Hadoop 2.6.0 在 WordCount 示例中无法减少任务
Hadoop 2.6.0 does not work reduce tasks in WordCount example
我已经在多个节点（物理）上安装了 Hadoop 集群设置。
我有一台用于 NameNode、ResourceManager 和 JobHistory 服务器的服务器。
我有两台用于 DataNode 的服务器。配置的时候我跟着this tutorial
我尝试测试 MapReduce 程序，例如 WordCount、Terasoft、Teragen 等，我都可以从 hadoop/share/hadoop/mapreduce/hadoop-mapreduce-examples-2.6.0.jar
启动
因此，我启动了 Teragen 和 randomwriter，它们以成功状态完成（因为没有 Reduce 任务，只有 Map 任务），但是当我尝试启动 WordCount 或 WordMean 时，Map 任务完成（1 个任务），但是 Reduce始终为 0%。它只是停止完成。在 yarn-root-resourcemanager-yamaster.log 成功映射任务后，我只看到一行：
INFO org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler: Null container completed...
我试图找到解决方案，我找到了 similar question on SOF, 但没有正确答案，实际上我不知道如何在资源管理器中查看免费的 reducers。我有什么：
- Hadoop 网络界面：master:50070
- 资源管理器：master:8088
- JobHistory 服务器：master:19888/jobhistory
更新：
我尝试使用 key -D mapd.reduce.tasks=0:
在不减少任务的情况下启动 wordcount 示例程序
hadoop jar hadoop-mapreduce-examples-2.6.0.jar wordcount -D mapd.reduce.tasks=0 /bigtext.txt /bigtext_wc_1.txt
并且有效。我得到了字数统计结果。错了，因为没有减少，但我的程序完成了。
15/02/03 12:40:37 INFO mapreduce.Job: Running job: job_1422950901990_0004
15/02/03 12:40:52 INFO mapreduce.Job: Job job_1422950901990_0004 running in uber mode : false
15/02/03 12:40:52 INFO mapreduce.Job: map 0% reduce 0%
15/02/03 12:41:03 INFO mapreduce.Job: map 100% reduce 0%
15/02/03 12:41:04 INFO mapreduce.Job: Job job_1422950901990_0004 completed successfully
15/02/03 12:41:05 INFO mapreduce.Job: Counters: 30
更新#2：
来自应用程序日志的更多信息：
2015-02-03 15:02:12,008 INFO [IPC Server handler 0 on 55452] org.apache.hadoop.mapred.TaskAttemptListenerImpl: Progress of TaskAttempt attempt_1422959549820_0005_m_000000_0 is : 1.0
2015-02-03 15:02:12,025 INFO [IPC Server handler 1 on 55452] org.apache.hadoop.mapred.TaskAttemptListenerImpl: Done acknowledgement from attempt_1422959549820_0005_m_000000_0
2015-02-03 15:02:12,028 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: attempt_1422959549820_0005_m_000000_0 TaskAttempt Transitioned from RUNNING to SUCCESS_CONTAINER_CLEANUP
2015-02-03 15:02:12,029 INFO [ContainerLauncher #1] org.apache.hadoop.mapreduce.v2.app.launcher.ContainerLauncherImpl: Processing the event EventType: CONTAINER_REMOTE_CLEANUP for container container_1422959549820_0005_01_000002 taskAttempt attempt_1422959549820_0005_m_000000_0
2015-02-03 15:02:12,030 INFO [ContainerLauncher #1] org.apache.hadoop.mapreduce.v2.app.launcher.ContainerLauncherImpl: KILLING attempt_1422959549820_0005_m_000000_0
2015-02-03 15:02:12,030 INFO [ContainerLauncher #1] org.apache.hadoop.yarn.client.api.impl.ContainerManagementProtocolProxy: Opening proxy : slave102.hadoop.ot.ru:51573
2015-02-03 15:02:12,063 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: attempt_1422959549820_0005_m_000000_0 TaskAttempt Transitioned from SUCCESS_CONTAINER_CLEANUP to SUCCEEDED
2015-02-03 15:02:12,084 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskImpl: Task succeeded with attempt attempt_1422959549820_0005_m_000000_0
2015-02-03 15:02:12,087 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskImpl: task_1422959549820_0005_m_000000 Task Transitioned from RUNNING to SUCCEEDED
2015-02-03 15:02:12,094 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl: Num completed Tasks: 1
2015-02-03 15:02:12,792 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: Before Scheduling: PendingReds:1 ScheduledMaps:0 ScheduledReds:0 AssignedMaps:1 AssignedReds:0 CompletedMaps:1 CompletedReds:0 ContAlloc:1 ContRel:0 HostLocal:1 RackLocal:0
2015-02-03 15:02:12,794 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: Recalculating schedule, headroom=<memory:4096, vCores:-1>
2015-02-03 15:02:12,794 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: Reduce slow start threshold reached. Scheduling reduces.
