diff --git a/README.md b/README.md
index 8f686743f042ba8dbb6b9d2b672357b83ad9c864..873ec9882f9a8befd103363214338dfb2c6f2ad7 100644
--- a/README.md
+++ b/README.md
@@ -13,7 +13,7 @@ This README file only contains basic setup instructions.
## Building
Spark requires Scala 2.10. The project is built using Simple Build Tool (SBT),
-which can be obtained from [here](http://www.scala-sbt.org). To build Spark and its example programs, run:
+which can be obtained [here](http://www.scala-sbt.org). To build Spark and its example programs, run:
sbt assembly
@@ -38,24 +38,11 @@ locally with one thread, or "local[N]" to run locally with N threads.
## Running tests
-### With sbt (Much faster to run compared to maven)
-Once you have built spark with `sbt assembly` mentioned in [Building](#Building) section. Test suits can be run as follows using sbt.
+Testing first requires [Building](#Building) Spark. Once Spark is built, tests
+can be run using:
`sbt test`
-### With maven.
-1. Export these necessary environment variables as follows.
-
- `export SCALA_HOME=<scala distribution>`
-
- `export MAVEN_OPTS="-Xmx1512m -XX:MaxPermSize=512m"`
-
-2. Build assembly by
-`mvn package -DskipTests`
-
-3. Run tests
-`mvn test`
-
## A Note About Hadoop Versions
Spark uses the Hadoop core library to talk to HDFS and other Hadoop-supported
diff --git a/core/src/main/scala/org/apache/spark/SparkContext.scala b/core/src/main/scala/org/apache/spark/SparkContext.scala
index c6f6883b011527d17917625c16967318835b8a62..e80e43af6d4e78478d882c5e935805ccacea5216 100644
--- a/core/src/main/scala/org/apache/spark/SparkContext.scala
+++ b/core/src/main/scala/org/apache/spark/SparkContext.scala
@@ -173,7 +173,8 @@ class SparkContext(
value <- Option(System.getenv(key))) {
executorEnvs(key) = value
}
- // A workaround for SPARK_TESTING and SPARK_HOME
+ // Convert java options to env vars as a work around
+ // since we can't set env vars directly in sbt.
for { (envKey, propKey) <- Seq(("SPARK_HOME", "spark.home"), ("SPARK_TESTING", "spark.testing"))
value <- Option(System.getenv(envKey)).orElse(Option(System.getProperty(propKey)))} {
executorEnvs(envKey) = value
diff --git a/core/src/test/scala/org/apache/spark/DriverSuite.scala b/core/src/test/scala/org/apache/spark/DriverSuite.scala
index 89c5631ad8f07a49fc39e85820bb9b5d800cde42..7e1e55fa3b29962426d4289399e2195ea63f107e 100644
--- a/core/src/test/scala/org/apache/spark/DriverSuite.scala
+++ b/core/src/test/scala/org/apache/spark/DriverSuite.scala
@@ -30,8 +30,7 @@ import org.apache.spark.util.Utils
class DriverSuite extends FunSuite with Timeouts {
test("driver should exit after finishing") {
- val sparkHome = Option(System.getenv("SPARK_HOME"))
- .orElse(Option(System.getProperty("spark.home"))).get
+ val sparkHome = sys.env.get("SPARK_HOME").orElse(sys.props.get("spark.home")).get
// Regression test for SPARK-530: "Spark driver process doesn't exit after finishing"
val masters = Table(("master"), ("local"), ("local-cluster[2,1,512]"))
forAll(masters) { (master: String) =>
diff --git a/core/src/test/scala/org/apache/spark/FileServerSuite.scala b/core/src/test/scala/org/apache/spark/FileServerSuite.scala
index 506f7484fb7316f61223143de4ea207819cbe302..a2eb9a4e84696c9326c5523e673add86b9d620d8 100644
--- a/core/src/test/scala/org/apache/spark/FileServerSuite.scala
+++ b/core/src/test/scala/org/apache/spark/FileServerSuite.scala
@@ -27,54 +27,39 @@ import org.scalatest.FunSuite
class FileServerSuite extends FunSuite with LocalSparkContext {
@transient var tmpFile: File = _
- @transient var testJarFile: String = _
-
+ @transient var tmpJarUrl: String = _
override def beforeAll() {
super.beforeAll()
- val buffer = new Array[Byte](10240)
- val tmpdir = new File(Files.createTempDir(), "test")
- tmpdir.mkdir()
- val tmpJarEntry = new File(tmpdir, "FileServerSuite2.txt")
- val pw = new PrintWriter(tmpJarEntry)
- pw.println("test String in the file named FileServerSuite2.txt")
+ val tmpDir = new File(Files.createTempDir(), "test")
+ tmpDir.mkdir()
+
+ val textFile = new File(tmpDir, "FileServerSuite.txt")
+ val pw = new PrintWriter(textFile)
+ pw.println("100")
pw.close()
- // The ugliest code possible, was translated from java.
