Merge branch 'master' into spark-2706

Author: zhzhan

core/pom.xml

@@ -306,26 +306,20 @@
       </plugin>
       <!-- Unzip py4j so we can include its files in the jar -->
       <plugin>
-        <groupId>org.codehaus.mojo</groupId>
-        <artifactId>exec-maven-plugin</artifactId>
-        <version>1.2.1</version>
+        <groupId>org.apache.maven.plugins</groupId>
+        <artifactId>maven-antrun-plugin</artifactId>
         <executions>
           <execution>
             <phase>generate-resources</phase>
             <goals>
-              <goal>exec</goal>
+              <goal>run</goal>
             </goals>
           </execution>
         </executions>
         <configuration>
-          <executable>unzip</executable>
-          <workingDirectory>../python</workingDirectory>
-          <arguments>
-            <argument>-o</argument>
-            <argument>lib/py4j*.zip</argument>
-            <argument>-d</argument>
-            <argument>build</argument>
-          </arguments>
+          <tasks>
+            <unzip src="../python/lib/py4j-0.8.2.1-src.zip" dest="../python/build" />
+          </tasks>
         </configuration>
       </plugin>
       <plugin>

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

@@ -149,6 +149,9 @@ class SimpleFutureAction[T] private[spark](jobWaiter: JobWaiter[_], resultFunc:
       case JobFailed(e: Exception) => scala.util.Failure(e)
     }
   }
+
+  /** Get the corresponding job id for this action. */
+  def jobId = jobWaiter.jobId
 }
 
 

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

@@ -294,7 +294,7 @@ private[spark] class SecurityManager(sparkConf: SparkConf) extends Logging {
   def checkUIViewPermissions(user: String): Boolean = {
     logDebug("user=" + user + " aclsEnabled=" + aclsEnabled() + " viewAcls=" +
       viewAcls.mkString(","))
-    if (aclsEnabled() && (user != null) && (!viewAcls.contains(user))) false else true
+    !aclsEnabled || user == null || viewAcls.contains(user)
   }
 
   /**
@@ -309,7 +309,7 @@ private[spark] class SecurityManager(sparkConf: SparkConf) extends Logging {
   def checkModifyPermissions(user: String): Boolean = {
     logDebug("user=" + user + " aclsEnabled=" + aclsEnabled() + " modifyAcls=" +
       modifyAcls.mkString(","))
-    if (aclsEnabled() && (user != null) && (!modifyAcls.contains(user))) false else true
+    !aclsEnabled || user == null || modifyAcls.contains(user)
   }
 
 

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

@@ -224,26 +224,7 @@ class SparkContext(config: SparkConf) extends Logging {
   ui.bind()
 
   /** A default Hadoop Configuration for the Hadoop code (e.g. file systems) that we reuse. */
-  val hadoopConfiguration: Configuration = {
-    val hadoopConf = SparkHadoopUtil.get.newConfiguration()
-    // Explicitly check for S3 environment variables
-    if (System.getenv("AWS_ACCESS_KEY_ID") != null &&
-        System.getenv("AWS_SECRET_ACCESS_KEY") != null) {
-      hadoopConf.set("fs.s3.awsAccessKeyId", System.getenv("AWS_ACCESS_KEY_ID"))
-      hadoopConf.set("fs.s3n.awsAccessKeyId", System.getenv("AWS_ACCESS_KEY_ID"))
-      hadoopConf.set("fs.s3.awsSecretAccessKey", System.getenv("AWS_SECRET_ACCESS_KEY"))
-      hadoopConf.set("fs.s3n.awsSecretAccessKey", System.getenv("AWS_SECRET_ACCESS_KEY"))
-    }
-    // Copy any "spark.hadoop.foo=bar" system properties into conf as "foo=bar"
-    conf.getAll.foreach { case (key, value) =>
-      if (key.startsWith("spark.hadoop.")) {
-        hadoopConf.set(key.substring("spark.hadoop.".length), value)
-      }
-    }
-    val bufferSize = conf.get("spark.buffer.size", "65536")
-    hadoopConf.set("io.file.buffer.size", bufferSize)
-    hadoopConf
-  }
+  val hadoopConfiguration = SparkHadoopUtil.get.newConfiguration(conf)
 