2015-02-03 15:02:12,795 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: All maps assigned. Ramping up all remaining reduces:1
2015-02-03 15:02:12,795 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: After Scheduling: PendingReds:0 ScheduledMaps:0 ScheduledReds:1 AssignedMaps:1 AssignedReds:0 CompletedMaps:1 CompletedReds:0 ContAlloc:1 ContRel:0 HostLocal:1 RackLocal:0
2015-02-03 15:02:13,805 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerRequestor: getResources() for application_1422959549820_0005: ask=1 release= 0 newContainers=0 finishedContainers=1 resourcelimit=<memory:6144, vCores:0> knownNMs=4
2015-02-03 15:02:13,806 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: Received completed container container_1422959549820_0005_01_000002
2015-02-03 15:02:13,808 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: After Scheduling: PendingReds:0 ScheduledMaps:0 ScheduledReds:1 AssignedMaps:0 AssignedReds:0 CompletedMaps:1 CompletedReds:0 ContAlloc:1 ContRel:0 HostLocal:1 RackLocal:0
2015-02-03 15:02:13,808 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: Diagnostics report from attempt_1422959549820_0005_m_000000_0: Container killed by the ApplicationMaster.
Container killed on request. Exit code is 143
Container exited with a non-zero exit code 143
集群的配置文件。
hdfs-site.xml
<configuration>
<!-- Properties for NameNode -->
<property>
<name>dfs.namenode.name.dir</name>
<value>/grid/hadoop1/nn</value>
<description>Path on the local filesystem where the NameNode stores the namespace and transactions logs persistently. If this is a comma-delimited list of directories then the name table is replicated in all of the directories, for redundancy.</description>
</property>
<property>
<name>dfs.namenode.hosts</name>
<value>/opt/current/hadoop/etc/hadoop/slaves</value>
<description>List of permitted DataNodes.If necessary, use these files to control the list of allowable datanodes.</description>
</property>
<property>
<name>dfs.namenode.hosts.exclude</name>
<value>/opt/current/hadoop/etc/hadoop/excludes</value>
<description>List of excluded DataNodes. If necessary, use these files to control the list of allowable datanodes.</description>
</property>
<property>
<name>dfs.blocksize</name>
<value>268435456</value>
<description>HDFS blocksize of 256MB for large file-systems.</description>
</property>
<property>
<name>dfs.namenode.handler.count</name>
<value>100</value>
<description>More NameNode server threads to handle RPCs from large number of DataNodes.</description>
</property>
<!-- Properties for DataNode -->
<property>
<name>dfs.datanode.data.dir</name>
<value>/grid/hadoop1/dn</value>
<description>Comma separated list of paths on the local filesystem of a DataNode where it should store its blocks. If this is a comma-delimited list of directories, then data will be stored in all named directories, typically on different devices.</description>
</property>
</configuration>
核心-site.xml
<configuration>
<property>
<name>fs.defaultFS</name>
<value>hdfs://master:8020</value>
<description>Default hdfs filesystem on namenode host like - hdfs://host:port/</description>
</property>
<property>
<name>io.file.buffer.size</name>
<value>131072</value>
<description>Size of read/write buffer used in SequenceFiles.</description>
</property>
</configuration>
mapred-site.xml
<configuration>
<!-- Configurations for MapReduce Applications -->
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
<description>Execution framework set to Hadoop YARN.</description>
</property>
<property>
<name>mapreduce.map.memory.mb</name>
<value>1536</value>
<description>Larger resource limit for maps.</description>
</property>
<property>
<name>mapreduce.map.java.opts</name>
<value>-Xmx1024M</value>
<description>Larger heap-size for child jvms of maps.</description>
</property>
<property>
<name>mapreduce.reduce.memory.mb</name>
<value>3072</value>
<description>Larger resource limit for reduces.</description>
