Spark On YARN内存分配

「爱情、让人受尽委屈。」 2022-07-05 15:09 379阅读 0赞

# Spark On YARN内存分配 #

时间 2015-06-09 00:00:00   [JavaChen's Blog][JavaChen_s Blog]

原文   [http://blog.javachen.com/2015/06/09/memory-in-spark-on-yarn.html][http_blog.javachen.com_2015_06_09_memory-in-spark-on-yarn.html]

主题  [Spark ][Spark] [YARN][]

本文主要了解Spark On YARN部署模式下的内存分配情况，因为没有深入研究Spark的源代码，所以只能根据日志去看相关的源代码，从而了解“为什么会这样，为什么会那样”。

## 说明 ##

按照Spark应用程序中的driver分布方式不同，Spark on YARN有两种模式： `yarn-client` 模式、 `yarn-cluster` 模式。

当在YARN上运行Spark作业，每个Spark executor作为一个YARN容器运行。Spark可以使得多个Tasks在同一个容器里面运行。

下图是yarn-cluster模式的作业执行图，图片来源于网络：

![taobao.png][]

关于Spark On YARN相关的配置参数，请参考Spark配置参数。本文主要讨论内存分配情况，所以只需要关注以下几个内心相关的参数：

*  `spark.driver.memory` ：默认值512m
 *  `spark.executor.memory` ：默认值512m
 *  `spark.yarn.am.memory` ：默认值512m
 *  `spark.yarn.executor.memoryOverhead` ：值为 `executorMemory * 0.07, with minimum of 384`
 *  `spark.yarn.driver.memoryOverhead` ：值为 `driverMemory * 0.07, with minimum of 384`
 *  `spark.yarn.am.memoryOverhead` ：值为 `AM memory * 0.07, with minimum of 384`

注意：

*  `--executor-memory/spark.executor.memory` 控制 executor 的堆的大小，但是 JVM 本身也会占用一定的堆空间，比如内部的 String 或者直接 byte buffer， `spark.yarn.XXX.memoryOverhead` 属性决定向 YARN 请求的每个 executor 或dirver或am 的额外堆内存大小，默认值为 `max(384, 0.07 * spark.executor.memory` )
 *  在 executor 执行的时候配置过大的 memory 经常会导致过长的GC延时，64G是推荐的一个 executor 内存大小的上限。
 *  HDFS client 在大量并发线程时存在性能问题。大概的估计是每个 executor 中最多5个并行的 task 就可以占满写入带宽。

另外，因为任务是提交到YARN上运行的，所以YARN中有几个关键参数，参考YARN的内存和CPU配置：

*  `yarn.app.mapreduce.am.resource.mb` ：AM能够申请的最大内存，默认值为1536MB
 *  `yarn.nodemanager.resource.memory-mb` ：nodemanager能够申请的最大内存，默认值为8192MB
 *  `yarn.scheduler.minimum-allocation-mb` ：调度时一个container能够申请的最小资源，默认值为1024MB
 *  `yarn.scheduler.maximum-allocation-mb` ：调度时一个container能够申请的最大资源，默认值为8192MB

## 测试 ##

Spark集群测试环境为：

*  master：64G内存，16核cpu
 *  worker：128G内存，32核cpu
 *  worker：128G内存，32核cpu
 *  worker：128G内存，32核cpu
 *  worker：128G内存，32核cpu

注意：YARN集群部署在Spark集群之上的，每一个worker节点上同时部署了一个NodeManager，并且YARN集群中的配置如下：

<property>
    	  <name>yarn.nodemanager.resource.memory-mb</name>
    	  <value>106496</value> 
      </property>
      <property>
    	  <name>yarn.scheduler.minimum-allocation-mb</name>
    	  <value>2048</value>
      </property>
      <property>
    	  <name>yarn.scheduler.maximum-allocation-mb</name>
    	  <value>106496</value>
      </property>
      <property>
    	  <name>yarn.app.mapreduce.am.resource.mb</name>
    	  <value>2048</value>
      </property>

将spark的日志基本调为DEBUG，并将log4j.logger.org.apache.hadoop设置为WARN建设不必要的输出，修改/etc/spark/conf/log4j.properties：

# Set everything to be logged to the console
    log4j.rootCategory=DEBUG, console
    log4j.appender.console=org.apache.log4j.ConsoleAppender
    log4j.appender.console.target=System.err
    log4j.appender.console.layout=org.apache.log4j.PatternLayout
    log4j.appender.console.layout.ConversionPattern=%d{yy/MM/dd HH:mm:ss} %p %c{1}: %m%n
    
