一、Hadoop原理介绍
1、请参考原理篇：Hadoop1-认识Hadoop大数据处理架构
二、centos7单机部署hadoop 前期准备
1、创建用户
[root@web3 ~]# useradd -m hadoop -s /bin/bash #---创建hadoop用户
[root@web3 ~]# passwd hadoop #---创建密码
Changing password for user hadoop.
New password:
BAD PASSWORD: The password is a palindrome
Retype new password:
passwd: all authentication tokens updated successfully.
[root@web3 ~]#
2、添加用户权限
[root@web3 ~]# chmod u+w /etc/sudoers #---给sudo文件写权限
[root@web3 ~]# cat /etc/sudoers |grep hadoop #---这里自行vim添加，这里是用cat命令展示为添加后的效果
hadoop ALL=(ALL) ALL
[root@web3 ~]#
[root@web3 ~]# chmod u-w /etc/sudoers #---给sudo去掉写权限
3、安装软件openssh，生成授权，免密码
1）安装
[root@web3 ~]# su hadoop #切换用户hadoop
[hadoop@web3 root]$ sudo yum install openssh-clients openssh-server #---安装openssh
1）操作步骤
cd .ssh/
ssh-keygen -t rsa
cat id_rsa.pub >> authorized_keys
chmod ./authorized_keys
4、安装java
sudo yum install java-1.8.-openjdk java-1.8.-openjdk-devel
#---用rpm -ql查看java相关目录
[hadoop@web3 bin]$ rpm -ql java-1.8.0-openjdk.x86_64 1:1.8.0.222.b10-1.el7_7
/usr/lib/jvm/java-1.8.0-openjdk-1.8.0.222.b10-1.el7_7.x86_64/jre/bin/policytool
/usr/lib/jvm/java-1.8.-openjdk-1.8.0.222.b10-.el7_7.x86_64/jre/lib/amd64/libawt_xawt.so
/usr/lib/jvm/java-1.8.-openjdk-1.8.0.222.b10-.el7_7.x86_64/jre/lib/amd64/libjawt.so
/usr/lib/jvm/java-1.8.-openjdk-1.8.0.222.b10-.el7_7.x86_64/jre/lib/amd64/libjsoundalsa.so
/usr/lib/jvm/java-1.8.-openjdk-1.8.0.222.b10-.el7_7.x86_64/jre/lib/amd64/libsplashscreen.so
/usr/share/applications/java-1.8.-openjdk-1.8.0.222.b10-.el7_7.x86_64-policytool.desktop
/usr/share/icons/hicolor/16x16/apps/java-1.8.-openjdk.png
/usr/share/icons/hicolor/24x24/apps/java-1.8.-openjdk.png
/usr/share/icons/hicolor/32x32/apps/java-1.8.-openjdk.png
/usr/share/icons/hicolor/48x48/apps/java-1.8.-openjdk.png
package :1.8.0.222.b10-.el7_7 is not installed
[hadoop@web3 bin]$
5、添加环境变量
[hadoop@web3 bin]$ cat ~/.bashrc
# .bashrc
# Source global definitions
if [ -f /etc/bashrc ]; then
. /etc/bashrc
fi
# Uncomment the following line if you don't like systemctl's auto-paging feature:
# export SYSTEMD_PAGER=
# User specific aliases and functions
#---添加此环境变量
export JAVA_HOME=/usr/lib/jvm/java-1.8.0-openjdk-1.8.0.222.b10-1.el7_7.x86_64
#---输出检查
[hadoop@web3 jvm]$ echo $JAVA_HOME
/usr/lib/jvm/java-1.8.0-openjdk-1.8.0.222.b10-1.el7_7.x86_64
#---输出jave版本
[hadoop@web3 jvm]$ java -version
openjdk version "1.8.0_222"
OpenJDK Runtime Environment (build 1.8.0_222-b10)
OpenJDK 64-Bit Server VM (build 25.222-b10, mixed mode)
#---使用变量输出java版本
[hadoop@web3 jvm]$ $JAVA_HOME/bin/java -version
openjdk version "1.8.0_222"
OpenJDK Runtime Environment (build 1.8.0_222-b10)