   // Optionally log Spark events
   private[spark] val eventLogger: Option[EventLoggingListener] = {
@@ -815,7 +796,7 @@ class SparkContext(config: SparkConf) extends Logging {
    * Add a file to be downloaded with this Spark job on every node.
    * The `path` passed can be either a local file, a file in HDFS (or other Hadoop-supported
    * filesystems), or an HTTP, HTTPS or FTP URI.  To access the file in Spark jobs,
-   * use `SparkFiles.get(path)` to find its download location.
+   * use `SparkFiles.get(fileName)` to find its download location.
    */
   def addFile(path: String) {
     val uri = new URI(path)
@@ -827,7 +808,8 @@ class SparkContext(config: SparkConf) extends Logging {
     addedFiles(key) = System.currentTimeMillis
 
     // Fetch the file locally in case a job is executed using DAGScheduler.runLocally().
-    Utils.fetchFile(path, new File(SparkFiles.getRootDirectory()), conf, env.securityManager)
+    Utils.fetchFile(path, new File(SparkFiles.getRootDirectory()), conf, env.securityManager,
+      hadoopConfiguration)
 
     logInfo("Added file " + path + " at " + key + " with timestamp " + addedFiles(key))
     postEnvironmentUpdate()
@@ -1637,4 +1619,3 @@ private[spark] class WritableConverter[T](
     val writableClass: ClassTag[T] => Class[_ <: Writable],
     val convert: Writable => T)
   extends Serializable
-

core/src/main/scala/org/apache/spark/api/java/JavaRDDLike.scala

@@ -26,7 +26,7 @@ import scala.reflect.ClassTag
 import com.google.common.base.Optional
 import org.apache.hadoop.io.compress.CompressionCodec
 
-import org.apache.spark.{Partition, SparkContext, TaskContext}
+import org.apache.spark.{FutureAction, Partition, SparkContext, TaskContext}
 import org.apache.spark.annotation.Experimental
 import org.apache.spark.api.java.JavaPairRDD._
 import org.apache.spark.api.java.JavaSparkContext.fakeClassTag
@@ -574,4 +574,17 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable {
 
   def name(): String = rdd.name
 
+  /**
+   * :: Experimental ::
+   * The asynchronous version of the foreach action.
+   *
+   * @param f the function to apply to all the elements of the RDD
+   * @return a FutureAction for the action
+   */
+  @Experimental
+  def foreachAsync(f: VoidFunction[T]): FutureAction[Unit] = {
+    import org.apache.spark.SparkContext._
+    rdd.foreachAsync(x => f.call(x))
+  }
+
 }

core/src/main/scala/org/apache/spark/api/java/JavaSparkContext.scala

@@ -545,7 +545,7 @@ class JavaSparkContext(val sc: SparkContext) extends JavaSparkContextVarargsWork
    * Add a file to be downloaded with this Spark job on every node.
    * The `path` passed can be either a local file, a file in HDFS (or other Hadoop-supported
    * filesystems), or an HTTP, HTTPS or FTP URI.  To access the file in Spark jobs,
-   * use `SparkFiles.get(path)` to find its download location.
+   * use `SparkFiles.get(fileName)` to find its download location.
    */
   def addFile(path: String) {
     sc.addFile(path)

core/src/main/scala/org/apache/spark/broadcast/TorrentBroadcast.scala

@@ -28,6 +28,7 @@ import org.apache.spark.{Logging, SparkConf, SparkEnv, SparkException}
 import org.apache.spark.io.CompressionCodec
 import org.apache.spark.storage.{BroadcastBlockId, StorageLevel}
 import org.apache.spark.util.ByteBufferInputStream
+import org.apache.spark.util.io.ByteArrayChunkOutputStream
 