- val tmpFile2 = new File(tmpdir, "test.jar")
- val stream = new FileOutputStream(tmpFile2)
- val jar = new JarOutputStream(stream, new java.util.jar.Manifest())
- val jarAdd = new JarEntry(tmpJarEntry.getName)
- jarAdd.setTime(tmpJarEntry.lastModified)
- jar.putNextEntry(jarAdd)
- val in = new FileInputStream(tmpJarEntry)
+
+ val jarFile = new File(tmpDir, "test.jar")
+ val jarStream = new FileOutputStream(jarFile)
+ val jar = new JarOutputStream(jarStream, new java.util.jar.Manifest())
+
+ val jarEntry = new JarEntry(textFile.getName)
+ jar.putNextEntry(jarEntry)
+
+ val in = new FileInputStream(textFile)
+ val buffer = new Array[Byte](10240)
var nRead = 0
- while (nRead <= 0) {
+ while (nRead <= 0) {
nRead = in.read(buffer, 0, buffer.length)
jar.write(buffer, 0, nRead)
}
+
in.close()
jar.close()
- stream.close()
- testJarFile = tmpFile2.toURI.toURL.toString
- }
-
- override def beforeEach() {
- super.beforeEach()
- // Create a sample text file
- val tmpdir = new File(Files.createTempDir(), "test")
- tmpdir.mkdir()
- tmpFile = new File(tmpdir, "FileServerSuite.txt")
- val pw = new PrintWriter(tmpFile)
- pw.println("100")
- pw.close()
- }
+ jarStream.close()
- override def afterEach() {
- super.afterEach()
- // Clean up downloaded file
- if (tmpFile.exists) {
- tmpFile.delete()
- }
+ tmpFile = textFile
+ tmpJarUrl = jarFile.toURI.toURL.toString
}
test("Distributing files locally") {
@@ -108,10 +93,10 @@ class FileServerSuite extends FunSuite with LocalSparkContext {
test ("Dynamically adding JARS locally") {
sc = new SparkContext("local[4]", "test")
- sc.addJar(testJarFile)
+ sc.addJar(tmpJarUrl)
val testData = Array((1, 1))
- sc.parallelize(testData).foreach { (x) =>
- if (Thread.currentThread.getContextClassLoader.getResource("FileServerSuite2.txt") == null) {
+ sc.parallelize(testData).foreach { x =>
+ if (Thread.currentThread.getContextClassLoader.getResource("FileServerSuite.txt") == null) {
throw new SparkException("jar not added")
}
}
@@ -133,10 +118,10 @@ class FileServerSuite extends FunSuite with LocalSparkContext {
test ("Dynamically adding JARS on a standalone cluster") {
sc = new SparkContext("local-cluster[1,1,512]", "test")
- sc.addJar(testJarFile)
+ sc.addJar(tmpJarUrl)
val testData = Array((1,1))
- sc.parallelize(testData).foreach { (x) =>
- if (Thread.currentThread.getContextClassLoader.getResource("FileServerSuite2.txt") == null) {
+ sc.parallelize(testData).foreach { x =>
+ if (Thread.currentThread.getContextClassLoader.getResource("FileServerSuite.txt") == null) {
throw new SparkException("jar not added")
}
}
@@ -144,10 +129,10 @@ class FileServerSuite extends FunSuite with LocalSparkContext {
test ("Dynamically adding JARS on a standalone cluster using local: URL") {
sc = new SparkContext("local-cluster[1,1,512]", "test")
- sc.addJar(testJarFile.replace("file", "local"))
+ sc.addJar(tmpJarUrl.replace("file", "local"))
val testData = Array((1,1))
- sc.parallelize(testData).foreach { (x) =>
- if (Thread.currentThread.getContextClassLoader.getResource("FileServerSuite2.txt") == null) {
+ sc.parallelize(testData).foreach { x =>
+ if (Thread.currentThread.getContextClassLoader.getResource("FileServerSuite.txt") == null) {
throw new SparkException("jar not added")
}
}
