你的位置:首页 > 数据库

[数据库]一、Hadoop伪分布式环境搭建


Hadoop 2.x伪分布式环境搭建步骤:

1、修改hadoop-env.sh、yarn-env.sh、mapred-env.sh

方法:使用notepad++(beifeng用户)代开这三个文件

添加代码:export JAVA_HOME=/opt/modules/jdk1.7.0_67

2、修改core-site.

1)修改core-site.

<configuration>
  <property>
    <name>fs.defaultFS</name>
    <value>hdfs://Hadoop-senior02.beifeng.com:8020</value>
  </property>
  <property>
    <name>hadoop.tmp.dir</name>
    <value>/opt/modules/hadoop-2.5.0/data</value>
  </property>
</configuration>

2)修改hdfs-site.

<configuration>
  <property>
    <name>dfs.replication</name>
    <value>1</value>
  </property>
  <property>
    <name>dfs.namenode.http-address</name>
    <value>Hadoop-senior02.beifeng.com:50070</value>
  </property>
</configuration>

3)修改yarn-site.

<configuration>
  <property>
    <name>yarn.nodemanager.aux-services</name>
    <value>mapreduce_shuffle</value>
  </property>
  <property>
    <name>yarn.resourcemanager.hostname</name>
    <value>Hadoop-senior02.beifeng.com</value>
  </property>
  <property>
    <name>yarn.log-aggregation-enable</name>
    <value>true</value>
  </property>
  <property>
    <name>yarn.log-aggregation.retain-seconds</name>
    <value>86400</value>
  </property>
</configuration>

4)修改mapred-site.

<configuration>
  <property>
    <name>mapreduce.framework.name</name>
    <value>yarn</value>
  </property>
  <property>
    <name>mapreduce.jobhistory.webapp.address</name>
    <value>0.0.0.0:19888</value>
  </property>
</configuration>

3、启动hdfs

1)格式化namenode:$ bin/hdfs namenode -format

2)启动namenode:$sbin/hadoop-daemon.sh start namenode

3)启动datanode:$sbin/hadoop-daemon.sh start datanode

4)hdfs监控web页面:http://hadoop-senior02.beifeng.com:50070

4、启动yarn

1)启动resourcemanager:$sbin/yarn-daemon.sh start resourcemanager

2)启动nodemanager:sbin/yarn-daemon.sh start nodemanager

3)yarn监控web页面:http://hadoop-senior02.beifeng.com:8088

5、测试wordcount jar包

1)定位路径:/opt/modules/hadoop-2.5.0

2)代码测试:bin/yarn jar share/hadoop/mapreduce/hadoop-mapreduce-examples-2.5.0.jar wordcount /input/sort.txt /output6/

运行过程:

16/05/08 06:39:13 INFO client.RMProxy: Connecting to ResourceManager at Hadoop-senior02.beifeng.com/192.168.241.130:8032
16/05/08 06:39:15 INFO input.FileInputFormat: Total input paths to process : 1
16/05/08 06:39:15 INFO mapreduce.JobSubmitter: number of splits:1
16/05/08 06:39:15 INFO mapreduce.JobSubmitter: Submitting tokens for job: job_1462660542807_0001
16/05/08 06:39:16 INFO impl.YarnClientImpl: Submitted application application_1462660542807_0001
16/05/08 06:39:16 INFO mapreduce.Job: The url to track the job: http://Hadoop-senior02.beifeng.com:8088/proxy/application_1462660542807_0001/
16/05/08 06:39:16 INFO mapreduce.Job: Running job: job_1462660542807_0001
16/05/08 06:39:36 INFO mapreduce.Job: Job job_1462660542807_0001 running in uber mode : false
16/05/08 06:39:36 INFO mapreduce.Job: map 0% reduce 0%
16/05/08 06:39:48 INFO mapreduce.Job: map 100% reduce 0%
16/05/08 06:40:04 INFO mapreduce.Job: map 100% reduce 100%
16/05/08 06:40:04 INFO mapreduce.Job: Job job_1462660542807_0001 completed successfully
16/05/08 06:40:04 INFO mapreduce.Job: Counters: 49

3)结果查看:bin/hdfs dfs -text /output6/par*

运行结果:

hadoop 2
jps 1
mapreduce 2
yarn 1

6、MapReduce历史服务器

1)启动:sbin/mr-jobhistory-daemon.sh start historyserver

2)web ui界面:http://hadoop-senior02.beifeng.com:19888

7、hdfs、yarn、mapreduce功能

1)hdfs:分布式文件系统,高容错性的文件系统,适合部署在廉价的机器上。

             hdfs是一个主从结构,分为namenode和datanode,其中namenode是命名空间,datanode是存储空间,datanode以数据块的形式进行存储,每个数据块128M

2)yarn:通用资源管理系统,为上层应用提供统一的资源管理和调度。

       yarn分为resourcemanager和nodemanager,resourcemanager负责资源调度和分配,nodemanager负责数据处理和资源

3)mapreduce:MapReduce是一种计算模型,分为Map(映射)和Reduce(归约)。

     map将每一行数据处理后,以键值对的形式出现,并传给reduce;reduce将map传过来的数据进行汇总和统计。