OpenJDK 64-Bit Server VM (build 25.222-b10, mixed mode)
[hadoop@web3 jvm]$
java -version 与$JAVA_HOME/bin/jave -version运行一直即代表添加成功
6、开始安装hadoop 3.1.2
下载路径：http://mirror.bit.edu.cn/apache/hadoop/common/
上传到服务器
[hadoop@web3 root]$ cd
[hadoop@web3 ~]$ ll
total
-rw-r--r-- hadoop hadoop Oct : hadoop-3.1..tar.gz
drwxrwxr-x hadoop hadoop Oct : ssh
[hadoop@web3 ~]$ sudo tar -zxf hadoop-3.1.2.tar.gz -C /usr/local
[sudo] password for hadoop:
[hadoop@web3 ~]$ cd /usr/local
[hadoop@web3 local]$ sudo mv hadoop-3.1.2/ ./hadoop
[hadoop@web3 local]$ ll
total
drwxr-xr-x. root root Nov bin
drwxr-xr-x. root root Nov etc
drwxr-xr-x. root root Nov games
drwxr-xr-x hadoop Jan hadoop
drwxr-xr-x. root root Nov include
drwxr-xr-x. root root Nov lib
drwxr-xr-x. root root Nov lib64
drwxr-xr-x. root root Nov libexec
drwxr-xr-x. root root Nov sbin
drwxr-xr-x. root root Aug share
drwxr-xr-x. root root Nov src
[hadoop@web3 local]$ chown -R hadoop:hadoop ./hadoop
[hadoop@web3 local]$ cd hadoop/
[hadoop@web3 hadoop]$ ./bin/hadoop version
Hadoop 3.1.
Source code repository https://github.com/apache/hadoop.git -r 1019dde65bcf12e05ef48ac71e84550d589e5d9a
Compiled by sunilg on --29T01:39Z
Compiled with protoc 2.5.
From source with checksum 64b8bdd4ca6e77cce75a93eb09ab2a9
This command was run using /usr/local/hadoop/share/hadoop/common/hadoop-common-3.1..jar
[hadoop@web3 hadoop]$ pwd
/usr/local/hadoop
[hadoop@web3 hadoop]$
三、Hadoop单机配置-非分布式
hadoop默认模式为非分布式模式，无需进行其他配置即可运行，非分布式即但java进程，方便进行调试
hadoop附带了丰富的例子（./bin/hadoop jar ./share/hadoop/mapreduce/hadoop-mapreduce-examples-3.1.2.jar 可以看到所有例子），包括wordcount、terasort、join、grep等
1、现在运行grep测试一下
这个实例是运行grep例子，将input文件夹所有文件作为输入，筛选当中符合正则表达式dfs[a-z.]+的单词并统计出现的次数，最后输出结果到output文件夹中
mkdir ./input
cp ./etc/hadoop/*.xml ./input #---将配置文件作为输入文件
./bin/hadoop jar ./share/hadoop/mapreduce/hadoop-mapreduce-examples-3.1.2.jar grep ./input ./output 'dfs[a-z.]+'
2、正确的运行结果
运行hadoop实例，成功的话会输出很多作业的相关信息，最后的输出信息就是下面图示，作业结果会输出在指定的output文件夹中，通过命令cat ./output/* 查看结果，符合正则的单词dfsadmin出现了一次。
[hadoop@web3 hadoop]$ cat ./output/* #如果要重新运行
1 dfsadmin
[hadoop@web3 hadoop]$
四、Hadoop伪分布式配置
hadoop可以在单台节点以伪分布式运行，hadoop进程以分离的java进程来运行，节点作为namenode也作为datanode，同时，读取的时HDFS中的文件
1、设置环境变量
[hadoop@web3 hadoop]$ vim ~/.bashrc # .bashrc # Source global definitions
if [ -f /etc/bashrc ]; then
. /etc/bashrc
fi # Uncomment the following line if you don't like systemctl's auto-paging feature:
# export SYSTEMD_PAGER= # User specific aliases and functions
#Java environment variables
export JAVA_HOME=/usr/lib/jvm/java-1.8.-openjdk-1.8.0.222.b10-.el7_7.x86_64
#Hadoop environment Variables