    # Settings to quiet third party logs that are too verbose
    log4j.logger.org.eclipse.jetty=WARN
    log4j.logger.org.apache.hadoop=WARN
    log4j.logger.org.eclipse.jetty.util.component.AbstractLifeCycle=ERROR
    log4j.logger.org.apache.spark.repl.SparkIMain$exprTyper=INFO
    log4j.logger.org.apache.spark.repl.SparkILoop$SparkILoopInterpreter=INFO

接下来是运行测试程序，以官方自带的SparkPi例子为例， `下面主要测试client模式，至于cluster模式请参考下面的过程` 。运行下面命令：

spark-submit --class org.apache.spark.examples.SparkPi \
    	--master yarn-client  \
    	--num-executors 4 \
    	--driver-memory 2g \
    	--executor-memory 3g \
    	--executor-cores 4 \
    	/usr/lib/spark/lib/spark-examples-1.3.0-cdh5.4.0-hadoop2.6.0-cdh5.4.0.jar \
    	100000

观察输出日志（无关的日志被略去）：

15/06/08 13:57:01 INFO SparkContext: Running Spark version 1.3.0
    15/06/08 13:57:02 INFO SecurityManager: Changing view acls to: root
    15/06/08 13:57:02 INFO SecurityManager: Changing modify acls to: root
    
    15/06/08 13:57:03 INFO MemoryStore: MemoryStore started with capacity 1060.3 MB
    
    15/06/08 13:57:04 DEBUG YarnClientSchedulerBackend: ClientArguments called with: --arg bj03-bi-pro-hdpnamenn:51568 --num-executors 4 --num-executors 4 --executor-memory 3g --executor-memory 3g --executor-cores 4 --executor-cores 4 --name Spark Pi
    15/06/08 13:57:04 DEBUG YarnClientSchedulerBackend: [actor] handled message (24.52531 ms) ReviveOffers from Actor[akka://sparkDriver/user/CoarseGrainedScheduler#864850679]
    15/06/08 13:57:05 INFO Client: Requesting a new application from cluster with 4 NodeManagers
    15/06/08 13:57:05 INFO Client: Verifying our application has not requested more than the maximum memory capability of the cluster (106496 MB per container)
    15/06/08 13:57:05 INFO Client: Will allocate AM container, with 896 MB memory including 384 MB overhead
    15/06/08 13:57:05 INFO Client: Setting up container launch context for our AM
    