diff --git a/core/src/test/scala/org/apache/spark/deploy/worker/ExecutorRunnerTest.scala b/core/src/test/scala/org/apache/spark/deploy/worker/ExecutorRunnerTest.scala
index 894a72284b3fe840893d0e55b29a59be8294b940..f58b1ee05a64b5f1ac21cbbb3912d5e23d818ac6 100644
--- a/core/src/test/scala/org/apache/spark/deploy/worker/ExecutorRunnerTest.scala
+++ b/core/src/test/scala/org/apache/spark/deploy/worker/ExecutorRunnerTest.scala
@@ -19,18 +19,14 @@ package org.apache.spark.deploy.worker
import java.io.File
-import scala.util.Try
-
import org.scalatest.FunSuite
import org.apache.spark.deploy.{ExecutorState, Command, ApplicationDescription}
-
class ExecutorRunnerTest extends FunSuite {
test("command includes appId") {
def f(s:String) = new File(s)
- val sparkHome = Try(sys.env("SPARK_HOME")).toOption
- .orElse(Option(System.getProperty("spark.home"))).get
+ val sparkHome = sys.env.get("SPARK_HOME").orElse(sys.env.get("spark.home")).get
val appDesc = new ApplicationDescription("app name", 8, 500, Command("foo", Seq(),Map()),
sparkHome, "appUiUrl")
val appId = "12345-worker321-9876"
diff --git a/make-distribution.sh b/make-distribution.sh
index a2c8e645971432f75251efa665c82ee7deb7c124..8ae8a4cf742a161451ca13c12c08282b17f69382 100755
--- a/make-distribution.sh
+++ b/make-distribution.sh
@@ -45,7 +45,8 @@ DISTDIR="$FWDIR/dist"
export TERM=dumb # Prevents color codes in SBT output
if ! test `which sbt` ;then
- echo -e "You need sbt installed and available on path, please follow the instructions here: http://www.scala-sbt.org/release/docs/Getting-Started/Setup.html"
+ echo -e "You need sbt installed and available on your path."
+ echo -e "Download sbt from http://www.scala-sbt.org/"
exit -1;
fi
diff --git a/project/SparkBuild.scala b/project/SparkBuild.scala
index 8290e7cf438c743e28674ad9fe3b2a4a84f4d723..5f57c964bde615b5323d62da2cb703b5b2a7a7a8 100644
--- a/project/SparkBuild.scala
+++ b/project/SparkBuild.scala
@@ -72,6 +72,7 @@ object SparkBuild extends Build {
val sparkHome = System.getProperty("user.dir")
System.setProperty("spark.home", sparkHome)
System.setProperty("spark.testing", "1")
+
// Allows build configuration to be set through environment variables
lazy val hadoopVersion = Properties.envOrElse("SPARK_HADOOP_VERSION", DEFAULT_HADOOP_VERSION)
lazy val isNewHadoop = Properties.envOrNone("SPARK_IS_NEW_HADOOP") match {
diff --git a/python/lib/py4j-0.8.1-src.zip b/python/lib/py4j-0.8.1-src.zip
new file mode 100644
index 0000000000000000000000000000000000000000..2069a328d1f2e6a94df057c6a3930048ae3f3832
Binary files /dev/null and b/python/lib/py4j-0.8.1-src.zip differ
diff --git a/python/lib/py4j-0.8.1.zip b/python/lib/py4j-0.8.1.zip
deleted file mode 100644
index 3231e31164e51daad4254e8f09a49336205563b8..0000000000000000000000000000000000000000
Binary files a/python/lib/py4j-0.8.1.zip and /dev/null differ
diff --git a/python/pyspark/__init__.py b/python/pyspark/__init__.py
index 2b2c3a061a71d49b0bc5f9815be1bfa2d90cb191..a51d5af79b0dafd6e685b21d76efd69405e613aa 100644
--- a/python/pyspark/__init__.py
+++ b/python/pyspark/__init__.py
@@ -40,7 +40,7 @@ Public classes:
import sys
import os
-sys.path.insert(0, os.path.join(os.environ["SPARK_HOME"], "python/lib/py4j0.7.egg"))
+sys.path.insert(0, os.path.join(os.environ["SPARK_HOME"], "python/lib/py4j-0.8.1-src.zip"))
from pyspark.conf import SparkConf