export HADOOP_HOME=/usr/local/hadoop
export HADOOP_INSTALL=$HADOOP_HOME
export HADOOP_MAPRED_HOME=$HADOOP_HOME
export HADOOP_COMMON_HOME=$HADOOP_HOME
export HADOOP_HDFS_HOME=$HADOOP_HOME
export YARN_HOME=$HADOOP_HOME
export HADOOP_COMMON_LIB_NATIVE_DIR=$HADOOP_HOME/lib/native
export PATH=$PATH:$HADOOP_HOME/sbin:$HADOOP_HOME/bin
~
更新环境变量
2、修改被指文件
配置文件位于/usr/local/hadoop/etc/hadoop/中，伪分布式需要两个配置文件core-site.xml和hdfs-site.xml，hadoop的配置文件时xml格式，每个配置声明property的name和value的方式来实现
core-site.xml
修改标注红色字体部分
[hadoop@web3 hadoop]$ vim ./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>
<property>
<name>hadoop.tmp.dir</name>
<value>file:/usr/local/hadoop/tmp</value>
<description>Abase for other temporary directories.</description>
</property>
<property>
<name>fs.defaultFS</name>
<value>hdfs://localhost:9000</value>
</property>
</configuration>
~
~
hdfs-site.xml
[hadoop@web3 hadoop]$ vim ./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>
<property>
<name>dfs.replication</name>
<value>1</value>
</property>
<property>
<name>dfs.namenode.name.dir</name>
<value>file:/usr/local/hadoop/tmp/dfs/name</value>
</property>
<property>
<name>dfs.datanode.data.dir</name>
<value>file:/usr/local/hadoop/tmp/dfs/data</value>
</property>
</configuration>
3、执行namenode的格式化
[hadoop@web3 hadoop]$ ./bin/hdfs namenode -format
WARNING: /usr/local/hadoop/logs does not exist. Creating.
-- ::, INFO namenode.NameNode: STARTUP_MSG:
/************************************************************
STARTUP_MSG: Starting NameNode
STARTUP_MSG: host = web3/192.168.216.53
STARTUP_MSG: args = [-format]
STARTUP_MSG: version = 3.1.2
STARTUP_MSG: classpath = /usr/local/hadoop/etc/hadoop:/usr/local/hadoop/share/hadoop/common/lib/accessors-smart-1.2.jar:/usr/local/hadoop/share/hadoop/common/lib/asm-
。。。。。。。。。。。。。。。。。这里省略一堆。。。。。。。。。。。。。。。。。。。。。。。。。
2019-10-18 18:56:47,031 INFO namenode.FSDirectory: XAttrs enabled? true
2019-10-18 18:56:47,032 INFO namenode.NameNode: Caching file names occurring more than 10 times
2019-10-18 18:56:47,046 INFO snapshot.SnapshotManager: Loaded config captureOpenFiles: false, skipCaptureAccessTimeOnlyChange: false, snapshotDiffAllowSnapRootDescendant: true, maxSnapshotLimit: 65536
2019-10-18 18:56:47,049 INFO snapshot.SnapshotManager: SkipList is disabled
2019-10-18 18:56:47,057 INFO util.GSet: Computing capacity for map cachedBlocks
2019-10-18 18:56:47,057 INFO util.GSet: VM type = 64-bit
2019-10-18 18:56:47,057 INFO util.GSet: 0.25% max memory 411 MB = 1.0 MB
2019-10-18 18:56:47,058 INFO util.GSet: capacity = 2^17 = 131072 entries
2019-10-18 18:56:47,083 INFO metrics.TopMetrics: NNTop conf: dfs.namenode.top.window.num.buckets = 10
