hadoop HA集群的安装

1.hadoop集群规化

ip	主机名	安装软件	角色	运行进程
10.124.147.22	hadoop1	jdk、zookeeper、hadoop	namenode/zookeeper/jobhistoryserver	DFSZKFailoverController、NameNode、JobHistoryServer、QuorumPeerMain
10.124.147.23	hadoop2	jdk、zookeeper、hadoop	namenode/zookeeper	DFSZKFailoverController、NameNode、QuorumPeerMain
10.124.147.32	hadoop3	jdk、zookeeper、hadoop	resourcemanager/zookeeper	ResourceManager、QuorumPeerMain
10.124.147.33	hadoop4	jdk、zookeeper、hadoop	resourcemanager/zookeeper	ResourceManager、QuorumPeerMain
10.110.92.161	hadoop5	jdk、hadoop	datanode/journalnode	NodeManager、JournalNode、DataNode
10.110.92.162	hadoop6	jdk、hadoop	datanode/journalnode	NodeManager、JournalNode、DataNode
10.122.147.37	hadoop7	jdk、hadoop	datanode/journalnode	NodeManager、JournalNode、DataNode

2.基本环境

system os: centos 6.5

hadoop: 2.7.3

zoopkeeper: 3.4.12

jdk: 1.8.0

3.环境准备

3.1 hosts设定

[root@10-124-147-23 local]# cat /etc/hosts
127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4
10.124.147.22 hadoop1 10-124-147-22
10.124.147.23 hadoop2 10-124-147-23
10.124.147.32 hadoop3 10-124-147-32
10.124.147.33 hadoop4 10-124-147-33
10.110.92.161 hadoop5 10-110-92-161
10.110.92.162 hadoop6 10-110-92-162
10.122.147.37 hadoop7 10-122-147-37

在此需要注意两点

1. 127.0.0.1之后不要有主机名，比如上面的10-124-147-22的
2. 最好将ipv6地址栏的localhosts删除
3. 此处除了hadoop1之外，我还设定了10-124-147-22，是因为不想更改主机名，实际实际的时候，直接进行hostname更改即可

3.2 java环境安装

3.2.1 jdk安装包解压

[root@10-124-147-23 letv]# tar xvf jdk-8u141-linux-x64.tar.gz 
[root@10-124-147-23 letv]# ln -svfn /letv/jdk1.8.0_141 /usr/local/java

3.2.2 profile环境的变更

[root@10-124-147-23 letv]# tail -3 /etc/profile
export JAVA_HOME=/usr/local/java
export HADOOP_HOME=/usr/local/hadoop
export PATH=$HADOOP_HOME/bin:$JAVA_HOME/bin:$PATH

[root@10-124-147-23 letv]# source /etc/profile

3.3 zookeeper集群的安装

3.3.1 zookeeper安装包的解压

[root@10-124-147-23 letv]# tar xvf zookeeper-3.4.12.tar.gz
[root@10-124-147-23 letv]# ln -svnf /letv/zookeeper-3.4.12 /usr/local/zookeeper
[root@10-124-147-23 letv]# cd /usr/local/zookeeper/conf
[root@10-124-147-23 conf]# ll
total 16
-rw-rw-r-- 1 1000 1000  535 Mar 27 12:32 configuration.xsl
-rw-rw-r-- 1 1000 1000 2161 Mar 27 12:32 log4j.properties
-rw-rw-r-- 1 1000 1000  922 Mar 27 12:32 zoo_sample.cfg
[root@10-124-147-23 conf]# cp zoo_sample.cfg zoo.cfg

3.3.2 zoo.cfg配置文件修改

[root@10-124-147-23 conf]# grep  ^[^#] zoo.cfg
tickTime=2000
initLimit=10
syncLimit=5
dataDir=/usr/local/zookeeper/data
clientPort=2181
server.1=hadoop1:2888:3888
server.2=hadoop2:2888:3888
server.3=hadoop3:2888:3888
server.4=hadoop4:2888:3888