    15/06/08 13:57:07 DEBUG Client: ===============================================================================
    15/06/08 13:57:07 DEBUG Client: Yarn AM launch context:
    15/06/08 13:57:07 DEBUG Client:     user class: N/A
    15/06/08 13:57:07 DEBUG Client:     env:
    15/06/08 13:57:07 DEBUG Client:         CLASSPATH -> <CPS>/__spark__.jar<CPS>$HADOOP_CONF_DIR<CPS>$HADOOP_COMMON_HOME/*<CPS>$HADOOP_COMMON_HOME/lib/*<CPS>$HADOOP_HDFS_HOME/*<CPS>$HADOOP_HDFS_HOME/lib/*<CPS>$HADOOP_MAPRED_HOME/*<CPS>$HADOOP_MAPRED_HOME/lib/*<CPS>$HADOOP_YARN_HOME/*<CPS>$HADOOP_YARN_HOME/lib/*<CPS>$HADOOP_MAPRED_HOME/share/hadoop/mapreduce/*<CPS>$HADOOP_MAPRED_HOME/share/hadoop/mapreduce/lib/*<CPS>:/usr/lib/spark/lib/spark-assembly.jar::/usr/lib/hadoop/lib/*:/usr/lib/hadoop/*:/usr/lib/hadoop-hdfs/lib/*:/usr/lib/hadoop-hdfs/*:/usr/lib/hadoop-mapreduce/lib/*:/usr/lib/hadoop-mapreduce/*:/usr/lib/hadoop-yarn/lib/*:/usr/lib/hadoop-yarn/*:/usr/lib/hive/lib/*:/usr/lib/flume-ng/lib/*:/usr/lib/paquet/lib/*:/usr/lib/avro/lib/*
    15/06/08 13:57:07 DEBUG Client:         SPARK_DIST_CLASSPATH -> :/usr/lib/spark/lib/spark-assembly.jar::/usr/lib/hadoop/lib/*:/usr/lib/hadoop/*:/usr/lib/hadoop-hdfs/lib/*:/usr/lib/hadoop-hdfs/*:/usr/lib/hadoop-mapreduce/lib/*:/usr/lib/hadoop-mapreduce/*:/usr/lib/hadoop-yarn/lib/*:/usr/lib/hadoop-yarn/*:/usr/lib/hive/lib/*:/usr/lib/flume-ng/lib/*:/usr/lib/paquet/lib/*:/usr/lib/avro/lib/*
    15/06/08 13:57:07 DEBUG Client:         SPARK_YARN_CACHE_FILES_FILE_SIZES -> 97237208
    15/06/08 13:57:07 DEBUG Client:         SPARK_YARN_STAGING_DIR -> .sparkStaging/application_1433742899916_0001
    15/06/08 13:57:07 DEBUG Client:         SPARK_YARN_CACHE_FILES_VISIBILITIES -> PRIVATE
    15/06/08 13:57:07 DEBUG Client:         SPARK_USER -> root
    15/06/08 13:57:07 DEBUG Client:         SPARK_YARN_MODE -> true
    15/06/08 13:57:07 DEBUG Client:         SPARK_YARN_CACHE_FILES_TIME_STAMPS -> 1433743027399
    15/06/08 13:57:07 DEBUG Client:         SPARK_YARN_CACHE_FILES -> hdfs://mycluster:8020/user/root/.sparkStaging/application_1433742899916_0001/spark-assembly-1.3.0-cdh5.4.0-hadoop2.6.0-cdh5.4.0.jar#__spark__.jar
    15/06/08 13:57:07 DEBUG Client:     resources:
    15/06/08 13:57:07 DEBUG Client:         __spark__.jar -> resource { scheme: "hdfs" host: "mycluster" port: 8020 file: "/user/root/.sparkStaging/application_1433742899916_0001/spark-assembly-1.3.0-cdh5.4.0-hadoop2.6.0-cdh5.4.0.jar" } size: 97237208 timestamp: 1433743027399 type: FILE visibility: PRIVATE
    15/06/08 13:57:07 DEBUG Client:     command:
    15/06/08 13:57:07 DEBUG Client:         /bin/java -server -Xmx512m -Djava.io.tmpdir=/tmp '-Dspark.eventLog.enabled=true' '-Dspark.executor.instances=4' '-Dspark.executor.memory=3g' '-Dspark.executor.cores=4' '-Dspark.driver.port=51568' '-Dspark.serializer=org.apache.spark.serializer.KryoSerializer' '-Dspark.driver.appUIAddress=http://bj03-bi-pro-hdpnamenn:4040' '-Dspark.executor.id=<driver>' '-Dspark.kryo.classesToRegister=scala.collection.mutable.BitSet,scala.Tuple2,scala.Tuple1,org.apache.spark.mllib.recommendation.Rating' '-Dspark.driver.maxResultSize=8g' '-Dspark.jars=file:/usr/lib/spark/lib/spark-examples-1.3.0-cdh5.4.0-hadoop2.6.0-cdh5.4.0.jar' '-Dspark.driver.memory=2g' '-Dspark.eventLog.dir=hdfs://mycluster:8020/user/spark/applicationHistory' '-Dspark.app.name=Spark Pi' '-Dspark.fileserver.uri=http://X.X.X.X:49172' '-Dspark.tachyonStore.folderName=spark-81ae0186-8325-40f2-867b-65ee7c922357' -Dspark.yarn.app.container.log.dir=<LOG_DIR> org.apache.spark.deploy.yarn.ExecutorLauncher --arg 'bj03-bi-pro-hdpnamenn:51568' --executor-memory 3072m --executor-cores 4 --num-executors  4 1> <LOG_DIR>/stdout 2> <LOG_DIR>/stderr
    15/06/08 13:57:07 DEBUG Client: ===============================================================================

从 `Will allocate AM container, with 896 MB memory including 384 MB overhead` 日志可以看到，AM占用了 `896 MB` 内存，除掉 `384 MB` 的overhead内存，实际上只有 `512 MB` ，即 `spark.yarn.am.memory` 的默认值，另外可以看到YARN集群有4个NodeManager，每个container最多有106496 MB内存。

Yarn AM launch context启动了一个Java进程，设置的JVM内存为 `512m` ，见 `/bin/java -server -Xmx512m` 。