2019-10-18 18:56:47,084 INFO metrics.TopMetrics: NNTop conf: dfs.namenode.top.num.users = 10
2019-10-18 18:56:47,084 INFO metrics.TopMetrics: NNTop conf: dfs.namenode.top.windows.minutes = 1,5,25
2019-10-18 18:56:47,090 INFO namenode.FSNamesystem: Retry cache on namenode is enabled
2019-10-18 18:56:47,090 INFO namenode.FSNamesystem: Retry cache will use 0.03 of total heap and retry cache entry expiry time is 600000 millis
2019-10-18 18:56:47,094 INFO util.GSet: Computing capacity for map NameNodeRetryCache
2019-10-18 18:56:47,094 INFO util.GSet: VM type = 64-bit
2019-10-18 18:56:47,094 INFO util.GSet: 0.029999999329447746% max memory 411 MB = 126.3 KB
2019-10-18 18:56:47,094 INFO util.GSet: capacity = 2^14 = 16384 entries
2019-10-18 18:56:47,154 INFO namenode.FSImage: Allocated new BlockPoolId: BP-178131724-192.168.216.53-1571396207141
2019-10-18 18:56:47,182 INFO common.Storage: Storage directory /usr/local/hadoop/tmp/dfs/name has been successfully formatted.
2019-10-18 18:56:47,201 INFO namenode.FSImageFormatProtobuf: Saving image file /usr/local/hadoop/tmp/dfs/name/current/fsimage.ckpt_0000000000000000000 using no compression
2019-10-18 18:56:47,421 INFO namenode.FSImageFormatProtobuf: Image file /usr/local/hadoop/tmp/dfs/name/current/fsimage.ckpt_0000000000000000000 of size 393 bytes saved in 0 seconds .
2019-10-18 18:56:47,443 INFO namenode.NNStorageRetentionManager: Going to retain 1 images with txid >= 0
2019-10-18 18:56:47,454 INFO namenode.NameNode: SHUTDOWN_MSG:
/************************************************************
SHUTDOWN_MSG: Shutting down NameNode at web3/192.168.216.53
************************************************************/
[hadoop@web3 hadoop]$
执行后查看最后几行的info信息，查看是否成功
可以看到已经成功格式化
4、开启namenode和datanode守护进程
[hadoop@web3 hadoop]$ ./sbin/start-dfs.sh #--开启namenode和datanode守护进程
Starting namenodes on [localhost]
localhost: Warning: Permanently added 'localhost' (ECDSA) to the list of known hosts.
Starting datanodes
Starting secondary namenodes [web3]
web3: Warning: Permanently added 'web3,fe80::9416:80e8:f210:1e24%ens33' (ECDSA) to the list of known hosts.
-- ::, WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
[hadoop@web3 hadoop]$ jps #--检查是否启动，jps 看到namenode和datanode就说明启动了
NameNode
DataNode
SecondaryNameNode
Jps
[hadoop@web3 hadoop]$
如提示WARN util.NativeCodeLoader，整个提示不会影响正常启动
5、查看监听端口并访问web界面
1）查看监听端口
如下：应该是43332整个端口
[hadoop@web3 hadoop]$ netstat -unltop
(Not all processes could be identified, non-owned process info
will not be shown, you would have to be root to see it all.)