修改dataDir值，因为同时要建立zookeeper集群，下面写下对应的server地址

[root@10-124-147-23 conf]# echo 1 > /usr/local/zookeeper/data/myid

将当前主机在zookeeper集群中的id值写入，然后启动zookeeper

3.3.3 启动zookeeper

[root@10-124-147-23 bin]# pwd
/usr/local/zookeeper/bin
[root@10-124-147-23 bin]# ./zkServer.sh start

同理，启动其它主机的zookeeper，操作同上，唯一区别的就是/usr/local/zookeeper/data/myid中的值，需要彼此不一样

3.3.4 查看zookeeper状态

[root@10-124-147-23 bin]# ./zkServer.sh status
ZooKeeper JMX enabled by default
Using config: /usr/local/zookeeper/bin/../conf/zoo.cfg
Mode: follower

[root@10-124-147-33 ~]# /usr/local/zookeeper/bin/zkServer.sh status
ZooKeeper JMX enabled by default
Using config: /usr/local/zookeeper/bin/../conf/zoo.cfg
Mode: leader

4.hadoop的安装

hadoop2.0官方提供了两种HDFS HA的解决方案，一种是NFS，另一种是QJM。这里我们使用简单的QJM。在该方案中，主备NameNode之间通过一组JournalNode同步元数据信息，一条数据只要成功写入多数JournalNode即认为写入成功。JournalNode的个数需要为奇数个

4.1 hadoop解压

[root@10-124-147-33 letv]# tar xvf hadoop-2.7.6.tar.gz 
[root@10-124-147-23 ~]# ln -svnf /letv/hadoop-2.7.6 /usr/local/hadoop

4.2 hadoop环境

本次安装hadoop，只需要指定java环境和hadoop环境即可，因为zookeeper和hadoop都需要运行java环境，上述安装环境已经指定

[root@10-124-147-23 letv]# tail -3 /etc/profile
export JAVA_HOME=/usr/local/java
export HADOOP_HOME=/usr/local/hadoop
export PATH=$HADOOP_HOME/bin:$JAVA_HOME/bin:$PATH

4.3 hadoop配置文件的修改

hadoop配置文件位于etc/hadoop目录之下，主要控制文件有以下6个

4.3.1 hadoop-env.sh

[root@10-124-147-23 ~]# grep JAVA_HOME /usr/local/hadoop/etc/hadoop/hadoop-env.sh
# The only required environment variable is JAVA_HOME.  All others are
# set JAVA_HOME in this file, so that it is correctly defined on
export JAVA_HOME=/usr/local/java

此处需要指向java环境的实际路径，不能直接使用${JAVA_HOME}来指定，此处并不能直接识别此变量，具体原因未知。

4.3.2 hdfs-site.xml

[root@10-124-147-23 ~]# cat /usr/local/hadoop/etc/hadoop/hdfs-site.xml
<?xml version="1.0" encoding="UTF-8"?>
<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
<!--
  Licensed under the Apache License, Version 2.0 (the "License");
  you may not use this file except in compliance with the License.
  You may obtain a copy of the License at
    http://www.apache.org/licenses/LICENSE-2.0
  Unless required by applicable law or agreed to in writing, software
  distributed under the License is distributed on an "AS IS" BASIS,
  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  See the License for the specific language governing permissions and
  limitations under the License. See accompanying LICENSE file.
-->
<!-- Put site-specific property overrides in this file. -->
<configuration>
<!--指定hdfs的nameservice为ns1，需要和core-site.xml中的保持一致 -->
<property>
<name>dfs.nameservices</name>
<value>ns1</value>
</property>

<!-- ns1下面有两个NameNode，分别是nn1，nn2 -->
<property>
<name>dfs.ha.namenodes.ns1</name>
<value>nn1,nn2</value>
</property>

<!-- nn1的RPC通信地址 -->
<property>
<name>dfs.namenode.rpc-address.ns1.nn1</name>
<value>hadoop1:9000</value>
</property>

<!-- nn1的http通信地址 -->
<property>
<name>dfs.namenode.http-address.ns1.nn1</name>
<value>hadoop1:50070</value>
</property>