</property>
<property>
<name>mapreduce.reduce.java.opts</name>
<value>-Xmx2560M</value>
<description>Larger heap-size for child jvms of reduces.</description>
</property>
<property>
<name>mapreduce.task.io.sort.mb</name>
<value>512</value>
<description>Higher memory-limit while sorting data for efficiency.</description>
</property>
<property>
<name>mapreduce.task.io.sort.factor</name>
<value>100</value>
<description>More streams merged at once while sorting files.</description>
</property>
<property>
<name>mapreduce.reduce.shuffle.parallelcopies</name>
<value>50</value>
<description>Higher number of parallel copies run by reduces to fetch outputs from very large number of maps.</description>
</property>
<!-- Configurations for MapReduce JobHistory Server -->
<property>
<name>mapreduce.jobhistory.address</name>
<value>master:10020</value>
<description>MapReduce JobHistory Server host:port. Default port is 10020.</description>
</property>
<property>
<name>mapreduce.jobhistory.webapp.address</name>
<value>master:19888</value>
<description>MapReduce JobHistory Server Web UI host:port. Default port is 19888.</description>
</property>
<property>
<name>mapreduce.jobhistory.intermediate-done-dir</name>
<value>/mr-history/tmp</value>
<description>Directory where history files are written by MapReduce jobs.</description>
</property>
<property>
<name>mapreduce.jobhistory.done-dir</name>
<value>/mr-history/done</value>
<description>Directory where history files are managed by the MR JobHistory Server.</description>
</property>
</configuration>
纱-site.xml
<configuration>
<!-- Site specific YARN configuration properties -->
<!-- Configurations for ResourceManager and NodeManager -->
<property>
<name>yarn.acl.enable</name>
<value>yes</value>
<description>Enable ACLs? Defaults to false.</description>
</property>
<property>
<name>yarn.admin.acl</name>
<value>false</value>
<description>ACL to set admins on the cluster. ACLs are of for comma-separated-usersspacecomma-separated-groups. Defaults to special value of * which means anyone. Special value of just space means no one has access.</description>
</property>
<property>
<name>yarn.log-aggregation-enable</name>
<value>false</value>
<description>Configuration to enable or disable log aggregation</description>
</property>
<!-- Configurations for ResourceManager -->
<property>
<name>yarn.resourcemanager.address</name>
<value>master:8050</value>
<description>Value: host:port. If set, overrides the hostname set in yarn.resourcemanager.hostname.</description>
</property>
<property>
<name>yarn.resourcemanager.scheduler.address</name>
<value>master:8030</value>
<description>ResourceManager host:port for ApplicationMasters to talk to Scheduler to obtain resources. If set, overrides the hostname set in yarn.resourcemanager.hostname.</description>
</property>
<property>
<name>yarn.resourcemanager.resource-tracker.address</name>
<value>master:8025</value>
<description>ResourceManager host:port for NodeManagers. If set, overrides the hostname set in yarn.resourcemanager.hostname.</description>
</property>
<property>
<name>yarn.resourcemanager.admin.address</name>
<value>master:8141</value>
<description>ResourceManager host:port for administrative commands. If set, overrides the hostname set in yarn.resourcemanager.hostname.</description>
</property>
<property>
<name>yarn.resourcemanager.webapp.address</name>
<value>master:8088</value>
<description>web-ui host:port. If set, overrides the hostname set in</description>
</property>
<property>
<name>yarn.resourcemanager.hostname</name>
<value>master</value>
<description>ResourceManager host</description>
</property>
<property>
<name>yarn.resourcemanager.scheduler.class</name>
<value>org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler</value>
<description>ResourceManager Scheduler class.</description>
</property>
<property>
<name>yarn.scheduler.maximum-allocation-mb</name>