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name Timer
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN - off (0.00//)
tcp 192.168.122.1: 0.0.0.0:* LISTEN - off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN - off (0.00//)
tcp 127.0.0.1: 0.0.0.0:* LISTEN - off (0.00//)
tcp 127.0.0.1: 0.0.0.0:* LISTEN - off (0.00//)
tcp 127.0.0.1: 0.0.0.0:* LISTEN - off (0.00//)
tcp 0 0 127.0.0.1:43332 0.0.0.0:* LISTEN 17553/java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 127.0.0.1: 0.0.0.0:* LISTEN /java off (0.00//)
tcp6 ::: :::* LISTEN - off (0.00//)
tcp6 ::: :::* LISTEN - off (0.00//)
tcp6 ::: :::* LISTEN - off (0.00//)
tcp6 ::: :::* LISTEN - off (0.00//)
tcp6 ::: :::* LISTEN - off (0.00//)
udp 0.0.0.0: 0.0.0.0:* - off (0.00//)
udp 0.0.0.0: 0.0.0.0:* - off (0.00//)
udp 192.168.122.1: 0.0.0.0:* - off (0.00//)
udp 0.0.0.0: 0.0.0.0:* - off (0.00//)
[hadoop@web3 hadoop]$
2）访问web端
成功启动后web访问一下，可以查看namenode和datanode信息，还可以在线查看hdfs中的文件如下图：
http://localhost:43332
五、Hadoop伪分布式实例
1、HDFS中创建用户目录
#--HDFS中创建用户目录
[hadoop@web3 hadoop]$ ./bin/hdfs dfs -mkdir -p /user/hadoop
2019-10-18 22:56:44,350 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
2、创建一个input目录，并复制/usr/local/hadoop/etc/hadoop文件中的所有xml文件
[hadoop@web3 hadoop]$ ./bin/hdfs dfs -mkdir input
2019-10-18 22:58:03,745 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
[hadoop@web3 hadoop]$ ./bin/hdfs dfs -put ./etc/hadoop/*.xml input
2019-10-18 22:58:39,703 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
3、查看HDFS文件列表
[hadoop@web3 hadoop]$ ./bin/hdfs dfs -ls input
-- ::, WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
Found items
-rw-r--r-- hadoop supergroup -- : input/capacity-scheduler.xml
-rw-r--r-- hadoop supergroup -- : input/core-site.xml
-rw-r--r-- hadoop supergroup -- : input/hadoop-policy.xml
-rw-r--r-- hadoop supergroup -- : input/hdfs-site.xml
-rw-r--r-- hadoop supergroup -- : input/httpfs-site.xml
-rw-r--r-- hadoop supergroup -- : input/kms-acls.xml
-rw-r--r-- hadoop supergroup -- : input/kms-site.xml
-rw-r--r-- hadoop supergroup -- : input/mapred-site.xml
-rw-r--r-- hadoop supergroup -- : input/yarn-site.xml
[hadoop@web3 hadoop]$
4、实例演示
伪分布式运行mapreduce作业的方式和单机一样，区别在于伪分布式读取的是HDFS中的文件
[hadoop@web3 hadoop]$ ./bin/hadoop jar ./share/hadoop/mapreduce/hadoop-mapreduce-examples-3.1.2.jar grep input output 'dfs[a-z.]+'
2019-10-18 23:06:38,782 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
2019-10-18 23:06:40,494 INFO impl.MetricsConfig: loaded properties from hadoop-metrics2.properties
2019-10-18 23:06:40,809 INFO impl.MetricsSystemImpl: Scheduled Metric snapshot period at 10 second(s).