这里为什么会取默认值呢？查看打印上面这行日志的代码，见org.apache.spark.deploy.yarn.Client：

private def verifyClusterResources(newAppResponse: GetNewApplicationResponse): Unit = { 
        val maxMem = newAppResponse.getMaximumResourceCapability().getMemory()
        logInfo("Verifying our application has not requested more than the maximum " +
          s"memory capability of the cluster ($maxMem MB per container)")
        val executorMem = args.executorMemory + executorMemoryOverhead
        if (executorMem > maxMem) { 
          throw new IllegalArgumentException(s"Required executor memory (${args.executorMemory}" +
            s"+$executorMemoryOverhead MB) is above the max threshold ($maxMem MB) of this cluster!")
        }
        val amMem = args.amMemory + amMemoryOverhead
        if (amMem > maxMem) { 
          throw new IllegalArgumentException(s"Required AM memory (${args.amMemory}" +
            s"+$amMemoryOverhead MB) is above the max threshold ($maxMem MB) of this cluster!")
        }
        logInfo("Will allocate AM container, with %d MB memory including %d MB overhead".format(
          amMem,
          amMemoryOverhead))
      }

args.amMemory来自ClientArguments类，这个类中会校验输出参数：

private def validateArgs(): Unit = { if (numExecutors <= 0) { throw new IllegalArgumentException( "You must specify at least 1 executor!\n" + getUsageMessage()) } if (executorCores < sparkConf.getInt("spark.task.cpus", 1)) { throw new SparkException("Executor cores must not be less than " + "spark.task.cpus.") } if (isClusterMode) { for (key <- Seq(amMemKey, amMemOverheadKey, amCoresKey)) { if (sparkConf.contains(key)) { println(s"$key is set but does not apply in cluster mode.") } } amMemory = driverMemory amCores = driverCores } else { for (key <- Seq(driverMemOverheadKey, driverCoresKey)) { if (sparkConf.contains(key)) { println(s"$key is set but does not apply in client mode.") } } sparkConf.getOption(amMemKey) .map(Utils.memoryStringToMb) .foreach { mem => amMemory = mem } sparkConf.getOption(amCoresKey) .map(_.toInt) .foreach { cores => amCores = cores } } }

从上面代码可以看到当 isClusterMode 为true时，则args.amMemory值为driverMemory的值；否则，则从 `spark.yarn.am.memory` 中取，如果没有设置该属性，则取默认值512m。isClusterMode 为true的条件是 userClass 不为空， `def isClusterMode: Boolean = userClass != null` ，即输出参数需要有 `--class` 参数，而从下面日志可以看到ClientArguments的输出参数中并没有该参数。

15/06/08 13:57:04 DEBUG YarnClientSchedulerBackend: ClientArguments called with: --arg bj03-bi-pro-hdpnamenn:51568 --num-executors 4 --num-executors 4 --executor-memory 3g --executor-memory 3g --executor-cores 4 --executor-cores 4 --name Spark Pi

故，要想设置AM申请的内存值，要么使用cluster模式，要么在client模式中，是有 `--conf` 手动设置 `spark.yarn.am.memory` 属性，例如：

spark-submit --class org.apache.spark.examples.SparkPi \
    	--master yarn-client  \
    	--num-executors 4 \
    	--driver-memory 2g \
    	--executor-memory 3g \
    	--executor-cores 4 \
    	--conf spark.yarn.am.memory=1024m \
    	/usr/lib/spark/lib/spark-examples-1.3.0-cdh5.4.0-hadoop2.6.0-cdh5.4.0.jar \
    	100000

打开YARN管理界面，可以看到：

a. Spark Pi 应用启动了5个Container，使用了18G内存、5个CPU core

![777vQ3.jpg_web][]

b. YARN为AM启动了一个Container，占用内存为2048M

![AbIJFbQ.jpg_web][]

c. YARN启动了4个Container运行任务，每一个Container占用内存为4096M

![ja6Rv2.jpg_web][]

为什么会是 `2G +4G *4=18G` 呢？第一个Container只申请了2G内存，是因为我们的程序只为AM申请了512m内存，而 `yarn.scheduler.minimum-allocation-mb` 参数决定了最少要申请2G内存。至于其余的Container，我们设置了executor-memory内存为3G，为什么每一个Container占用内存为4096M呢？

为了找出规律，多测试几组数据，分别测试并收集executor-memory为3G、4G、5G、6G时每个executor对应的Container内存申请情况：

*  executor-memory=3g：2G+4G \* 4=18G
 *  executor-memory=4g：2G+6G \* 4=26G
 *  executor-memory=5g：2G+6G \* 4=26G
 *  executor-memory=6g：2G+8G \* 4=34G