<value>6144</value>
<description>Maximum limit of memory to allocate to each container request at the Resource Manager. In MBs</description>
</property>
<property>
<name>yarn.scheduler.minimum-allocation-mb</name>
<value>2048</value>
<description>Minimum limit of memory to allocate to each container request at the Resource Manager. In MBs</description>
</property>
<property>
<name>yarn.resourcemanager.nodes.include-path</name>
<value>/opt/current/hadoop/etc/hadoop/slaves</value>
<description>List of permitted NodeManagers. If necessary, use these files to control the list of allowable NodeManagers.</description>
</property>
<property>
<name>yarn.resourcemanager.nodes.exclude-path</name>
<value>/opt/current/hadoop/etc/hadoop/excludes</value>
<description>List of excluded NodeManagers. If necessary, use these files to control the list of allowable NodeManagers.</description>
</property>
<!-- Configurations for NodeManager -->
<property>
<name>yarn.nodemanager.resource.memory-mb</name>
<value>2048</value>
<description>Resource i.e. available physical memory, in MB, for given NodeManager. Defines total available resources on the NodeManager to be made available to running containers</description>
</property>
<property>
<name>yarn.nodemanager.vmem-pmem-ratio</name>
<value>2.1</value>
<description>Maximum ratio by which virtual memory usage of tasks may exceed physical memory. The virtual memory usage of each task may exceed its physical memory limit by this ratio. The total amount of virtual memory used by tasks on the NodeManager may exceed its physical memory usage by this ratio.</description>
</property>
<property>
<name>yarn.nodemanager.local-dirs</name>
<value>/grid/hadoop1/yarn/local</value>
<description>Comma-separated list of paths on the local filesystem where intermediate data is written.Multiple paths help spread disk i/o.</description>
</property>
<property>
<name>yarn.nodemanager.log-dirs</name>
<value>/var/log/hadoop-yarn/containers</value>
<description>Where to store container logs.</description>
</property>
<property>
<name>yarn.nodemanager.log.retain-second</name>
<value>10800</value>
<description>Default time (in seconds) to retain log files on the NodeManager Only applicable if log-aggregation is disabled.</description>
</property>
<property>
<name>yarn.nodemanager.remote-app-log-dir</name>
<value>/logs</value>
<description>HDFS directory where the application logs are moved on application completion. Need to set appropriate permissions. Only applicable if log-aggregation is enabled.</description>
</property>
<property>
<name>yarn.nodemanager.remote-app-log-dir-suffix</name>
<value>logs</value>
<description>Suffix appended to the remote log dir. Logs will be aggregated to ${yarn.nodemanager.remote-app-log-dir}/${user}/${thisParam} Only applicable if log-aggregation is enabled.</description>
</property>
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
<description>Shuffle service that needs to be set for Map Reduce applications.</description>
</property>
</configuration>
最后 /etc/hosts:
127.0.0.1 localhost
## BigData Hadoop Lab ##
#Name Node
172.25.28.100 master.hadoop.ot.ru master
172.25.28.101 secondary.hadoop.ot.ru secondary
#DataNodes on DL Servers
172.25.28.102 slave102.hadoop.ot.ru slave102
172.25.28.103 slave103.hadoop.ot.ru slave103
172.25.28.104 slave104.hadoop.ot.ru slave104
172.25.28.105 slave105.hadoop.ot.ru slave105
172.25.28.106 slave106.hadoop.ot.ru slave106
172.25.28.107 slave107.hadoop.ot.ru slave107
#DataNodes on ARM Servers
172.25.40.25 slave25.hadoop.ot.ru slave25
172.25.40.26 slave26.hadoop.ot.ru slave26
172.25.40.27 slave27.hadoop.ot.ru slave27
172.25.40.28 slave28.hadoop.ot.ru slave28
答案是内存不够。每个任务容器（map 或 reduce）对我的机器来说太大了。
这个错误：
2015-02-03 15:02:13,808 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: Diagnostics report from attempt_1422959549820_0005_m_000000_0: Container killed by the ApplicationMaster.
