apache
diff --git a/‎core/src/main/scala/org/apache/spark/SparkContext.scala‎
Lines changed: 36 additions & 16 deletions b/‎core/src/main/scala/org/apache/spark/SparkContext.scala‎
Lines changed: 36 additions & 16 deletions
diff --git a/‎core/src/main/scala/org/apache/spark/SparkEnv.scala‎
Lines changed: 6 additions & 2 deletions b/‎core/src/main/scala/org/apache/spark/SparkEnv.scala‎
Lines changed: 6 additions & 2 deletions
diff --git a/‎core/src/main/scala/org/apache/spark/deploy/master/Master.scala‎
Lines changed: 7 additions & 5 deletions b/‎core/src/main/scala/org/apache/spark/deploy/master/Master.scala‎
Lines changed: 7 additions & 5 deletions
diff --git a/‎core/src/main/scala/org/apache/spark/executor/CoarseGrainedExecutorBackend.scala‎
Lines changed: 9 additions & 7 deletions b/‎core/src/main/scala/org/apache/spark/executor/CoarseGrainedExecutorBackend.scala‎
Lines changed: 9 additions & 7 deletions
diff --git a/‎core/src/main/scala/org/apache/spark/executor/Executor.scala‎
Lines changed: 1 addition & 0 deletions b/‎core/src/main/scala/org/apache/spark/executor/Executor.scala‎
Lines changed: 1 addition & 0 deletions
diff --git a/‎core/src/main/scala/org/apache/spark/executor/ExecutorSource.scala‎
Lines changed: 1 addition & 2 deletions b/‎core/src/main/scala/org/apache/spark/executor/ExecutorSource.scala‎
Lines changed: 1 addition & 2 deletions
diff --git a/‎core/src/main/scala/org/apache/spark/executor/MesosExecutorBackend.scala‎
Lines changed: 2 additions & 1 deletion b/‎core/src/main/scala/org/apache/spark/executor/MesosExecutorBackend.scala‎
Lines changed: 2 additions & 1 deletion
diff --git a/‎core/src/main/scala/org/apache/spark/metrics/MetricsSystem.scala‎
Lines changed: 35 additions & 5 deletions b/‎core/src/main/scala/org/apache/spark/metrics/MetricsSystem.scala‎
Lines changed: 35 additions & 5 deletions
diff --git a/‎core/src/main/scala/org/apache/spark/scheduler/DAGSchedulerSource.scala‎
Lines changed: 2 additions & 2 deletions b/‎core/src/main/scala/org/apache/spark/scheduler/DAGSchedulerSource.scala‎
Lines changed: 2 additions & 2 deletions
diff --git a/‎core/src/main/scala/org/apache/spark/scheduler/EventLoggingListener.scala‎
Lines changed: 18 additions & 15 deletions b/‎core/src/main/scala/org/apache/spark/scheduler/EventLoggingListener.scala‎
Lines changed: 18 additions & 15 deletions
@@ -187,6 +187,15 @@ class SparkContext(config: SparkConf) extends Logging {
   val master = conf.get("spark.master")
   val appName = conf.get("spark.app.name")
 
+  private[spark] val isEventLogEnabled = conf.getBoolean("spark.eventLog.enabled", false)
+  private[spark] val eventLogDir: Option[String] = {
+    if (isEventLogEnabled) {
+      Some(conf.get("spark.eventLog.dir", EventLoggingListener.DEFAULT_LOG_DIR).stripSuffix("/"))
+    } else {
+      None
+    }
+  }
+
   // Generate the random name for a temp folder in Tachyon
   // Add a timestamp as the suffix here to make it more safe
   val tachyonFolderName = "spark-" + randomUUID.toString()
@@ -200,6 +209,7 @@ class SparkContext(config: SparkConf) extends Logging {
   private[spark] val listenerBus = new LiveListenerBus
 