<!-- nn2的RPC通信地址 -->
<property>
<name>dfs.namenode.rpc-address.ns1.nn2</name>
<value>hadoop2:9000</value>
</property>

<!-- nn2的http通信地址 -->
<property>
<name>dfs.namenode.http-address.ns1.nn2</name>
<value>hadoop2:50070</value>
</property>

<!-- 指定NameNode的元数据在JournalNode上的存放位置 -->
<property>
<name>dfs.namenode.shared.edits.dir</name>
<value>qjournal://hadoop5:8485;hadoop6:8485;hadoop7:8485/ns1</value>
</property>

<!-- 指定JournalNode在本地磁盘存放数据的位置 -->
<property>
<name>dfs.journalnode.edits.dir</name>
<value>/usr/local/hadoop/data/journaldata</value>
</property>

<!-- 开启NameNode失败自动切换 -->
<property>
<name>dfs.ha.automatic-failover.enabled</name>
<value>true</value>
</property>

<!-- 配置失败自动切换实现方式 -->
<property>
<name>dfs.client.failover.proxy.provider.ns1</name>
<value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
</property>

<!-- 配置隔离机制方法，多个机制用换行分割，即每个机制占用一行-->
<property>
<name>dfs.ha.fencing.methods</name>
<value>
sshfence
shell(/bin/true)
</value>
</property>

<!-- 使用sshfence隔离机制时需要ssh免登陆 -->
<property>
<name>dfs.ha.fencing.ssh.private-key-files</name>
<value>/home/hadoop/.ssh/id_rsa</value>
</property>

<!-- 配置sshfence隔离机制超时时间 -->
<property>
<name>dfs.ha.fencing.ssh.connect-timeout</name>
<value>30000</value>
</property>
</configuration>

在hadoop 3中，hdfs的web通讯端口50070 已经变更为9870

4.3.3 mapred-site.xml

[root@10-124-147-23 ~]# cat /usr/local/hadoop/etc/hadoop/mapred-site.xml
<?xml version="1.0"?>
<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
<!--
  Licensed under the Apache License, Version 2.0 (the "License");
  you may not use this file except in compliance with the License.
  You may obtain a copy of the License at
    http://www.apache.org/licenses/LICENSE-2.0
  Unless required by applicable law or agreed to in writing, software
  distributed under the License is distributed on an "AS IS" BASIS,
  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  See the License for the specific language governing permissions and
  limitations under the License. See accompanying LICENSE file.
-->
<!-- Put site-specific property overrides in this file. -->

<configuration>
<!-- 指定mr框架为yarn方式 -->
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property>

<property>
<name>mapreduce.jobhistory.address</name>
<value>hadoop1:10020</value>
</property>

<property>
<name>mapreduce.jobhistory.webapp.address</name>
<value>hadoop1:19888</value>
</property>
</configuration>

4.3.4 core-site.xml

[root@10-124-147-23 ~]# cat /usr/local/hadoop/etc/hadoop/core-site.xml
<?xml version="1.0" encoding="UTF-8"?>
<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
<!--
  Licensed under the Apache License, Version 2.0 (the "License");
  you may not use this file except in compliance with the License.
  You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

  Unless required by applicable law or agreed to in writing, software
  distributed under the License is distributed on an "AS IS" BASIS,
  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  See the License for the specific language governing permissions and
  limitations under the License. See accompanying LICENSE file.
-->

<!-- Put site-specific property overrides in this file. -->

<configuration>
<!-- 指定hdfs的nameservice为ns1 -->
<property>
<name>fs.defaultFS</name>
<value>hdfs://ns1</value>
</property>
<!-- 指定hadoop临时目录 -->
<property>
<name>hadoop.tmp.dir</name>
<value>/usr/local/hadoop/data/tmp</value>
</property>
<!-- 指定zookeeper地址 -->
<property>
<name>ha.zookeeper.quorum</name>
<value>hadoop1:2181,hadoop2:2181,hadoop3:2181,hadoop4:2181</value>
</property>
</configuration>