Container killed on request. Exit code is 143
Container exited with a non-zero exit code 143
告诉我这件事。
这是我的大多数服务器的最佳设置：
yarn.scheduler.minimum-allocation-mb=768
yarn.scheduler.maximum-allocation-mb=3072
yarn.nodemanager.resource.memory-mb=3072
mapreduce.map.memory.mb=768
mapreduce.map.java.opts=-Xmx512m
mapreduce.reduce.memory.mb=1536
mapreduce.reduce.java.opts=-Xmx1024m
yarn.app.mapreduce.am.resource.mb=768
yarn.app.mapreduce.am.command-opts=-Xmx512m
我已经在多个节点（物理）上安装了 Hadoop 集群设置。 我有一台用于 NameNode、ResourceManager 和 JobHistory 服务器的服务器。 我有两台用于 DataNode 的服务器。配置的时候我跟着this tutorial
我尝试测试 MapReduce 程序，例如 WordCount、Terasoft、Teragen 等，我都可以从 hadoop/share/hadoop/mapreduce/hadoop-mapreduce-examples-2.6.0.jar
因此，我启动了 Teragen 和 randomwriter，它们以成功状态完成（因为没有 Reduce 任务，只有 Map 任务），但是当我尝试启动 WordCount 或 WordMean 时，Map 任务完成（1 个任务），但是 Reduce始终为 0%。它只是停止完成。在 yarn-root-resourcemanager-yamaster.log 成功映射任务后，我只看到一行：
INFO org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler: Null container completed...
我试图找到解决方案，我找到了 similar question on SOF, 但没有正确答案，实际上我不知道如何在资源管理器中查看免费的 reducers。我有什么：
- Hadoop 网络界面：master:50070
- 资源管理器：master:8088
- JobHistory 服务器：master:19888/jobhistory
更新： 我尝试使用 key -D mapd.reduce.tasks=0:
在不减少任务的情况下启动 wordcount 示例程序hadoop jar hadoop-mapreduce-examples-2.6.0.jar wordcount -D mapd.reduce.tasks=0 /bigtext.txt /bigtext_wc_1.txt
并且有效。我得到了字数统计结果。错了，因为没有减少，但我的程序完成了。
15/02/03 12:40:37 INFO mapreduce.Job: Running job: job_1422950901990_0004
15/02/03 12:40:52 INFO mapreduce.Job: Job job_1422950901990_0004 running in uber mode : false
15/02/03 12:40:52 INFO mapreduce.Job: map 0% reduce 0%
15/02/03 12:41:03 INFO mapreduce.Job: map 100% reduce 0%
15/02/03 12:41:04 INFO mapreduce.Job: Job job_1422950901990_0004 completed successfully
15/02/03 12:41:05 INFO mapreduce.Job: Counters: 30
更新#2：
来自应用程序日志的更多信息：
2015-02-03 15:02:12,008 INFO [IPC Server handler 0 on 55452] org.apache.hadoop.mapred.TaskAttemptListenerImpl: Progress of TaskAttempt attempt_1422959549820_0005_m_000000_0 is : 1.0
2015-02-03 15:02:12,025 INFO [IPC Server handler 1 on 55452] org.apache.hadoop.mapred.TaskAttemptListenerImpl: Done acknowledgement from attempt_1422959549820_0005_m_000000_0
2015-02-03 15:02:12,028 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: attempt_1422959549820_0005_m_000000_0 TaskAttempt Transitioned from RUNNING to SUCCESS_CONTAINER_CLEANUP
2015-02-03 15:02:12,029 INFO [ContainerLauncher #1] org.apache.hadoop.mapreduce.v2.app.launcher.ContainerLauncherImpl: Processing the event EventType: CONTAINER_REMOTE_CLEANUP for container container_1422959549820_0005_01_000002 taskAttempt attempt_1422959549820_0005_m_000000_0
2015-02-03 15:02:12,030 INFO [ContainerLauncher #1] org.apache.hadoop.mapreduce.v2.app.launcher.ContainerLauncherImpl: KILLING attempt_1422959549820_0005_m_000000_0
2015-02-03 15:02:12,030 INFO [ContainerLauncher #1] org.apache.hadoop.yarn.client.api.impl.ContainerManagementProtocolProxy: Opening proxy : slave102.hadoop.ot.ru:51573
2015-02-03 15:02:12,063 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: attempt_1422959549820_0005_m_000000_0 TaskAttempt Transitioned from SUCCESS_CONTAINER_CLEANUP to SUCCEEDED
2015-02-03 15:02:12,084 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskImpl: Task succeeded with attempt attempt_1422959549820_0005_m_000000_0
2015-02-03 15:02:12,087 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskImpl: task_1422959549820_0005_m_000000 Task Transitioned from RUNNING to SUCCEEDED
2015-02-03 15:02:12,094 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl: Num completed Tasks: 1
2015-02-03 15:02:12,792 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: Before Scheduling: PendingReds:1 ScheduledMaps:0 ScheduledReds:0 AssignedMaps:1 AssignedReds:0 CompletedMaps:1 CompletedReds:0 ContAlloc:1 ContRel:0 HostLocal:1 RackLocal:0
2015-02-03 15:02:12,794 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: Recalculating schedule, headroom=<memory:4096, vCores:-1>
2015-02-03 15:02:12,794 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: Reduce slow start threshold reached. Scheduling reduces.
