From 5c24759ddc25cbafbedbaafbf053d38015a7774e Mon Sep 17 00:00:00 2001 From: wangfei Date: Mon, 15 Dec 2014 16:46:21 -0800 Subject: [Minor][Core] fix comments in MapOutputTracker Using driver and executor in the comments of ```MapOutputTracker``` is more clear. Author: wangfei Closes #3700 from scwf/commentFix and squashes the following commits: aa68524 [wangfei] master and worker should be driver and executor --- .../scala/org/apache/spark/MapOutputTracker.scala | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) (limited to 'core') diff --git a/core/src/main/scala/org/apache/spark/MapOutputTracker.scala b/core/src/main/scala/org/apache/spark/MapOutputTracker.scala index e45885338e..a074ab8ece 100644 --- a/core/src/main/scala/org/apache/spark/MapOutputTracker.scala +++ b/core/src/main/scala/org/apache/spark/MapOutputTracker.scala @@ -72,7 +72,7 @@ private[spark] class MapOutputTrackerMasterActor(tracker: MapOutputTrackerMaster /** * Class that keeps track of the location of the map output of * a stage. This is abstract because different versions of MapOutputTracker - * (driver and worker) use different HashMap to store its metadata. + * (driver and executor) use different HashMap to store its metadata. */ private[spark] abstract class MapOutputTracker(conf: SparkConf) extends Logging { private val timeout = AkkaUtils.askTimeout(conf) @@ -81,11 +81,11 @@ private[spark] abstract class MapOutputTracker(conf: SparkConf) extends Logging var trackerActor: ActorRef = _ /** - * This HashMap has different behavior for the master and the workers. + * This HashMap has different behavior for the driver and the executors. * - * On the master, it serves as the source of map outputs recorded from ShuffleMapTasks. - * On the workers, it simply serves as a cache, in which a miss triggers a fetch from the - * master's corresponding HashMap. + * On the driver, it serves as the source of map outputs recorded from ShuffleMapTasks. + * On the executors, it simply serves as a cache, in which a miss triggers a fetch from the + * driver's corresponding HashMap. * * Note: because mapStatuses is accessed concurrently, subclasses should make sure it's a * thread-safe map. @@ -99,7 +99,7 @@ private[spark] abstract class MapOutputTracker(conf: SparkConf) extends Logging protected var epoch: Long = 0 protected val epochLock = new AnyRef - /** Remembers which map output locations are currently being fetched on a worker. */ + /** Remembers which map output locations are currently being fetched on an executor. */ private val fetching = new HashSet[Int] /** @@ -196,8 +196,8 @@ private[spark] abstract class MapOutputTracker(conf: SparkConf) extends Logging /** * Called from executors to update the epoch number, potentially clearing old outputs - * because of a fetch failure. Each worker task calls this with the latest epoch - * number on the master at the time it was created. + * because of a fetch failure. Each executor task calls this with the latest epoch + * number on the driver at the time it was created. */ def updateEpoch(newEpoch: Long) { epochLock.synchronized { @@ -229,7 +229,7 @@ private[spark] class MapOutputTrackerMaster(conf: SparkConf) private var cacheEpoch = epoch /** - * Timestamp based HashMap for storing mapStatuses and cached serialized statuses in the master, + * Timestamp based HashMap for storing mapStatuses and cached serialized statuses in the driver, * so that statuses are dropped only by explicit de-registering or by TTL-based cleaning (if set). * Other than these two scenarios, nothing should be dropped from this HashMap. */ @@ -339,7 +339,7 @@ private[spark] class MapOutputTrackerMaster(conf: SparkConf) } /** - * MapOutputTracker for the workers, which fetches map output information from the driver's + * MapOutputTracker for the executors, which fetches map output information from the driver's * MapOutputTrackerMaster. */ private[spark] class MapOutputTrackerWorker(conf: SparkConf) extends MapOutputTracker(conf) { -- cgit v1.2.3