4.3.5 yarn-site.xml

[root@10-124-147-23 ~]# cat /usr/local/hadoop/etc/hadoop/yarn-site.xml
<?xml version="1.0"?>
<!--
  Licensed under the Apache License, Version 2.0 (the "License");
  you may not use this file except in compliance with the License.
  You may obtain a copy of the License at
    http://www.apache.org/licenses/LICENSE-2.0
  Unless required by applicable law or agreed to in writing, software
  distributed under the License is distributed on an "AS IS" BASIS,
  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  See the License for the specific language governing permissions and
  limitations under the License. See accompanying LICENSE file.
-->


<configuration>

<!-- Site specific YARN configuration properties -->
<!-- 开启RM高可靠 -->
<property>
<name>yarn.resourcemanager.ha.enabled</name>
<value>true</value>
</property>

<!-- 指定RM的cluster id -->
<property>
<name>yarn.resourcemanager.cluster-id</name>
<value>yrc</value>
</property>

<!-- 指定RM的名字 -->
<property>
<name>yarn.resourcemanager.ha.rm-ids</name>
<value>rm1,rm2</value>
</property>

<!-- 分别指定RM的地址 -->
<property>
<name>yarn.resourcemanager.hostname.rm1</name>
<value>hadoop3</value>
</property>

<property>
<name>yarn.resourcemanager.hostname.rm2</name>
<value>hadoop4</value>
</property>

<!-- 指定zk集群地址 -->
<property>
<name>yarn.resourcemanager.zk-address</name>
<value>hadoop1:2181,hadoop2:2181,hadoop3:2181,hadoop4:2181</value>
</property>

<!-- 在RM节点接管后，任务状态可以恢复-->
<property>
<name>yarn.resourcemanager.recovery.enabled</name>
<value>true</value>
</property>

<!-- 设置存储yarn中状态信息的地方，默认为hdfs，这里设置为zookeeper-->
<property>
<name>yarn.resourcemanager.store.class</name>
<value>org.apache.hadoop.yarn.server.resourcemanager.recovery.ZKRMStateStore</value>
</property>

<!-- 使在yarn上能够运行mapreduce_shuffle程序-->
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
</configuration>

4.3.6 slave

[root@10-124-147-23 ~]# cat /usr/local/hadoop/etc/hadoop/slaves 
hadoop5
hadoop6
hadoop7

这里的slaves分两种，对于hadoop1而言，其为namenode，所以其slaves是hdfs系统中的slaves，也就是datanode，在本文中，设定hadoop5,hadoop6,hadoop7为datanode

而对于hadoop3而言，其为resourcemanager，故其slaves是yarn系统中的slaves，也就是nodemanager，nodemanager是对每机机器的资源状态进行监控，同时将监控结果向resourcemanager进行报告，一般一台datanode上面都会有着nodemanager进程。

本文中journalnode,nodemanager,datanode三个角色都是位于同一机器，实际上journalnode只是参与到namenodeHA模式中，与后两者并不挂钩，因为集群中不允许同时有两个namenode同时工作 ，否则数据地址空间就会出错，但是为了HA，所以standby的namenode需要保持与active状态的namenode数据一致，两个namenode为了数据同步，会通过一组称作journalnodes的独立进程进行相互通信。当active状态的namenode的命名空间有任何修改时，会告知大部分的journalnodes进程。standby状态的namenode有能力读取journalnodes中的变更信息，并且一直监控edit log的变化，把变化应用于自己的命名空间。standby可以确保在集群出错时，命名空间状态已经完全同步了。

一般正常生产中，journalnode设定为5个，基本上zookeeper个数也是设定为5个，文中我zookeeper设定4个其实不太合理。

综上，所以对于hadoop3而言，其slaves也可以设定为hadoop5,hadoop6,hadoop7

所以本文中所有节点，hadoop配置可以保持一致

4.3.7 ssh-key验证

实际生产中其实只需要namenode之间ssh-key免密即可，实验环境中，因为需要在namenode中直接通过脚本启动其它slaves节点，所以需要进行ssh-key免密的设定