2015-02-03 15:02:12,795 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: All maps assigned. Ramping up all remaining reduces:1
2015-02-03 15:02:12,795 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: After Scheduling: PendingReds:0 ScheduledMaps:0 ScheduledReds:1 AssignedMaps:1 AssignedReds:0 CompletedMaps:1 CompletedReds:0 ContAlloc:1 ContRel:0 HostLocal:1 RackLocal:0
2015-02-03 15:02:13,805 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerRequestor: getResources() for application_1422959549820_0005: ask=1 release= 0 newContainers=0 finishedContainers=1 resourcelimit=<memory:6144, vCores:0> knownNMs=4
2015-02-03 15:02:13,806 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: Received completed container container_1422959549820_0005_01_000002
2015-02-03 15:02:13,808 INFO [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: After Scheduling: PendingReds:0 ScheduledMaps:0 ScheduledReds:1 AssignedMaps:0 AssignedReds:0 CompletedMaps:1 CompletedReds:0 ContAlloc:1 ContRel:0 HostLocal:1 RackLocal:0
2015-02-03 15:02:13,808 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: Diagnostics report from attempt_1422959549820_0005_m_000000_0: Container killed by the ApplicationMaster.
Container killed on request. Exit code is 143
Container exited with a non-zero exit code 143
集群的配置文件。
hdfs-site.xml
<configuration>
<!-- Properties for NameNode -->
<property>
<name>dfs.namenode.name.dir</name>
<value>/grid/hadoop1/nn</value>
<description>Path on the local filesystem where the NameNode stores the namespace and transactions logs persistently. If this is a comma-delimited list of directories then the name table is replicated in all of the directories, for redundancy.</description>
</property>
<property>
<name>dfs.namenode.hosts</name>
<value>/opt/current/hadoop/etc/hadoop/slaves</value>
<description>List of permitted DataNodes.If necessary, use these files to control the list of allowable datanodes.</description>
</property>
<property>
<name>dfs.namenode.hosts.exclude</name>
<value>/opt/current/hadoop/etc/hadoop/excludes</value>
<description>List of excluded DataNodes. If necessary, use these files to control the list of allowable datanodes.</description>
</property>
<property>
<name>dfs.blocksize</name>
<value>268435456</value>
<description>HDFS blocksize of 256MB for large file-systems.</description>
</property>
<property>
<name>dfs.namenode.handler.count</name>
<value>100</value>
<description>More NameNode server threads to handle RPCs from large number of DataNodes.</description>
</property>
<!-- Properties for DataNode -->
<property>
<name>dfs.datanode.data.dir</name>
<value>/grid/hadoop1/dn</value>
<description>Comma separated list of paths on the local filesystem of a DataNode where it should store its blocks. If this is a comma-delimited list of directories, then data will be stored in all named directories, typically on different devices.</description>
</property>
</configuration>
核心-site.xml
<configuration>
<property>
<name>fs.defaultFS</name>
<value>hdfs://master:8020</value>
<description>Default hdfs filesystem on namenode host like - hdfs://host:port/</description>
</property>
<property>
<name>io.file.buffer.size</name>
<value>131072</value>
<description>Size of read/write buffer used in SequenceFiles.</description>
</property>
</configuration>
mapred-site.xml
<configuration>
<!-- Configurations for MapReduce Applications -->
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
<description>Execution framework set to Hadoop YARN.</description>
</property>
<property>
<name>mapreduce.map.memory.mb</name>
<value>1536</value>
<description>Larger resource limit for maps.</description>
</property>
<property>
<name>mapreduce.map.java.opts</name>
<value>-Xmx1024M</value>
<description>Larger heap-size for child jvms of maps.</description>
</property>
<property>
<name>mapreduce.reduce.memory.mb</name>
<value>3072</value>
<description>Larger resource limit for reduces.</description>
</property>
<property>
<name>mapreduce.reduce.java.opts</name>
<value>-Xmx2560M</value>