2019-10-18 23:06:40,810 INFO impl.MetricsSystemImpl: JobTracker metrics system started
2019-10-18 23:06:41,480 INFO input.FileInputFormat: Total input files to process : 9
2019-10-18 23:06:41,591 INFO mapreduce.JobSubmitter: number of splits:9
2019-10-18 23:06:42,290 INFO mapreduce.JobSubmitter: Submitting tokens for job: job_local1738759870_0001
2019-10-18 23:06:42,293 INFO mapreduce.JobSubmitter: Executing with tokens: []
。。。。。。。。。。。。。。#省略若干#。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。
Shuffle Errors
BAD_ID=
CONNECTION=
IO_ERROR=
WRONG_LENGTH=
WRONG_MAP=
WRONG_REDUCE=
File Input Format Counters
Bytes Read=
File Output Format Counters
Bytes Written=
#---检查运行结果
[hadoop@web3 hadoop]$ ./bin/hdfs dfs -cat output/*
2019-10-18 23:07:19,640 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
1 dfsadmin
1 dfs.replication
1 dfs.namenode.name.dir
1 dfs.datanode.data.dir
[hadoop@web3 hadoop]$
5、实例2，也可以把结果取回本地
删除本地output
rm -r ./output
将hdfs中的output拷贝到本机
./bin/hdfs dfs -get output ./output
查看
cat ./output/*
[hadoop@web3 hadoop]$ rm -r ./output
[hadoop@web3 hadoop]$ ./bin/hdfs dfs -get output ./output
-- ::, WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
[hadoop@web3 hadoop]$ cat ./output/*
1 dfsadmin
1 dfs.replication
1 dfs.namenode.name.dir
1 dfs.datanode.data.dir
[hadoop@web3 hadoop]$
删除hdfs output
注意hadoop运行程序时，输出目录不能存在，否则会提示错误
[hadoop@web3 hadoop]$ ./bin/hdfs dfs -rm -r output
-- ::, WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
Deleted output
[hadoop@web3 hadoop]$
六、启动YARN
伪分布式启动YARN也可以，一般不会影响程序执行，上面./sbin/start-dfs.sh启动hadoop，仅仅时启动了MapReduce环境，还可以启动YARN,让YARN来复制资源管理与任务调度。
还有上面例子未见JobTracker和TaskTracker，这时因为新版hadoop使用了新的MapReduce框架（MapReduce V2，也称为YARN，Yet Another Resource Negotiator）
YARN是从MapReduce中分离出来的，复制资源管理与任务调度。YARN运行于MapReduce之上，提供了高可用性、高扩展性
1、编辑mapred-site.xml
[hadoop@web3 hadoop]$ cat ./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>
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property>
</configuration>
2、编辑yarn-site.xml
[hadoop@web3 hadoop]$ cat ./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 -->
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
</configuration>
3、启动YARN
[hadoop@web3 hadoop]$ ./sbin/start-yarn.sh
Starting resourcemanager
Starting nodemanagers
[hadoop@web3 hadoop]$ jps
DataNode
ResourceManager #---启动后多了一个ResourceManager
Jps
NodeManager #---启动后多了一个NodeManager
SecondaryNameNode
NameNode
#---开启历史服务器，能在web中查看任务运行情况
[hadoop@web3 hadoop]$ ./sbin/mr-jobhistory-daemon.sh start historyserver
4、提示
启动YARN之后，运行实例的方法还是一样的，仅仅是资源管理方式、任务调度不同。观察日志可以发现，不启用YARN时，是“mapred.LocalJobRunner”在跑，启用YARN之后，是“mapred.YARNRuner”在跑任务，启用YARN有个好处是可以通过web界面查看任务情况
http://localhost:8088/cluster
通过netstat -untlop可以看到监听到了8088
[hadoop@web3 hadoop]$ netstat -untlop
(Not all processes could be identified, non-owned process info
will not be shown, you would have to be root to see it all.)