   // Create the Spark execution environment (cache, map output tracker, etc)
+  conf.set("spark.executor.id", "driver")
   private[spark] val env = SparkEnv.create(
     conf,
     "<driver>",
@@ -232,19 +242,6 @@ class SparkContext(config: SparkConf) extends Logging {
   /** A default Hadoop Configuration for the Hadoop code (e.g. file systems) that we reuse. */
   val hadoopConfiguration = SparkHadoopUtil.get.newConfiguration(conf)
 
-  // Optionally log Spark events
-  private[spark] val eventLogger: Option[EventLoggingListener] = {
-    if (conf.getBoolean("spark.eventLog.enabled", false)) {
-      val logger = new EventLoggingListener(appName, conf, hadoopConfiguration)
-      logger.start()
-      listenerBus.addListener(logger)
-      Some(logger)
-    } else None
-  }
-
-  // At this point, all relevant SparkListeners have been registered, so begin releasing events
-  listenerBus.start()
-
   val startTime = System.currentTimeMillis()
 
   // Add each JAR given through the constructor
@@ -309,6 +306,29 @@ class SparkContext(config: SparkConf) extends Logging {
   // constructor
   taskScheduler.start()
 
+  val applicationId: String = taskScheduler.applicationId()
+  conf.set("spark.app.id", applicationId)
+
+  val metricsSystem = env.metricsSystem
+
+  // The metrics system for Driver need to be set spark.app.id to app ID.
+  // So it should start after we get app ID from the task scheduler and set spark.app.id.
+  metricsSystem.start()
+
+  // Optionally log Spark events
+  private[spark] val eventLogger: Option[EventLoggingListener] = {
+    if (isEventLogEnabled) {
+      val logger =
+        new EventLoggingListener(applicationId, eventLogDir.get, conf, hadoopConfiguration)
+      logger.start()
+      listenerBus.addListener(logger)
+      Some(logger)
+    } else None
+  }
+
+  // At this point, all relevant SparkListeners have been registered, so begin releasing events
+  listenerBus.start()
+
   private[spark] val cleaner: Option[ContextCleaner] = {
     if (conf.getBoolean("spark.cleaner.referenceTracking", true)) {
       Some(new ContextCleaner(this))
@@ -411,8 +431,8 @@ class SparkContext(config: SparkConf) extends Logging {
   // Post init
   taskScheduler.postStartHook()
 
-  private val dagSchedulerSource = new DAGSchedulerSource(this.dagScheduler, this)
-  private val blockManagerSource = new BlockManagerSource(SparkEnv.get.blockManager, this)
+  private val dagSchedulerSource = new DAGSchedulerSource(this.dagScheduler)
+  private val blockManagerSource = new BlockManagerSource(SparkEnv.get.blockManager)
 
   private def initDriverMetrics() {
     SparkEnv.get.metricsSystem.registerSource(dagSchedulerSource)
@@ -1278,7 +1298,7 @@ class SparkContext(config: SparkConf) extends Logging {
   private def postApplicationStart() {
     // Note: this code assumes that the task scheduler has been initialized and has contacted
     // the cluster manager to get an application ID (in case the cluster manager provides one).
-    listenerBus.post(SparkListenerApplicationStart(appName, taskScheduler.applicationId(),
+    listenerBus.post(SparkListenerApplicationStart(appName, Some(applicationId),
       startTime, sparkUser))
   }
 
 
@@ -259,11 +259,15 @@ object SparkEnv extends Logging {
       }
 
     val metricsSystem = if (isDriver) {
+      // Don't start metrics system right now for Driver.
+      // We need to wait for the task scheduler to give us an app ID.
+      // Then we can start the metrics system.
       MetricsSystem.createMetricsSystem("driver", conf, securityManager)
     } else {
-      MetricsSystem.createMetricsSystem("executor", conf, securityManager)
+      val ms = MetricsSystem.createMetricsSystem("executor", conf, securityManager)
+      ms.start()
+      ms
     }
-    metricsSystem.start()
 