<description>Larger heap-size for child jvms of reduces.</description>
</property>
<property>
<name>mapreduce.task.io.sort.mb</name>
<value>512</value>
<description>Higher memory-limit while sorting data for efficiency.</description>
</property>
<property>
<name>mapreduce.task.io.sort.factor</name>
<value>100</value>
<description>More streams merged at once while sorting files.</description>
</property>
<property>
<name>mapreduce.reduce.shuffle.parallelcopies</name>
<value>50</value>
<description>Higher number of parallel copies run by reduces to fetch outputs from very large number of maps.</description>
</property>
<!-- Configurations for MapReduce JobHistory Server -->
<property>
<name>mapreduce.jobhistory.address</name>
<value>master:10020</value>
<description>MapReduce JobHistory Server host:port. Default port is 10020.</description>
</property>
<property>
<name>mapreduce.jobhistory.webapp.address</name>
<value>master:19888</value>
<description>MapReduce JobHistory Server Web UI host:port. Default port is 19888.</description>
</property>
<property>
<name>mapreduce.jobhistory.intermediate-done-dir</name>
<value>/mr-history/tmp</value>
<description>Directory where history files are written by MapReduce jobs.</description>
</property>
<property>
<name>mapreduce.jobhistory.done-dir</name>
<value>/mr-history/done</value>
<description>Directory where history files are managed by the MR JobHistory Server.</description>
</property>
</configuration>
纱-site.xml
<configuration>
<!-- Site specific YARN configuration properties -->
<!-- Configurations for ResourceManager and NodeManager -->
<property>
<name>yarn.acl.enable</name>
<value>yes</value>
<description>Enable ACLs? Defaults to false.</description>
</property>
<property>
<name>yarn.admin.acl</name>
<value>false</value>
<description>ACL to set admins on the cluster. ACLs are of for comma-separated-usersspacecomma-separated-groups. Defaults to special value of * which means anyone. Special value of just space means no one has access.</description>
</property>
<property>
<name>yarn.log-aggregation-enable</name>
<value>false</value>
<description>Configuration to enable or disable log aggregation</description>
</property>
<!-- Configurations for ResourceManager -->
<property>
<name>yarn.resourcemanager.address</name>
<value>master:8050</value>
<description>Value: host:port. If set, overrides the hostname set in yarn.resourcemanager.hostname.</description>
</property>
<property>
<name>yarn.resourcemanager.scheduler.address</name>
<value>master:8030</value>
<description>ResourceManager host:port for ApplicationMasters to talk to Scheduler to obtain resources. If set, overrides the hostname set in yarn.resourcemanager.hostname.</description>
</property>
<property>
<name>yarn.resourcemanager.resource-tracker.address</name>
<value>master:8025</value>
<description>ResourceManager host:port for NodeManagers. If set, overrides the hostname set in yarn.resourcemanager.hostname.</description>
</property>
<property>
<name>yarn.resourcemanager.admin.address</name>
<value>master:8141</value>
<description>ResourceManager host:port for administrative commands. If set, overrides the hostname set in yarn.resourcemanager.hostname.</description>
</property>
<property>
<name>yarn.resourcemanager.webapp.address</name>
<value>master:8088</value>
<description>web-ui host:port. If set, overrides the hostname set in</description>
</property>
<property>
<name>yarn.resourcemanager.hostname</name>
<value>master</value>
<description>ResourceManager host</description>
</property>
<property>
<name>yarn.resourcemanager.scheduler.class</name>
<value>org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler</value>
<description>ResourceManager Scheduler class.</description>
</property>
<property>
<name>yarn.scheduler.maximum-allocation-mb</name>
<value>6144</value>
<description>Maximum limit of memory to allocate to each container request at the Resource Manager. In MBs</description>
</property>
<property>
<name>yarn.scheduler.minimum-allocation-mb</name>
<value>2048</value>