 /**
  * A BitTorrent-like implementation of [[org.apache.spark.broadcast.Broadcast]].
@@ -201,29 +202,12 @@ private object TorrentBroadcast extends Logging {
   }
 
   def blockifyObject[T: ClassTag](obj: T): Array[ByteBuffer] = {
-    // TODO: Create a special ByteArrayOutputStream that splits the output directly into chunks
-    // so we don't need to do the extra memory copy.
-    val bos = new ByteArrayOutputStream()
+    val bos = new ByteArrayChunkOutputStream(BLOCK_SIZE)
     val out: OutputStream = if (compress) compressionCodec.compressedOutputStream(bos) else bos
     val ser = SparkEnv.get.serializer.newInstance()
     val serOut = ser.serializeStream(out)
     serOut.writeObject[T](obj).close()
-    val byteArray = bos.toByteArray
-    val bais = new ByteArrayInputStream(byteArray)
-    val numBlocks = math.ceil(byteArray.length.toDouble / BLOCK_SIZE).toInt
-    val blocks = new Array[ByteBuffer](numBlocks)
-
-    var blockId = 0
-    for (i <- 0 until (byteArray.length, BLOCK_SIZE)) {
-      val thisBlockSize = math.min(BLOCK_SIZE, byteArray.length - i)
-      val tempByteArray = new Array[Byte](thisBlockSize)
-      bais.read(tempByteArray, 0, thisBlockSize)
-
-      blocks(blockId) = ByteBuffer.wrap(tempByteArray)
-      blockId += 1
-    }
-    bais.close()
-    blocks
+    bos.toArrays.map(ByteBuffer.wrap)
   }
 
   def unBlockifyObject[T: ClassTag](blocks: Array[ByteBuffer]): T = {

core/src/main/scala/org/apache/spark/deploy/SparkHadoopUtil.scala

@@ -24,15 +24,18 @@ import org.apache.hadoop.mapred.JobConf
 import org.apache.hadoop.security.Credentials
 import org.apache.hadoop.security.UserGroupInformation
 
-import org.apache.spark.{Logging, SparkContext, SparkException}
+import org.apache.spark.{Logging, SparkContext, SparkConf, SparkException}
+import org.apache.spark.annotation.DeveloperApi
 
 import scala.collection.JavaConversions._
 
 /**
+ * :: DeveloperApi ::
  * Contains util methods to interact with Hadoop from Spark.
  */
+@DeveloperApi
 class SparkHadoopUtil extends Logging {
-  val conf: Configuration = newConfiguration()
+  val conf: Configuration = newConfiguration(new SparkConf())
   UserGroupInformation.setConfiguration(conf)
 