     // Set the sparkFiles directory, used when downloading dependencies.  In local mode,
     // this is a temporary directory; in distributed mode, this is the executor's current working
 
@@ -33,8 +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,
-  SparkHadoopUtil}
+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
@@ -693,16 +693,18 @@ private[spark] class Master(
       app.desc.appUiUrl = notFoundBasePath
       return false
     }
-    val fileSystem = Utils.getHadoopFileSystem(eventLogDir,
+
+    val appEventLogDir = EventLoggingListener.getLogDirPath(eventLogDir, app.id)
+    val fileSystem = Utils.getHadoopFileSystem(appEventLogDir,
       SparkHadoopUtil.get.newConfiguration(conf))
-    val eventLogInfo = EventLoggingListener.parseLoggingInfo(eventLogDir, fileSystem)
+    val eventLogInfo = EventLoggingListener.parseLoggingInfo(appEventLogDir, fileSystem)
     val eventLogPaths = eventLogInfo.logPaths
     val compressionCodec = eventLogInfo.compressionCodec
 
     if (eventLogPaths.isEmpty) {
       // Event logging is enabled for this application, but no event logs are found
       val title = s"Application history not found (${app.id})"
-      var msg = s"No event logs found for application $appName in $eventLogDir."
+      var msg = s"No event logs found for application $appName in $appEventLogDir."
       logWarning(msg)
       msg += " Did you specify the correct logging directory?"
       msg = URLEncoder.encode(msg, "UTF-8")
 
@@ -106,6 +106,7 @@ private[spark] object CoarseGrainedExecutorBackend extends Logging {
       executorId: String,
       hostname: String,
       cores: Int,
+      appId: String,
       workerUrl: Option[String]) {
 
     SignalLogger.register(log)
@@ -122,7 +123,8 @@ private[spark] object CoarseGrainedExecutorBackend extends Logging {
       val driver = fetcher.actorSelection(driverUrl)
       val timeout = AkkaUtils.askTimeout(executorConf)
       val fut = Patterns.ask(driver, RetrieveSparkProps, timeout)
-      val props = Await.result(fut, timeout).asInstanceOf[Seq[(String, String)]]
+      val props = Await.result(fut, timeout).asInstanceOf[Seq[(String, String)]] ++
+        Seq[(String, String)](("spark.app.id", appId))
       fetcher.shutdown()
 
       // Create a new ActorSystem using driver's Spark properties to run the backend.
@@ -144,16 +146,16 @@ private[spark] object CoarseGrainedExecutorBackend extends Logging {
 
   def main(args: Array[String]) {
     args.length match {
-      case x if x < 4 =>
+      case x if x < 5 =>
         System.err.println(
           // Worker url is used in spark standalone mode to enforce fate-sharing with worker
           "Usage: CoarseGrainedExecutorBackend <driverUrl> <executorId> <hostname> " +
-          "<cores> [<workerUrl>]")
+          "<cores> <appid> [<workerUrl>] ")
         System.exit(1)
-      case 4 =>
-        run(args(0), args(1), args(2), args(3).toInt, None)
-      case x if x > 4 =>
-        run(args(0), args(1), args(2), args(3).toInt, Some(args(4)))
+      case 5 =>
+        run(args(0), args(1), args(2), args(3).toInt, args(4), None)
+      case x if x > 5 =>
+        run(args(0), args(1), args(2), args(3).toInt, args(4), Some(args(5)))
     }
   }
 }
@@ -74,6 +74,7 @@ private[spark] class Executor(
   val executorSource = new ExecutorSource(this, executorId)
 
   // Initialize Spark environment (using system properties read above)
+  conf.set("spark.executor.id", "executor." + executorId)
   private val env = {
     if (!isLocal) {
       val _env = SparkEnv.create(conf, executorId, slaveHostname, 0,
 