关于这个问题，我是查看源代码，根据org.apache.spark.deploy.yarn.ApplicationMaster -> YarnRMClient -> YarnAllocator的类查找路径找到YarnAllocator中有这样一段代码：

// Executor memory in MB.
      protected val executorMemory = args.executorMemory
      // Additional memory overhead.
      protected val memoryOverhead: Int = sparkConf.getInt("spark.yarn.executor.memoryOverhead",
        math.max((MEMORY_OVERHEAD_FACTOR * executorMemory).toInt, MEMORY_OVERHEAD_MIN))
      // Number of cores per executor.
      protected val executorCores = args.executorCores
      // Resource capability requested for each executors
      private val resource = Resource.newInstance(executorMemory + memoryOverhead, executorCores)

因为没有具体的去看YARN的源代码，所以这里猜测Container的大小是根据 `executorMemory + memoryOverhead` 计算出来的，大概的规则是每一个Container的大小必须为 `yarn.scheduler.minimum-allocation-mb` 值的整数倍，当 `executor-memory=3g` 时， `executorMemory + memoryOverhead` 为3G+384M=3456M，需要申请的Container大小为 `yarn.scheduler.minimum-allocation-mb` \* 2 =4096m=4G，其他依此类推。

注意：

*  Yarn always rounds up memory requirement to multiples of `yarn.scheduler.minimum-allocation-mb` , which by default is 1024 or 1GB.
 *  Spark adds an `overhead` to `SPARK_EXECUTOR_MEMORY/SPARK_DRIVER_MEMORY` before asking Yarn for the amount.

另外，需要注意memoryOverhead的计算方法，当executorMemory的值很大时，memoryOverhead的值相应会变大，这个时候就不是384m了，相应的Container申请的内存值也变大了，例如：当executorMemory设置为90G时，memoryOverhead值为 `math.max(0.07 * 90G, 384m)=6.3G` ，其对应的Container申请的内存为98G。

回头看看给AM对应的Container分配2G内存原因，512+384=896，小于2G，故分配2G，你可以在设置 `spark.yarn.am.memory` 的值之后再来观察。

打开Spark的管理界面 [http://ip:4040][http_ip_4040] ，可以看到driver和Executor中内存的占用情况：

![eAFze2.jpg_web][]

从上图可以看到Executor占用了1566.7 MB内存，这是怎样计算出来的？参考 [Spark on Yarn: Where Have All the Memory Gone?][Spark on Yarn_ Where Have All the Memory Gone] 这篇文章，totalExecutorMemory的计算方式为：

//yarn/common/src/main/scala/org/apache/spark/deploy/yarn/YarnSparkHadoopUtil.scala
      val MEMORY_OVERHEAD_FACTOR = 0.07
      val MEMORY_OVERHEAD_MIN = 384
    
    //yarn/common/src/main/scala/org/apache/spark/deploy/yarn/YarnAllocator.scala
      protected val memoryOverhead: Int = sparkConf.getInt("spark.yarn.executor.memoryOverhead",
        math.max((MEMORY_OVERHEAD_FACTOR * executorMemory).toInt, MEMORY_OVERHEAD_MIN))
    ......
          val totalExecutorMemory = executorMemory + memoryOverhead
          numPendingAllocate.addAndGet(missing)
          logInfo(s"Will allocate $missing executor containers, each with $totalExecutorMemory MB " +
            s"memory including $memoryOverhead MB overhead")

这里我们给executor-memory设置的3G内存，memoryOverhead的值为 `math.max(0.07 * 3072, 384)=384` ，其最大可用内存通过下面代码来计算：

//core/src/main/scala/org/apache/spark/storage/BlockManager.scala /** Return the total amount of storage memory available. */ private def getMaxMemory(conf: SparkConf): Long = { val memoryFraction = conf.getDouble("spark.storage.memoryFraction", 0.6) val safetyFraction = conf.getDouble("spark.storage.safetyFraction", 0.9) (Runtime.getRuntime.maxMemory * memoryFraction * safetyFraction).toLong }

即，对于executor-memory设置3G时，executor内存占用大约为 3072m \* 0.6 \* 0.9 = 1658.88m，注意：实际上是应该乘以 `Runtime.getRuntime.maxMemory` 的值，该值小于3072m。