主要的设定的是datanode中需要有两个namenode和两个resourcemanager的ssh-key信息，同时namenode和resourcemanger自身也需要自身的ssh-key，以便启动，所以文中hadoop1,hadoop2,hadoop3,hadoop44台主机的hadoop用户的ssh-key需要放置于每一台主机hadoop用户之下。

[root@10-124-147-23 ~]# useradd hadoop
[hadoop@10-124-147-23 ~]$ ssh-keygen
[hadoop@10-124-147-23 ~]$ cat .ssh/id_rsa.pub 
ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAQEAyQ9T7zTAlhqFM9XQoHTPzwfgDwAzwLUgqe7NnDpufiirK9QqCdLZFNE6PNtN7oNyWMu3r9UE5aMYv9uLMu22m+8xyTXXINYfPW9hsityu/N6a9DwhEC9joNS3DVjBR8YRMQG2sxtDbebbaG2R4BK77DZyoB0uyqRItxLIMYTiZ/00LCMJCoAINUQVzOrteVpLHAviRNnrwZewoD2sUgeZU0A0hT++RiE/prqI+jIFJSacduVaKsabRu/zKan9b8coC1b+GJnypqk+CPyahJL+0jgb9Jgrjm2Lt4erbBo/k3u16nSJpSoSdf7kr5HKv3ds5+fwcMQV5oKV1jv6ximIw== hadoop@10-124-147-23

然后切换至其它节点主要，依次创建hadoop用户，将namenode节点ssh-key写入

[root@10-124-147-33 letv]# useradd hadoop
[hadoop@10-124-147-33 ~]$ mkdir .ssh
[hadoop@10-124-147-33 ~]$ chmod g-w .ssh
以上这一步非常重要，因为正常情况下需要对hadoop用户进行密码设定之后，然后再使用ssh-copy-id将key自动写入到其它主机中，我们并没有对hadoop用户设定密码，而ssh中为了安全，g与o用户是对.ssh目录均无w权限的，所以需要将.ssh目录中g与o用户的w权限去掉。类似的还在后面中的authorized_keys文件
[hadoop@10-124-147-33 ~]$ vim .ssh/authorized_keys
将hadoop1中的id_rsa.pub写入
[hadoop@10-124-147-33 ~]$ chmod 600 .ssh/authorized_keys
[hadoop@10-124-147-33 ~]$ ll .ssh/authorized_keys
-rw------- 1 hadoop hadoop 1608 Jul 19 11:43 .ssh/authorized_keys
[hadoop@10-124-147-33 ~]$ ll -d .ssh/
drwxr-xr-x 2 hadoop hadoop 4096 Jul 19 11:43 .ssh/

4.3.8 hadoop 文件copy

将hadoop1中的hadoop目录整个scp至其它节点，同时注意/etc/profile文件，以及部分节点上面的java环境

4.4 hadoop的启动

4.4.1 启动journalnode

[hadoop@10-110-92-161 ~]$ cd /usr/local/hadoop/
[hadoop@10-110-92-161 hadoop]$ sbin/hadoop-daemon.sh start journalnode
[hadoop@10-110-92-161 hadoop]$ jps
1557 JournalNode
22439 Jps

三个节点的journalnode都要启动

4.4.2 格式化namenode

[hadoop@10-124-147-22 hadoop]$  hdfs namenode -format

4.4.3 启动active namenode

[hadoop@10-124-147-22 hadoop]$ sbin/hadoop-daemon.sh start namenode
[hadoop@10-124-147-22 hadoop]$ jps
2580 DFSZKFailoverController
29590 Jps
1487 NameNode

4.4.4 复制active namenode信息至standby namenode

格式化active namenode后会在根据core-site.xml中的hadoop.tmp.dir配置生成个文件，可能直接copy至standby namenode，也可以通过选项-bootstrapStandby直接从active namenode拉取，使用命令拉取的前提是active namenode进程需要启动

[hadoop@10-124-147-23 hadoop]$ hdfs namenode -bootstrapStandby
[hadoop@10-124-147-23 hadoop]$ sbin/hadoop-daemon.sh start namenode
[hadoop@10-124-147-23 hadoop]$ jps
899 NameNode
11846 Jps
1353 DFSZKFailoverController