   /**
@@ -64,11 +67,39 @@ class SparkHadoopUtil extends Logging {
     }
   }
 
+  @Deprecated
+  def newConfiguration(): Configuration = newConfiguration(null)
+
   /**
    * Return an appropriate (subclass) of Configuration. Creating config can initializes some Hadoop
    * subsystems.
    */
-  def newConfiguration(): Configuration = new Configuration()
+  def newConfiguration(conf: SparkConf): Configuration = {
+    val hadoopConf = new Configuration()
+
+    // Note: this null check is around more than just access to the "conf" object to maintain
+    // the behavior of the old implementation of this code, for backwards compatibility.
+    if (conf != null) {
+      // Explicitly check for S3 environment variables
+      if (System.getenv("AWS_ACCESS_KEY_ID") != null &&
+          System.getenv("AWS_SECRET_ACCESS_KEY") != null) {
+        hadoopConf.set("fs.s3.awsAccessKeyId", System.getenv("AWS_ACCESS_KEY_ID"))
+        hadoopConf.set("fs.s3n.awsAccessKeyId", System.getenv("AWS_ACCESS_KEY_ID"))
+        hadoopConf.set("fs.s3.awsSecretAccessKey", System.getenv("AWS_SECRET_ACCESS_KEY"))
+        hadoopConf.set("fs.s3n.awsSecretAccessKey", System.getenv("AWS_SECRET_ACCESS_KEY"))
+      }
+      // Copy any "spark.hadoop.foo=bar" system properties into conf as "foo=bar"
+      conf.getAll.foreach { case (key, value) =>
+        if (key.startsWith("spark.hadoop.")) {
+          hadoopConf.set(key.substring("spark.hadoop.".length), value)
+        }
+      }
+      val bufferSize = conf.get("spark.buffer.size", "65536")
+      hadoopConf.set("io.file.buffer.size", bufferSize)
+    }
+
+    hadoopConf
+  }
 
   /**
    * Add any user credentials to the job conf which are necessary for running on a secure Hadoop
@@ -86,7 +117,7 @@ class SparkHadoopUtil extends Logging {
 
   def getSecretKeyFromUserCredentials(key: String): Array[Byte] = { null }
 
-  def loginUserFromKeytab(principalName: String, keytabFilename: String) { 
+  def loginUserFromKeytab(principalName: String, keytabFilename: String) {
     UserGroupInformation.loginUserFromKeytab(principalName, keytabFilename)
   }
 

core/src/main/scala/org/apache/spark/deploy/history/FsHistoryProvider.scala

@@ -24,6 +24,7 @@ import scala.collection.mutable
 import org.apache.hadoop.fs.{FileStatus, Path}
 
 import org.apache.spark.{Logging, SecurityManager, SparkConf}
+import org.apache.spark.deploy.SparkHadoopUtil
 import org.apache.spark.scheduler._
 import org.apache.spark.ui.SparkUI
 import org.apache.spark.util.Utils
@@ -40,7 +41,8 @@ private[history] class FsHistoryProvider(conf: SparkConf) extends ApplicationHis
     .map { d => Utils.resolveURI(d) }
     .getOrElse { throw new IllegalArgumentException("Logging directory must be specified.") }
 
-  private val fs = Utils.getHadoopFileSystem(resolvedLogDir)
+  private val fs = Utils.getHadoopFileSystem(resolvedLogDir,
+    SparkHadoopUtil.get.newConfiguration(conf))
 
   // A timestamp of when the disk was last accessed to check for log updates
   private var lastLogCheckTimeMs = -1L

core/src/main/scala/org/apache/spark/deploy/master/Master.scala

@@ -33,7 +33,8 @@ import akka.remote.{DisassociatedEvent, RemotingLifecycleEvent}
 import akka.serialization.SerializationExtension
 
 import org.apache.spark.{Logging, SecurityManager, SparkConf, SparkException}
-import org.apache.spark.deploy.{ApplicationDescription, DriverDescription, ExecutorState}
+import org.apache.spark.deploy.{ApplicationDescription, DriverDescription, ExecutorState,
+  SparkHadoopUtil}
 import org.apache.spark.deploy.DeployMessages._
 import org.apache.spark.deploy.history.HistoryServer
 import org.apache.spark.deploy.master.DriverState.DriverState
@@ -673,7 +674,8 @@ private[spark] class Master(
       app.desc.appUiUrl = notFoundBasePath
       return false
     }
-    val fileSystem = Utils.getHadoopFileSystem(eventLogDir)
+    val fileSystem = Utils.getHadoopFileSystem(eventLogDir,
+      SparkHadoopUtil.get.newConfiguration(conf))
     val eventLogInfo = EventLoggingListener.parseLoggingInfo(eventLogDir, fileSystem)
     val eventLogPaths = eventLogInfo.logPaths
     val compressionCodec = eventLogInfo.compressionCodec