@@ -37,8 +37,7 @@ private[spark] class ExecutorSource(val executor: Executor, executorId: String)
 
   override val metricRegistry = new MetricRegistry()
 
-  // TODO: It would be nice to pass the application name here
-  override val sourceName = "executor.%s".format(executorId)
+  override val sourceName = "executor"
 
   // Gauge for executor thread pool's actively executing task counts
   metricRegistry.register(MetricRegistry.name("threadpool", "activeTasks"), new Gauge[Int] {
 
@@ -52,7 +52,8 @@ private[spark] class MesosExecutorBackend
       slaveInfo: SlaveInfo) {
     logInfo("Registered with Mesos as executor ID " + executorInfo.getExecutorId.getValue)
     this.driver = driver
-    val properties = Utils.deserialize[Array[(String, String)]](executorInfo.getData.toByteArray)
+    val properties = Utils.deserialize[Array[(String, String)]](executorInfo.getData.toByteArray) ++
+      Seq[(String, String)](("spark.app.id", frameworkInfo.getId.getValue))
     executor = new Executor(
       executorInfo.getExecutorId.getValue,
       slaveInfo.getHostname,
 
@@ -83,10 +83,10 @@ private[spark] class MetricsSystem private (
   def getServletHandlers = metricsServlet.map(_.getHandlers).getOrElse(Array())
 
   metricsConfig.initialize()
-  registerSources()
-  registerSinks()
 
   def start() {
+    registerSources()
+    registerSinks()
     sinks.foreach(_.start)
   }
 
@@ -98,19 +98,49 @@ private[spark] class MetricsSystem private (
     sinks.foreach(_.report())
   }
 
+  /**
+   * Build a name that uniquely identifies each metric source.
+   * The name is structured as follows: <app ID>.<executor ID (or "driver")>.<source name>.
+   * If either ID is not available, this defaults to just using <source name>.
+   *
+   * @param source Metric source to be named by this method.
+   * @return An unique metric name for each combination of
+   *         application, executor/driver and metric source.
+   */
+  def buildRegistryName(source: Source): String = {
+    val appId = conf.getOption("spark.app.id")
+    val executorId = conf.getOption("spark.executor.id")
+    val defaultName = MetricRegistry.name(source.sourceName)
+
+    if (instance == "driver" || instance == "executor") {
+      if (appId.isDefined && executorId.isDefined) {
+        MetricRegistry.name(appId.get, executorId.get, source.sourceName)
+      } else {
+        // Only Driver and Executor are set spark.app.id and spark.executor.id.
+        // For instance, Master and Worker are not related to a specific application.
+        val warningMsg = s"Using default name $defaultName for source because %s is not set."
+        if (appId.isEmpty) { logWarning(warningMsg.format("spark.app.id")) }
+        if (executorId.isEmpty) { logWarning(warningMsg.format("spark.executor.id")) }
+        defaultName
+      }
+    } else { defaultName }
+  }
+
   def registerSource(source: Source) {
     sources += source
     try {
-      registry.register(source.sourceName, source.metricRegistry)
+      val regName = buildRegistryName(source)
+      registry.register(regName, source.metricRegistry)
     } catch {
       case e: IllegalArgumentException => logInfo("Metrics already registered", e)
     }
   }
 
   def removeSource(source: Source) {
     sources -= source
+    val regName = buildRegistryName(source)
     registry.removeMatching(new MetricFilter {
-      def matches(name: String, metric: Metric): Boolean = name.startsWith(source.sourceName)
+      def matches(name: String, metric: Metric): Boolean = name.startsWith(regName)
     })
   }
 
@@ -125,7 +155,7 @@ private[spark] class MetricsSystem private (
         val source = Class.forName(classPath).newInstance()
         registerSource(source.asInstanceOf[Source])
       } catch {
-        case e: Exception => logError("Source class " + classPath + " cannot be instantialized", e)
+        case e: Exception => logError("Source class " + classPath + " cannot be instantiated", e)
       }
     }
   }
 
