hadoop - Gridmix 使所有工作失败

Question

我正在尝试在 Hadoop (2.6.1) 上运行 Gridmix

我能够在 yarn 上运行 map reduce 作业并运行 Rumen 来提取轨迹进行模拟，但是我无法完成最后一步并运行 gridmix。

如果我检查日志，我发现了一个错误:

{"org.apache.hadoop.mapreduce.jobhistory.TaskAttemptUnsuccessfulCompletion"

{"taskid":"task_1449829849459_0042_m_000000","taskType":"MAP","attemptId":"attempt_1449829849459_0042_m_000000_0","finishTime":1449841158377,"hostname":"simo","port":56154,"rackname":"/default-rack","status":"FAILED",

"error":"Error: 

java.lang.ClassNotFoundException:
  org.apache.hadoop.tools.rumen.ResourceUsageMetrics
  \n\tat java.net.URLClassLoader$1.run(URLClassLoader.java:366)
  \n\tat java.net.URLClassLoader$1.run(URLClassLoader.java:355)
  \n\tat java.security.AccessController.doPrivileged(Native Method)
  \n\tat java.net.URLClassLoader.findClass(URLClassLoader.java:354)
  \n\tat java.lang.ClassLoader.loadClass(ClassLoader.java:425)\n\tat     sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:308)
  \n\tat java.lang.ClassLoader.loadClass(ClassLoader.java:358)
  \n\tat java.lang.Class.getDeclaredConstructors0(Native Method)
  \n\tat java.lang.Class.privateGetDeclaredConstructors(Class.java:2595)
  \n\tat java.lang.Class.getConstructor0(Class.java:2895)
  \n\tat java.lang.Class.getDeclaredConstructor(Class.java:2066)
  \n\tat org.apache.hadoop.util.ReflectionUtils.newInstance(ReflectionUtils.java:125)
  \n\tat org.apache.hadoop.io.serializer.WritableSerialization$WritableDeserializer.deserialize(WritableSerialization.java:66)
  \n\tat org.apache.hadoop.io.serializer.WritableSerialization$WritableDeserializer.deserialize(WritableSerialization.java:42)
  \n\tat org.apache.hadoop.mapred.MapTask.getSplitDetails(MapTask.java:372)
  \n\tat org.apache.hadoop.mapred.MapTask.runNewMapper(MapTask.java:751)
  \n\tat org.apache.hadoop.mapred.MapTask.run(MapTask.java:341)
  \n\tat org.apache.hadoop.mapred.YarnChild$2.run(YarnChild.java:163)
  \n\tat java.security.AccessController.doPrivileged(Native Method)
  \n\tat javax.security.auth.Subject.doAs(Subject.java:415)
  \n\tat rg.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1656)
  \n\tat org.apache.hadoop.mapred.YarnChild.main(YarnChild.java:158)
\n",
"counters":
  {"org.apache.hadoop.mapreduce.jobhistory.JhCounters":
    {"name":"COUNTERS",
     "groups":
        [{"name":"org.apache.hadoop.mapreduce.TaskCounter",
          "displayName":"Map-Reduce Framework",
          "counts":[{"name":"CPU_MILLISECONDS","displayName":"CPU time spent (ms)","value":0},
                    {"name":"PHYSICAL_MEMORY_BYTES","displayName":"Physical memory (bytes) snapshot","value":0},{"name":"VIRTUAL_MEMORY_BYTES","displayName":"Virtual memory (bytes) snapshot","value":0}]}]}},"clockSplits":[286,287,287,287,287,287,286,287,287,287,287,287],"cpuUsages":[0,0,0,0,0,0,0,0,0,0,0,0],"vMemKbytes":[0,0,0,0,0,0,0,0,0,0,0,0],"physMemKbytes":[0,0,0,0,0,0,0,0,0,0,0,0]}}}

这个错误对我来说似乎很奇怪，我不太明白。

我通过命令行运行 gridmix:

bin/hadoop jar share/hadoop/tools/lib/hadoop-gridmix-2.6.1.jar iopath trace.json

和bin/hadoop classpath给我:

$ bin/hadoop classpath
/home/simo/hadoop-2.6.1/conf:
/home/simo/hadoop-2.6.1/share/hadoop/common/lib/*:
/home/simo/hadoop-2.6.1/share/hadoop/common/*:
/home/simo/hadoop-2.6.1/share/hadoop/hdfs:
/home/simo/hadoop-2.6.1/share/hadoop/hdfs/lib/*:
/home/simo/hadoop-2.6.1/share/hadoop/hdfs/*:
/home/simo/hadoop-2.6.1/share/hadoop/yarn/lib/*:
/home/simo/hadoop-2.6.1/share/hadoop/yarn/*:
/home/simo/hadoop-2.6.1/share/hadoop/mapreduce/lib/*:
/home/simo/hadoop-2.6.1/share/hadoop/mapreduce/*:
/home/simo/hadoop-2.6.1/share/hadoop/tools/lib/: <- here is hadoop-rumen-2.6.1.jar
/home/simo/hadoop-2.6.1/share/hadoop/tools/lib/*:
/usr/lib/jvm/java-7-openjdk-amd64/lib/:
/usr/lib/jvm/java-1.7.0-openjdk-amd64/lib/tools.jar:
HADOOP_CLASSPATH

其中应包含瘤胃文件。

Answer 1

异常来自任务，而不是启动 gridmix 运行的命令。因此，hadoop 类路径并不是导致此问题的真正原因。您需要将瘤胃 jar 放到您的任务类路径中。

如果您拥有集群的管理权限，则可以将 rumen jar(及其依赖项)添加到集群安装中，并确保 yarn.application.classpath 能够使用它。

或者，您可以将其作为 libjar 提供，以便它可以与您的作业一起发送到集群。您需要确保其所有必要的依赖项也包含在任务类路径中。