4.4.5 格式化zkfc

[hadoop@10-124-147-22 hadoop]$ hdfs zkfc -formatZK

4.4.6 启动hdfs

[hadoop@10-124-147-22 hadoop]$ sbin/start-dfs.sh

4.4.7 启动resourcemanager

[hadoop@10-124-147-32 hadoop]$ pwd
/usr/local/hadoop
[hadoop@10-124-147-32 hadoop]$ resourcemanager sbin/start-yarn.sh
[hadoop@10-124-147-32 hadoop]$ jps
30882 ResourceManager
26868 Jps

4.4.8 启动standby resourcemanager

[hadoop@10-124-147-33 hadoop]$ pwd
/usr/local/hadoop
[hadoop@10-124-147-33 hadoop]$ sbin/yarn-daemon.sh start resourcemanager
[hadoop@10-124-147-33 hadoop]$ jps
22675 Jps
26980 ResourceManager

4.4.9 集群状态检测

[hadoop@10-124-147-22 hadoop]$ hdfs haadmin -getServiceState nn1
active
[hadoop@10-124-147-22 hadoop]$ hdfs haadmin -getServiceState nn2
standby
[hadoop@10-124-147-22 hadoop]$ yarn rmadmin -getServiceState rm1
active
[hadoop@10-124-147-22 hadoop]$ yarn rmadmin -getServiceState rm2
standby

此时，可以通过web访问active namenode的50070端口和active resourcemanager的8080端口

4.4.10 启动history进程

在active namenode启动即可

[hadoop@10-124-147-22 hadoop]$ sbin/mr-jobhistory-daemon.sh start historyserver
[hadoop@10-124-147-22 hadoop]$ pwd
/usr/local/hadoop
[hadoop@10-124-147-22 hadoop]$ jps
2580 DFSZKFailoverController
31781 Jps
2711 JobHistoryServer
1487 NameNode

4.5 hadoop的简单使用

4.5.1 上传文件于hdfs

新建一个文件/tmp/test.txt
[hadoop@10-124-147-22 hadoop]$ cat /tmp/test.txt 
hello world
hello mysql
hello mongo
hello elasticsearch
hello hadoop
hello hdfs
hello yarn
hello namenode
hello datanode
hello resourcemanager
hello nodemanager
hello journalnode
[hadoop@10-124-147-22 hadoop]$ hadoop fs -put /tmp/test.txt /wordcount
将/tmp/test.txt文件上传于hdfs中，并重命名为wordcount
[hadoop@10-124-147-22 hadoop]$ hadoop fs -cat /wordcount
hello world
hello mysql
hello mongo
hello elasticsearch
hello hadoop
hello hdfs
hello yarn
hello namenode
hello datanode
hello resourcemanager
hello nodemanager
hello journalnode