@@ -22,10 +22,10 @@ import com.codahale.metrics.{Gauge,MetricRegistry}
 import org.apache.spark.SparkContext
 import org.apache.spark.metrics.source.Source
 
-private[spark] class DAGSchedulerSource(val dagScheduler: DAGScheduler, sc: SparkContext)
+private[spark] class DAGSchedulerSource(val dagScheduler: DAGScheduler)
     extends Source {
   override val metricRegistry = new MetricRegistry()
-  override val sourceName = "%s.DAGScheduler".format(sc.appName)
+  override val sourceName = "DAGScheduler"
 
   metricRegistry.register(MetricRegistry.name("stage", "failedStages"), new Gauge[Int] {
     override def getValue: Int = dagScheduler.failedStages.size
 
@@ -43,38 +43,29 @@ import org.apache.spark.util.{FileLogger, JsonProtocol, Utils}
  *   spark.eventLog.buffer.kb - Buffer size to use when writing to output streams
  */
 private[spark] class EventLoggingListener(
-    appName: String,
+    appId: String,
+    logBaseDir: String,
     sparkConf: SparkConf,
     hadoopConf: Configuration)
   extends SparkListener with Logging {
 
   import EventLoggingListener._
 
-  def this(appName: String, sparkConf: SparkConf) =
-    this(appName, sparkConf, SparkHadoopUtil.get.newConfiguration(sparkConf))
+  def this(appId: String, logBaseDir: String, sparkConf: SparkConf) =
+    this(appId, logBaseDir, sparkConf, SparkHadoopUtil.get.newConfiguration(sparkConf))
 
   private val shouldCompress = sparkConf.getBoolean("spark.eventLog.compress", false)
   private val shouldOverwrite = sparkConf.getBoolean("spark.eventLog.overwrite", false)
   private val testing = sparkConf.getBoolean("spark.eventLog.testing", false)
   private val outputBufferSize = sparkConf.getInt("spark.eventLog.buffer.kb", 100) * 1024
-  private val logBaseDir = sparkConf.get("spark.eventLog.dir", DEFAULT_LOG_DIR).stripSuffix("/")
-  private val name = appName.replaceAll("[ :/]", "-").replaceAll("[${}'\"]", "_")
-    .toLowerCase + "-" + System.currentTimeMillis
-  val logDir = Utils.resolveURI(logBaseDir) + "/" + name.stripSuffix("/")
-
+  val logDir = EventLoggingListener.getLogDirPath(logBaseDir, appId)
+  val logDirName: String = logDir.split("/").last
   protected val logger = new FileLogger(logDir, sparkConf, hadoopConf, outputBufferSize,
     shouldCompress, shouldOverwrite, Some(LOG_FILE_PERMISSIONS))
 
   // For testing. Keep track of all JSON serialized events that have been logged.
   private[scheduler] val loggedEvents = new ArrayBuffer[JValue]
 
-  /**
-   * Return only the unique application directory without the base directory.
-   */
-  def getApplicationLogDir(): String = {
-    name
-  }
-
   /**
    * Begin logging events.
    * If compression is used, log a file that indicates which compression library is used.
@@ -184,6 +175,18 @@ private[spark] object EventLoggingListener extends Logging {
     } else ""
   }
 
+  /**
+   * Return a file-system-safe path to the log directory for the given application.
+   *
+   * @param logBaseDir A base directory for the path to the log directory for given application.
+   * @param appId A unique app ID.
+   * @return A path which consists of file-system-safe characters.
+   */
+  def getLogDirPath(logBaseDir: String, appId: String): String = {
+    val name = appId.replaceAll("[ :/]", "-").replaceAll("[${}'\"]", "_").toLowerCase
+    Utils.resolveURI(logBaseDir) + "/" + name.stripSuffix("/")
+  }
+
   /**
    * Parse the event logging information associated with the logs in the given directory.
    *