the main class (org.apache.spark.deploy.SparkSubmit) and use our own. -rem Leaving out the first argument is surprisingly difficult to do in Windows. Note that this must -rem be done here because the Windows "shift" command does not work in a conditional block. -set BOOTSTRAP_ARGS= -shift -:start_parse -if "%~1" == "" goto end_parse -set BOOTSTRAP_ARGS=%BOOTSTRAP_ARGS% %~1 -shift -goto start_parse -:end_parse - -if not [%SPARK_SUBMIT_BOOTSTRAP_DRIVER%] == [] ( - set SPARK_CLASS=1 - "%RUNNER%" org.apache.spark.deploy.SparkSubmitDriverBootstrapper %BOOTSTRAP_ARGS% -) else ( - "%RUNNER%" -cp "%CLASSPATH%" %JAVA_OPTS% %* +rem The launcher library prints the command to be executed in a single line suitable for being +rem executed by the batch interpreter. So read all the output of the launcher into a variable. +for /f "tokens=*" %%i in ('cmd /C ""%RUNNER%" -cp %LAUNCHER_CP% org.apache.spark.launcher.Main %*"') do ( + set SPARK_CMD=%%i ) -:exit +%SPARK_CMD% diff --git a/bin/spark-shell b/bin/spark-shell index cca5aa0676..b3761b5e13 100755 --- a/bin/spark-shell +++ b/bin/spark-shell @@ -28,25 +28,24 @@ esac # Enter posix mode for bash set -o posix -## Global script variables -FWDIR="$(cd "`dirname "$0"`"/..; pwd)" +export FWDIR="$(cd "`dirname "$0"`"/..; pwd)" -function usage() { +usage() { + if [ -n "$1" ]; then + echo "$1" + fi echo "Usage: ./bin/spark-shell [options]" "$FWDIR"/bin/spark-submit --help 2>&1 | grep -v Usage 1>&2 - exit 0 + exit "$2" } +export -f usage if [[ "$@" = *--help ]] || [[ "$@" = *-h ]]; then - usage + usage "" 0 fi -source "$FWDIR"/bin/utils.sh -SUBMIT_USAGE_FUNCTION=usage -gatherSparkSubmitOpts "$@" - # SPARK-4161: scala does not assume use of the java classpath, -# so we need to add the "-Dscala.usejavacp=true" flag mnually. We +# so we need to add the "-Dscala.usejavacp=true" flag manually. We # do this specifically for the Spark shell because the scala REPL # has its own class loader, and any additional classpath specified # through spark.driver.extraClassPath is not automatically propagated. @@ -61,11 +60,11 @@ function main() { # (see https://github.com/sbt/sbt/issues/562). stty -icanon min 1 -echo > /dev/null 2>&1 export SPARK_SUBMIT_OPTS="$SPARK_SUBMIT_OPTS -Djline.terminal=unix" - "$FWDIR"/bin/spark-submit --class org.apache.spark.repl.Main "${SUBMISSION_OPTS[@]}" spark-shell "${APPLICATION_OPTS[@]}" + "$FWDIR"/bin/spark-submit --class org.apache.spark.repl.Main "$@" stty icanon echo > /dev/null 2>&1 else export SPARK_SUBMIT_OPTS - "$FWDIR"/bin/spark-submit --class org.apache.spark.repl.Main "${SUBMISSION_OPTS[@]}" spark-shell "${APPLICATION_OPTS[@]}" + "$FWDIR"/bin/spark-submit --class org.apache.spark.repl.Main "$@" fi } diff --git a/bin/spark-shell2.cmd b/bin/spark-shell2.cmd index 1d1a40da31..02f51fe59a 100644 --- a/bin/spark-shell2.cmd +++ b/bin/spark-shell2.cmd @@ -25,17 +25,28 @@ if %ERRORLEVEL% equ 0 ( exit /b 0 ) -call %SPARK_HOME%\bin\windows-utils.cmd %* -if %ERRORLEVEL% equ 1 ( +rem SPARK-4161: scala does not assume use of the java classpath, +rem so we need to add the "-Dscala.usejavacp=true" flag manually. We +rem do this specifically for the Spark shell because the scala REPL +rem has its own class loader, and any additional classpath specified +rem through spark.driver.extraClassPath is not automatically propagated. +if "x%SPARK_SUBMIT_OPTS%"=="x" ( + set SPARK_SUBMIT_OPTS=-Dscala.usejavacp=true + goto run_shell +) +set SPARK_SUBMIT_OPTS="%SPARK_SUBMIT_OPTS% -Dscala.usejavacp=true" + +:run_shell +call %SPARK_HOME%\bin\spark-submit2.cmd --class org.apache.spark.repl.Main %* +set SPARK_ERROR_LEVEL=%ERRORLEVEL% +if not "x%SPARK_LAUNCHER_USAGE_ERROR%"=="x" ( call :usage exit /b 1 ) - -cmd /V /E /C %SPARK_HOME%\bin\spark-submit.cmd --class org.apache.spark.repl.Main %SUBMISSION_OPTS% spark-shell %APPLICATION_OPTS% - -exit /b 0 +exit /b %SPARK_ERROR_LEVEL% :usage +echo %SPARK_LAUNCHER_USAGE_ERROR% echo "Usage: .\bin\spark-shell.cmd [options]" >&2 -%SPARK_HOME%\bin\spark-submit --help 2>&1 | findstr /V "Usage" 1>&2 -exit /b 0 +call %SPARK_HOME%\bin\spark-submit2.cmd --help 2>&1 | findstr /V "Usage" 1>&2 +goto :eof diff --git a/bin/spark-sql b/bin/spark-sql index 3b6cc420fe..ca1729f4cf 100755 --- a/bin/spark-sql +++ b/bin/spark-sql @@ -25,12 +25,15 @@ set -o posix # NOTE: This exact class name is matched downstream by SparkSubmit. # Any changes need to be reflected there. -CLASS="org.apache.spark.sql.hive.thriftserver.SparkSQLCLIDriver" +export CLASS="org.apache.spark.sql.hive.thriftserver.SparkSQLCLIDriver" # Figure out where Spark is installed -FWDIR="$(cd "`dirname "$0"`"/..; pwd)" +export FWDIR="$(cd "`dirname "$0"`"/..; pwd)" function usage { + if [ -n "$1" ]; then + echo "$1" + fi echo "Usage: ./bin/spark-sql [options] [cli option]" pattern="usage" pattern+="\|Spark assembly has been built with Hive" @@ -42,16 +45,13 @@ function usage { "$FWDIR"/bin/spark-submit --help 2>&1 | grep -v Usage 1>&2 echo echo "CLI options:" - "$FWDIR"/bin/spark-class $CLASS --help 2>&1 | grep -v "$pattern" 1>&2 + "$FWDIR"/bin/spark-class "$CLASS" --help 2>&1 | grep -v "$pattern" 1>&2 + exit "$2" } +export -f usage if [[ "$@" = *--help ]] || [[ "$@" = *-h ]]; then - usage - exit 0 + usage "" 0 fi -source "$FWDIR"/bin/utils.sh -SUBMIT_USAGE_FUNCTION=usage -gatherSparkSubmitOpts "$@" - -exec "$FWDIR"/bin/spark-submit --class $CLASS "${SUBMISSION_OPTS[@]}" spark-internal "${APPLICATION_OPTS[@]}" +exec "$FWDIR"/bin/spark-submit --class "$CLASS" "$@" diff --git a/bin/spark-submit b/bin/spark-submit index 3e5cbdbb24..bcff78edd5 100755 --- a/bin/spark-submit +++ b/bin/spark-submit @@ -17,58 +17,18 @@ # limitations under the License. # -# NOTE: Any changes in this file must be reflected in SparkSubmitDriverBootstrapper.scala! - -export SPARK_HOME="$(cd "`dirname "$0"`"/..; pwd)" -ORIG_ARGS=("$@") - -# Set COLUMNS for progress bar -export COLUMNS=`tput cols` - -while (($#)); do - if [ "$1" = "--deploy-mode" ]; then - SPARK_SUBMIT_DEPLOY_MODE=$2 - elif [ "$1" = "--properties-file" ]; then - SPARK_SUBMIT_PROPERTIES_FILE=$2 - elif [ "$1" = "--driver-memory" ]; then - export SPARK_SUBMIT_DRIVER_MEMORY=$2 - elif [ "$1" = "--driver-library-path" ]; then - export SPARK_SUBMIT_LIBRARY_PATH=$2 - elif [ "$1" = "--driver-class-path" ]; then - export SPARK_SUBMIT_CLASSPATH=$2 - elif [ "$1" = "--driver-java-options" ]; then - export SPARK_SUBMIT_OPTS=$2 - elif [ "$1" = "--master" ]; then - export MASTER=$2 - fi - shift -done - -if [ -z "$SPARK_CONF_DIR" ]; then - export SPARK_CONF_DIR="$SPARK_HOME/conf" -fi -DEFAULT_PROPERTIES_FILE="$SPARK_CONF_DIR/spark-defaults.conf" -if [ "$MASTER" == "yarn-cluster" ]; then - SPARK_SUBMIT_DEPLOY_MODE=cluster +SPARK_HOME="$(cd "`dirname "$0"`"/..; pwd)" + +# Only define a usage function if an upstream script hasn't done so. +if ! type -t usage >/dev/null 2>&1; then + usage() { + if [ -n "$1" ]; then + echo "$1" + fi + "$SPARK_HOME"/bin/spark-class org.apache.spark.deploy.SparkSubmit --help + exit "$2" + } + export -f usage fi -export SPARK_SUBMIT_DEPLOY_MODE=${SPARK_SUBMIT_DEPLOY_MODE:-"client"} -export SPARK_SUBMIT_PROPERTIES_FILE=${SPARK_SUBMIT_PROPERTIES_FILE:-"$DEFAULT_PROPERTIES_FILE"} - -# For client mode, the driver will be launched in the same JVM that launches -# SparkSubmit, so we may need to read the properties file for any extra class -# paths, library paths, java options and memory early on. Otherwise, it will -# be too late by the time the driver JVM has started. - -if [[ "$SPARK_SUBMIT_DEPLOY_MODE" == "client" && -f "$SPARK_SUBMIT_PROPERTIES_FILE" ]]; then - # Parse the properties file only if the special configs exist - contains_special_configs=$( - grep -e "spark.driver.extra*\|spark.driver.memory" "$SPARK_SUBMIT_PROPERTIES_FILE" | \ - grep -v "^[[:space:]]*#" - ) - if [ -n "$contains_special_configs" ]; then - export SPARK_SUBMIT_BOOTSTRAP_DRIVER=1 - fi -fi - -exec "$SPARK_HOME"/bin/spark-class org.apache.spark.deploy.SparkSubmit "${ORIG_ARGS[@]}" +exec "$SPARK_HOME"/bin/spark-class org.apache.spark.deploy.SparkSubmit "$@" diff --git a/bin/spark-submit2.cmd b/bin/spark-submit2.cmd index 446cbc74b7..08ddb18574 100644 --- a/bin/spark-submit2.cmd +++ b/bin/spark-submit2.cmd @@ -17,62 +17,19 @@ rem See the License for the specific language governing permissions and rem limitations under the License. rem -rem NOTE: Any changes in this file must be reflected in SparkSubmitDriverBootstrapper.scala! - -set SPARK_HOME=%~dp0.. -set ORIG_ARGS=%* - -rem Reset the values of all variables used -set SPARK_SUBMIT_DEPLOY_MODE=client - -if [%SPARK_CONF_DIR%] == [] ( - set SPARK_CONF_DIR=%SPARK_HOME%\conf -) -set SPARK_SUBMIT_PROPERTIES_FILE=%SPARK_CONF_DIR%\spark-defaults.conf -set SPARK_SUBMIT_DRIVER_MEMORY= -set SPARK_SUBMIT_LIBRARY_PATH= -set SPARK_SUBMIT_CLASSPATH= -set SPARK_SUBMIT_OPTS= -set SPARK_SUBMIT_BOOTSTRAP_DRIVER= - -:loop -if [%1] == [] goto continue - if [%1] == [--deploy-mode] ( - set SPARK_SUBMIT_DEPLOY_MODE=%2 - ) else if [%1] == [--properties-file] ( - set SPARK_SUBMIT_PROPERTIES_FILE=%2 - ) else if [%1] == [--driver-memory] ( - set SPARK_SUBMIT_DRIVER_MEMORY=%2 - ) else if [%1] == [--driver-library-path] ( - set SPARK_SUBMIT_LIBRARY_PATH=%2 - ) else if [%1] == [--driver-class-path] ( - set SPARK_SUBMIT_CLASSPATH=%2 - ) else if [%1] == [--driver-java-options] ( - set SPARK_SUBMIT_OPTS=%2 - ) else if [%1] == [--master] ( - set MASTER=%2 - ) - shift -goto loop -:continue - -if [%MASTER%] == [yarn-cluster] ( - set SPARK_SUBMIT_DEPLOY_MODE=cluster -) - -rem For client mode, the driver will be launched in the same JVM that launches -rem SparkSubmit, so we may need to read the properties file for any extra class -rem paths, library paths, java options and memory early on. Otherwise, it will -rem be too late by the time the driver JVM has started. - -if [%SPARK_SUBMIT_DEPLOY_MODE%] == [client] ( - if exist %SPARK_SUBMIT_PROPERTIES_FILE% ( - rem Parse the properties file only if the special configs exist - for /f %%i in ('findstr /r /c:"^[\t ]*spark.driver.memory" /c:"^[\t ]*spark.driver.extra" ^ - %SPARK_SUBMIT_PROPERTIES_FILE%') do ( - set SPARK_SUBMIT_BOOTSTRAP_DRIVER=1 - ) - ) +rem This is the entry point for running Spark submit. To avoid polluting the +rem environment, it just launches a new cmd to do the real work. + +set CLASS=org.apache.spark.deploy.SparkSubmit +call %~dp0spark-class2.cmd %CLASS% %* +set SPARK_ERROR_LEVEL=%ERRORLEVEL% +if not "x%SPARK_LAUNCHER_USAGE_ERROR%"=="x" ( + call :usage + exit /b 1 ) +exit /b %SPARK_ERROR_LEVEL% -cmd /V /E /C %SPARK_HOME%\bin\spark-class.cmd org.apache.spark.deploy.SparkSubmit %ORIG_ARGS% +:usage +echo %SPARK_LAUNCHER_USAGE_ERROR% +call %SPARK_HOME%\bin\spark-class2.cmd %CLASS% --help +goto :eof diff --git a/bin/utils.sh b/bin/utils.sh deleted file mode 100755 index 748dbe345a..0000000000 --- a/bin/utils.sh +++ /dev/null @@ -1,60 +0,0 @@ -#!/usr/bin/env bash - -# -# Licensed to the Apache Software Foundation (ASF) under one or more -# contributor license agreements. See the NOTICE file distributed with -# this work for additional information regarding copyright ownership. -# The ASF licenses this file to You under the Apache License, Version 2.0 -# (the "License"); you may not use this file except in compliance with -# the License. You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. -# - -# Gather all spark-submit options into SUBMISSION_OPTS -function gatherSparkSubmitOpts() { - - if [ -z "$SUBMIT_USAGE_FUNCTION" ]; then - echo "Function for printing usage of $0 is not set." 1>&2 - echo "Please set usage function to shell variable 'SUBMIT_USAGE_FUNCTION' in $0" 1>&2 - exit 1 - fi - - # NOTE: If you add or remove spark-submit options, - # modify NOT ONLY this script but also SparkSubmitArgument.scala - SUBMISSION_OPTS=() - APPLICATION_OPTS=() - while (($#)); do - case "$1" in - --master | --deploy-mode | --class | --name | --jars | --packages | --py-files | --files | \ - --conf | --repositories | --properties-file | --driver-memory | --driver-java-options | \ - --driver-library-path | --driver-class-path | --executor-memory | --driver-cores | \ - --total-executor-cores | --executor-cores | --queue | --num-executors | --archives | \ - --proxy-user) - if [[ $# -lt 2 ]]; then - "$SUBMIT_USAGE_FUNCTION" - exit 1; - fi - SUBMISSION_OPTS+=("$1"); shift - SUBMISSION_OPTS+=("$1"); shift - ;; - - --verbose | -v | --supervise) - SUBMISSION_OPTS+=("$1"); shift - ;; - - *) - APPLICATION_OPTS+=("$1"); shift - ;; - esac - done - - export SUBMISSION_OPTS - export APPLICATION_OPTS -} diff --git a/bin/windows-utils.cmd b/bin/windows-utils.cmd deleted file mode 100644 index 0cf9e87ca5..0000000000 --- a/bin/windows-utils.cmd +++ /dev/null @@ -1,60 +0,0 @@ -rem -rem Licensed to the Apache Software Foundation (ASF) under one or more -rem contributor license agreements. See the NOTICE file distributed with -rem this work for additional information regarding copyright ownership. -rem The ASF licenses this file to You under the Apache License, Version 2.0 -rem (the "License"); you may not use this file except in compliance with -rem the License. You may obtain a copy of the License at -rem -rem http://www.apache.org/licenses/LICENSE-2.0 -rem -rem Unless required by applicable law or agreed to in writing, software -rem distributed under the License is distributed on an "AS IS" BASIS, -rem WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -rem See the License for the specific language governing permissions and -rem limitations under the License. -rem - -rem Gather all spark-submit options into SUBMISSION_OPTS - -set SUBMISSION_OPTS= -set APPLICATION_OPTS= - -rem NOTE: If you add or remove spark-sumbmit options, -rem modify NOT ONLY this script but also SparkSubmitArgument.scala - -:OptsLoop -if "x%1"=="x" ( - goto :OptsLoopEnd -) - -SET opts="\<--master\> \<--deploy-mode\> \<--class\> \<--name\> \<--jars\> \<--py-files\> \<--files\>" -SET opts="%opts:~1,-1% \<--conf\> \<--properties-file\> \<--driver-memory\> \<--driver-java-options\>" -SET opts="%opts:~1,-1% \<--driver-library-path\> \<--driver-class-path\> \<--executor-memory\>" -SET opts="%opts:~1,-1% \<--driver-cores\> \<--total-executor-cores\> \<--executor-cores\> \<--queue\>" -SET opts="%opts:~1,-1% \<--num-executors\> \<--archives\> \<--packages\> \<--repositories\>" -SET opts="%opts:~1,-1% \<--proxy-user\>" - -echo %1 | findstr %opts% >nul -if %ERRORLEVEL% equ 0 ( - if "x%2"=="x" ( - echo "%1" requires an argument. >&2 - exit /b 1 - ) - set SUBMISSION_OPTS=%SUBMISSION_OPTS% %1 %2 - shift - shift - goto :OptsLoop -) -echo %1 | findstr "\<--verbose\> \<-v\> \<--supervise\>" >nul -if %ERRORLEVEL% equ 0 ( - set SUBMISSION_OPTS=%SUBMISSION_OPTS% %1 - shift - goto :OptsLoop -) -set APPLICATION_OPTS=%APPLICATION_OPTS% %1 -shift -goto :OptsLoop - -:OptsLoopEnd -exit /b 0 |