4.5.2 hadoop任务测试

hadoop中提供了简单的任务测试jar包，可以进行测试

[hadoop@10-124-147-22 hadoop]$ hadoop jar share/hadoop/mapreduce/hadoop-mapreduce-examples-2.7.6.jar pi 2 10
Number of Maps  = 2
Samples per Map = 10
Wrote input for Map #0
Wrote input for Map #1
Starting Job
18/07/23 15:41:47 INFO input.FileInputFormat: Total input paths to process : 2
18/07/23 15:41:47 INFO mapreduce.JobSubmitter: number of splits:2
18/07/23 15:41:47 INFO mapreduce.JobSubmitter: Submitting tokens for job: job_1532056892547_0003
18/07/23 15:41:47 INFO impl.YarnClientImpl: Submitted application application_1532056892547_0003
18/07/23 15:41:47 INFO mapreduce.Job: The url to track the job: http://hadoop3:8088/proxy/application_1532056892547_0003/
18/07/23 15:41:47 INFO mapreduce.Job: Running job: job_1532056892547_0003
18/07/23 15:41:53 INFO mapreduce.Job: Job job_1532056892547_0003 running in uber mode : false
18/07/23 15:41:53 INFO mapreduce.Job:  map 0% reduce 0%
18/07/23 15:41:58 INFO mapreduce.Job:  map 100% reduce 0%
18/07/23 15:42:03 INFO mapreduce.Job:  map 100% reduce 100%
18/07/23 15:42:04 INFO mapreduce.Job: Job job_1532056892547_0003 completed successfully
18/07/23 15:42:05 INFO mapreduce.Job: Counters: 49
	File System Counters
		FILE: Number of bytes read=50
		FILE: Number of bytes written=376437
		FILE: Number of read operations=0
		FILE: Number of large read operations=0
		FILE: Number of write operations=0
		HDFS: Number of bytes read=510
		HDFS: Number of bytes written=215
		HDFS: Number of read operations=11
		HDFS: Number of large read operations=0
		HDFS: Number of write operations=3
	Job Counters 
		Launched map tasks=2
		Launched reduce tasks=1
		Data-local map tasks=2
		Total time spent by all maps in occupied slots (ms)=5283
		Total time spent by all reduces in occupied slots (ms)=2804
		Total time spent by all map tasks (ms)=5283
		Total time spent by all reduce tasks (ms)=2804
		Total vcore-milliseconds taken by all map tasks=5283
		Total vcore-milliseconds taken by all reduce tasks=2804
		Total megabyte-milliseconds taken by all map tasks=5409792
		Total megabyte-milliseconds taken by all reduce tasks=2871296
	Map-Reduce Framework
		Map input records=2
		Map output records=4
		Map output bytes=36
		Map output materialized bytes=56
		Input split bytes=274
		Combine input records=0
		Combine output records=0
		Reduce input groups=2
		Reduce shuffle bytes=56
		Reduce input records=4
		Reduce output records=0
		Spilled Records=8
		Shuffled Maps =2
		Failed Shuffles=0
		Merged Map outputs=2
		GC time elapsed (ms)=219
		CPU time spent (ms)=3030
		Physical memory (bytes) snapshot=752537600
		Virtual memory (bytes) snapshot=6612717568
		Total committed heap usage (bytes)=552075264
	Shuffle Errors
		BAD_ID=0
		CONNECTION=0
		IO_ERROR=0
		WRONG_LENGTH=0
		WRONG_MAP=0
		WRONG_REDUCE=0
	File Input Format Counters 
		Bytes Read=236
	File Output Format Counters 
		Bytes Written=97
Job Finished in 18.492 seconds
Estimated value of Pi is 3.80000000000000000000