<description>Minimum limit of memory to allocate to each container request at the Resource Manager. In MBs</description>
</property>
<property>
<name>yarn.resourcemanager.nodes.include-path</name>
<value>/opt/current/hadoop/etc/hadoop/slaves</value>
<description>List of permitted NodeManagers. If necessary, use these files to control the list of allowable NodeManagers.</description>
</property>
<property>
<name>yarn.resourcemanager.nodes.exclude-path</name>
<value>/opt/current/hadoop/etc/hadoop/excludes</value>
<description>List of excluded NodeManagers. If necessary, use these files to control the list of allowable NodeManagers.</description>
</property>
<!-- Configurations for NodeManager -->
<property>
<name>yarn.nodemanager.resource.memory-mb</name>
<value>2048</value>
<description>Resource i.e. available physical memory, in MB, for given NodeManager. Defines total available resources on the NodeManager to be made available to running containers</description>
</property>
<property>
<name>yarn.nodemanager.vmem-pmem-ratio</name>
<value>2.1</value>
<description>Maximum ratio by which virtual memory usage of tasks may exceed physical memory. The virtual memory usage of each task may exceed its physical memory limit by this ratio. The total amount of virtual memory used by tasks on the NodeManager may exceed its physical memory usage by this ratio.</description>
</property>
<property>
<name>yarn.nodemanager.local-dirs</name>
<value>/grid/hadoop1/yarn/local</value>
<description>Comma-separated list of paths on the local filesystem where intermediate data is written.Multiple paths help spread disk i/o.</description>
</property>
<property>
<name>yarn.nodemanager.log-dirs</name>
<value>/var/log/hadoop-yarn/containers</value>
<description>Where to store container logs.</description>
</property>
<property>
<name>yarn.nodemanager.log.retain-second</name>
<value>10800</value>
<description>Default time (in seconds) to retain log files on the NodeManager Only applicable if log-aggregation is disabled.</description>
</property>
<property>
<name>yarn.nodemanager.remote-app-log-dir</name>
<value>/logs</value>
<description>HDFS directory where the application logs are moved on application completion. Need to set appropriate permissions. Only applicable if log-aggregation is enabled.</description>
</property>
<property>
<name>yarn.nodemanager.remote-app-log-dir-suffix</name>
<value>logs</value>
<description>Suffix appended to the remote log dir. Logs will be aggregated to ${yarn.nodemanager.remote-app-log-dir}/${user}/${thisParam} Only applicable if log-aggregation is enabled.</description>
</property>
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
<description>Shuffle service that needs to be set for Map Reduce applications.</description>
</property>
</configuration>
最后 /etc/hosts:
127.0.0.1 localhost
## BigData Hadoop Lab ##
#Name Node
172.25.28.100 master.hadoop.ot.ru master
172.25.28.101 secondary.hadoop.ot.ru secondary
#DataNodes on DL Servers
172.25.28.102 slave102.hadoop.ot.ru slave102
172.25.28.103 slave103.hadoop.ot.ru slave103
172.25.28.104 slave104.hadoop.ot.ru slave104
172.25.28.105 slave105.hadoop.ot.ru slave105
172.25.28.106 slave106.hadoop.ot.ru slave106
172.25.28.107 slave107.hadoop.ot.ru slave107
#DataNodes on ARM Servers
172.25.40.25 slave25.hadoop.ot.ru slave25
172.25.40.26 slave26.hadoop.ot.ru slave26
172.25.40.27 slave27.hadoop.ot.ru slave27
172.25.40.28 slave28.hadoop.ot.ru slave28
答案是内存不够。每个任务容器（map 或 reduce）对我的机器来说太大了。
这个错误：
2015-02-03 15:02:13,808 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: Diagnostics report from attempt_1422959549820_0005_m_000000_0: Container killed by the ApplicationMaster.
Container killed on request. Exit code is 143
Container exited with a non-zero exit code 143
告诉我这件事。
这是我的大多数服务器的最佳设置：
yarn.scheduler.minimum-allocation-mb=768
yarn.scheduler.maximum-allocation-mb=3072
yarn.nodemanager.resource.memory-mb=3072
mapreduce.map.memory.mb=768
mapreduce.map.java.opts=-Xmx512m
mapreduce.reduce.memory.mb=1536
mapreduce.reduce.java.opts=-Xmx1024m
yarn.app.mapreduce.am.resource.mb=768
yarn.app.mapreduce.am.command-opts=-Xmx512m