上图中driver占用了1060.3 MB，此时driver-memory的值是位2G，故driver中存储内存占用为：2048m \* 0.6 \* 0.9 =1105.92m，注意：实际上是应该乘以 `Runtime.getRuntime.maxMemory` 的值，该值小于2048m。

这时候，查看worker节点CoarseGrainedExecutorBackend进程启动脚本：

$ jps
    46841 Worker
    21894 CoarseGrainedExecutorBackend
    9345
    21816 ExecutorLauncher
    43369
    24300 NodeManager
    38012 JournalNode
    36929 QuorumPeerMain
    22909 Jps
    
    $ ps -ef|grep 21894
    nobody   21894 21892 99 17:28 ?        00:04:49 /usr/java/jdk1.7.0_71/bin/java -server -XX:OnOutOfMemoryError=kill %p -Xms3072m -Xmx3072m -Djava.io.tmpdir=/data/yarn/local/usercache/root/appcache/application_1433742899916_0069/container_1433742899916_0069_01_000003/tmp -Dspark.driver.port=60235 -Dspark.yarn.app.container.log.dir=/data/yarn/logs/application_1433742899916_0069/container_1433742899916_0069_01_000003 org.apache.spark.executor.CoarseGrainedExecutorBackend --driver-url akka.tcp://sparkDriver@bj03-bi-pro-hdpnamenn:60235/user/CoarseGrainedScheduler --executor-id 2 --hostname X.X.X.X --cores 4 --app-id application_1433742899916_0069 --user-class-path file:/data/yarn/local/usercache/root/appcache/application_1433742899916_0069/container_1433742899916_0069_01_000003/__app__.jar

可以看到每个CoarseGrainedExecutorBackend进程分配的内存为3072m，如果我们想查看每个executor的jvm运行情况，可以开启jmx。在/etc/spark/conf/spark-defaults.conf中添加下面一行代码：

spark.executor.extraJavaOptions -Dcom.sun.management.jmxremote.port=1099 -Dcom.sun.management.jmxremote.ssl=false -Dcom.sun.management.jmxremote.authenticate=false

然后，通过jconsole监控jvm堆内存运行情况，这样方便调试内存大小。

## 总结 ##

由上可知，在client模式下，AM对应的Container内存由 `spark.yarn.am.memory` 加上 `spark.yarn.am.memoryOverhead` 来确定，executor加上spark. `yarn.executor.memoryOverhead` 的值之后确定对应Container需要申请的内存大小，driver和executor的内存加上 `spark.yarn.driver.memoryOverhead` 或 `spark.yarn.executor.memoryOverhead` 的值之后再乘以0.54确定storage memory内存大小。在YARN中，Container申请的内存大小必须为 `yarn.scheduler.minimum-allocation-mb` 的整数倍。

下面这张图展示了Spark on YARN 内存结构，图片来自 [How-to: Tune Your Apache Spark Jobs (Part 2)][How-to_ Tune Your Apache Spark Jobs _Part 2] ：

![ZFruUvi.png_web][]

至于cluster模式下的分析，请参考上面的过程。希望这篇文章对你有所帮助！

[JavaChen_s Blog]: http://www.tuicool.com/sites/7JvUNj2
[http_blog.javachen.com_2015_06_09_memory-in-spark-on-yarn.html]: http://blog.javachen.com/2015/06/09/memory-in-spark-on-yarn.html?utm_source=tuicool&utm_medium=referral
[Spark]: http://www.tuicool.com/topics/11090132
[YARN]: http://www.tuicool.com/topics/11090119
[taobao.png]: http://www.guozhongxin.com/images/taobao.png
[777vQ3.jpg_web]: http://img2.tuicool.com/777vQ3.jpg!web
[AbIJFbQ.jpg_web]: http://img1.tuicool.com/AbIJFbQ.jpg!web
[ja6Rv2.jpg_web]: http://img2.tuicool.com/ja6Rv2.jpg!web
[http_ip_4040]: http://ip:4040/
[eAFze2.jpg_web]: http://img0.tuicool.com/eAFze2.jpg!web
[Spark on Yarn_ Where Have All the Memory Gone]: http://www.wdong.org/wordpress/blog/2015/01/08/spark-on-yarn-where-have-all-my-memory-gone/
[How-to_ Tune Your Apache Spark Jobs _Part 2]: http://blog.cloudera.com/blog/2015/03/how-to-tune-your-apache-spark-jobs-part-2/
[ZFruUvi.png_web]: http://img0.tuicool.com/ZFruUvi.png!web