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name Timer
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN - off (0.00//)
tcp 192.168.122.1: 0.0.0.0:* LISTEN - off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN - off (0.00//)
tcp 127.0.0.1: 0.0.0.0:* LISTEN - off (0.00//)
tcp 0 0 0.0.0.0:8088 0.0.0.0:* LISTEN 24982/java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 127.0.0.1: 0.0.0.0:* LISTEN - off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 127.0.0.1: 0.0.0.0:* LISTEN - off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 127.0.0.1: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 0.0.0.0: 0.0.0.0:* LISTEN /java off (0.00//)
tcp 127.0.0.1: 0.0.0.0:* LISTEN /java off (0.00//)
tcp6 ::: :::* LISTEN - off (0.00//)
tcp6 ::: :::* LISTEN - off (0.00//)
tcp6 ::: :::* LISTEN - off (0.00//)
tcp6 ::: :::* LISTEN - off (0.00//)
tcp6 ::: :::* LISTEN - off (0.00//)
udp 0.0.0.0: 0.0.0.0:* - off (0.00//)
udp 0.0.0.0: 0.0.0.0:* - off (0.00//)
udp 192.168.122.1: 0.0.0.0:* - off (0.00//)
udp 0.0.0.0: 0.0.0.0:* - off (0.00//)
[hadoop@web3 hadoop]$
5、访问web界面
6、运行一个任务
提示错误
Error: Could not find or load main class org.apache.hadoop.mapreduce.v2.app.MRAppMaster Please check whether your etc/hadoop/mapred-site.xml contains the below configuration:
看到如下提示，下面排错就按照提示修改mapred-site.xml
[-- ::52.678]Container exited with a non-zero exit code . Error file: prelaunch.err.
Last bytes of prelaunch.err :
Last bytes of stderr :
Error: Could not find or load main class org.apache.hadoop.mapreduce.v2.app.MRAppMaster Please check whether your etc/hadoop/mapred-site.xml contains the below configuration:
<property>
<name>yarn.app.mapreduce.am.env</name>
<value>HADOOP_MAPRED_HOME=${full path of your hadoop distribution directory}</value>
</property>
<property>
<name>mapreduce.map.env</name>
<value>HADOOP_MAPRED_HOME=${full path of your hadoop distribution directory}</value>
</property>
<property>
<name>mapreduce.reduce.env</name>
<value>HADOOP_MAPRED_HOME=${full path of your hadoop distribution directory}</value>
</property> [-- ::52.679]Container exited with a non-zero exit code . Error file: prelaunch.err.
Last bytes of prelaunch.err :
Last bytes of stderr :
Error: Could not find or load main class org.apache.hadoop.mapreduce.v2.app.MRAppMaster Please check whether your etc/hadoop/mapred-site.xml contains the below configuration:
<property>
<name>yarn.app.mapreduce.am.env</name>
<value>HADOOP_MAPRED_HOME=${full path of your hadoop distribution directory}</value>
</property>
<property>
<name>mapreduce.map.env</name>
<value>HADOOP_MAPRED_HOME=${full path of your hadoop distribution directory}</value>
</property>
<property>
<name>mapreduce.reduce.env</name>
<value>HADOOP_MAPRED_HOME=${full path of your hadoop distribution directory}</value>
</property>
7、排错
修改配置文件mapred-site.xml，
[root@web3 hadoop]# cat ./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>
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property>
<property>
<name>yarn.app.mapreduce.am.env</name>
<value>HADOOP_MAPRED_HOME=$HADOOP_HOME</value>
</property>
<property>
<name>mapreduce.map.env</name>
<value>HADOOP_MAPRED_HOME=$HADOOP_HOME</value>
</property>
<property>
<name>mapreduce.reduce.env</name>
<value>HADOOP_MAPRED_HOME=$HADOOP_HOME</value>
</property>
</configuration>
[root@web3 hadoop]#
7、再次运行
运行成功
[hadoop@web3 hadoop]$ ./bin/hadoop jar ./share/hadoop/mapreduce/hadoop-mapreduce-examples-3.1..jar grep input output 'dfs[a-z.]+'
很明显，YARN主要是为集群提供更好的资源管理与任务调度，在单机上反之会使程序跑的更慢，所以单机是否开启YARN要看实际情况
8、关闭YARN
./sbin/stop-yarn.sh
./sbin/mr-jobhistory-daemon.sh stop historyserver
本文参考1：http://dblab.xmu.edu.cn/blog/install-hadoop-in-centos/
本文参考2：http://hadoop.apache.org/docs/stable/hadoop-project-dist/hadoop-common/SingleCluster.html