在job执行的时候，可以查看resourcemangerweb端的8088端口，上面可以看到job的完成进度

再执行一个word count任务

可以执行字母统计，将hdfs中的wordcount文件统计，并将结果输出到wordcount-to-output
[hadoop@10-124-147-22 hadoop]$ hadoop jar share/hadoop/mapreduce/hadoop-mapreduce-examples-2.7.6.jar wordcount /wordcount /wordcount-to-output
18/07/23 15:45:12 INFO input.FileInputFormat: Total input paths to process : 1
18/07/23 15:45:13 INFO mapreduce.JobSubmitter: number of splits:1
18/07/23 15:45:13 INFO mapreduce.JobSubmitter: Submitting tokens for job: job_1532056892547_0004
18/07/23 15:45:13 INFO impl.YarnClientImpl: Submitted application application_1532056892547_0004
18/07/23 15:45:13 INFO mapreduce.Job: The url to track the job: http://hadoop3:8088/proxy/application_1532056892547_0004/
18/07/23 15:45:13 INFO mapreduce.Job: Running job: job_1532056892547_0004
18/07/23 15:45:19 INFO mapreduce.Job: Job job_1532056892547_0004 running in uber mode : false
18/07/23 15:45:19 INFO mapreduce.Job:  map 0% reduce 0%
18/07/23 15:45:23 INFO mapreduce.Job:  map 100% reduce 0%
18/07/23 15:45:29 INFO mapreduce.Job:  map 100% reduce 100%
18/07/23 15:45:29 INFO mapreduce.Job: Job job_1532056892547_0004 completed successfully
18/07/23 15:45:29 INFO mapreduce.Job: Counters: 49
	File System Counters
		FILE: Number of bytes read=197
		FILE: Number of bytes written=250631
		FILE: Number of read operations=0
		FILE: Number of large read operations=0
		FILE: Number of write operations=0
		HDFS: Number of bytes read=264
		HDFS: Number of bytes written=140
		HDFS: Number of read operations=6
		HDFS: Number of large read operations=0
		HDFS: Number of write operations=2
	Job Counters 
		Launched map tasks=1
		Launched reduce tasks=1
		Data-local map tasks=1
		Total time spent by all maps in occupied slots (ms)=2492
		Total time spent by all reduces in occupied slots (ms)=3007
		Total time spent by all map tasks (ms)=2492
		Total time spent by all reduce tasks (ms)=3007
		Total vcore-milliseconds taken by all map tasks=2492
		Total vcore-milliseconds taken by all reduce tasks=3007
		Total megabyte-milliseconds taken by all map tasks=2551808
		Total megabyte-milliseconds taken by all reduce tasks=3079168
	Map-Reduce Framework
		Map input records=12
		Map output records=24
		Map output bytes=275
		Map output materialized bytes=197
		Input split bytes=85
		Combine input records=24
		Combine output records=13
		Reduce input groups=13
		Reduce shuffle bytes=197
		Reduce input records=13
		Reduce output records=13
		Spilled Records=26
		Shuffled Maps =1
		Failed Shuffles=0
		Merged Map outputs=1
		GC time elapsed (ms)=155
		CPU time spent (ms)=2440
		Physical memory (bytes) snapshot=465940480
		Virtual memory (bytes) snapshot=4427837440
		Total committed heap usage (bytes)=350224384
	Shuffle Errors
		BAD_ID=0
		CONNECTION=0
		IO_ERROR=0
		WRONG_LENGTH=0
		WRONG_MAP=0
		WRONG_REDUCE=0
	File Input Format Counters 
		Bytes Read=179
	File Output Format Counters 
		Bytes Written=140

执行结果

[hadoop@10-124-147-22 hadoop]$ hadoop fs -ls /
Found 5 items
drwxrwx---   - hadoop supergroup          0 2018-07-20 11:21 /tmp
drwxr-xr-x   - hadoop supergroup          0 2018-07-20 11:47 /user
-rw-r--r--   3 hadoop supergroup        179 2018-07-20 11:22 /wordcount
drwxr-xr-x   - hadoop supergroup          0 2018-07-23 15:45 /wordcount-to-output
[hadoop@10-124-147-22 hadoop]$ hadoop fs -ls /wordcount-to-output
Found 2 items
-rw-r--r--   3 hadoop supergroup          0 2018-07-23 15:45 /wordcount-to-output/_SUCCESS
-rw-r--r--   3 hadoop supergroup        140 2018-07-23 15:45 /wordcount-to-output/part-r-00000
[hadoop@10-124-147-22 hadoop]$ hadoop fs -cat /wordcount-to-output/part-r-00000
datanode	1
elasticsearch	1
hadoop	1
hdfs	1
hello	12
journalnode	1
mongo	1
mysql	1
namenode	1
nodemanager	1
resourcemanager	1
world	1
yarn	1

5.其它

5.1 hadoop3相对比hadoop2进程端口更变

Namenode ports: 50470 --> 9871, 50070 --> 9870, 8020 --> 9820
Secondary NN ports: 50091 --> 9869, 50090 --> 9868
Datanode ports: 50020 --> 9867, 50010 --> 9866, 50475 --> 9865, 50075 --> 9864
KMS service ：16000 --> 9600

同时变更的还有slaves文件，在hadoop2中的slaves文件在hadoop3中变成works文件

5.2生产中datanode的启动

生产中hadoop集群里面的datanode一般都是几百上千台主机，实际上生产中的datanode都是在各自主机中自行单独启动，并不是直接通过namenode进行启动，所以上面4.3.7中的ssh-key在实际生产中并不无那么多需求。同时journalnode虽然消耗资源小，但是一般也不与datanode分布于同一台主机中。