2. Hadoop三种集群方式
1. 三种集群方式
-
本地模式
hdfs dfs -ls /
不需要启动任何进程 -
伪分布式
所有进程跑在一个机器上 -
完全分布式
每个机器运行不同的进程
2. 服务器基本配置
2.1 服务器配置及系统版本
- CPU: 2核
- 内存: 4G
- 系统版本: Centos7 1511
2.2 服务器IP及主机名设置
-
服务器数量: 五台机器
主机名 公网IP 内网IP hadoop-1 192.168.10.145 172.16.1.207 hadoop-2 192.168.10.149 172.16.1.206 hadoop-3 192.168.10.152 172.16.1.204 hadoop-4 192.168.10.153 172.16.1.208 hadoop-5 192.168.10.156 172.16.1.205 -
根据以上表格修改hosts表和主机名
修改Hosts
#vim /etc/hosts
192.168.10.145 hadoop-1
192.168.10.149 hadoop-2
192.168.10.152 hadoop-3
192.168.10.153 hadoop-4
192.168.10.156 hadoop-5 #scp /etc/hosts hadoop-2:/etc/
#scp /etc/hosts hadoop-3:/etc/
#scp /etc/hosts hadoop-4:/etc/
#scp /etc/hosts hadoop-5:/etc/ 设置主机名
#hostnamectl set-hostname hadoop-1
#hostnamectl set-hostname hadoop-2
#hostnamectl set-hostname hadoop-3
#hostnamectl set-hostname hadoop-4
#hostnamectl set-hostname hadoop-5 -
ssh认证
hadoop-1主机上执行
#ssh-keygen -t rsa -P ''
#ssh-copy-id 192.168.10.145
#scp -r .ssh 192.168.10.149:/root/
#scp -r .ssh 192.168.10.152:/root/
#scp -r .ssh 192.168.10.153:/root/
#scp -r .ssh 192.168.10.156:/root/ -
创建 /soft 存放jdk和Hadoop目录
#ssh hadoop-1 'mkdir /soft'
#ssh hadoop-2 'mkdir /soft'
#ssh hadoop-3 'mkdir /soft'
#ssh hadoop-4 'mkdir /soft'
#ssh hadoop-5 'mkdir /soft' -
安装jdk
#cd /root/
#确保已经下载了相关jdk包
#scp jdk-8u131-linux-x64.tar.gz hadoop-1:/soft/
#scp jdk-8u131-linux-x64.tar.gz hadoop-2:/soft/
#scp jdk-8u131-linux-x64.tar.gz hadoop-3:/soft/
#scp jdk-8u131-linux-x64.tar.gz hadoop-4:/soft/
#scp jdk-8u131-linux-x64.tar.gz hadoop-5:/soft/ 所有的服务器
#tar xf /soft/jdk-8u131-linux-x64.tar.gz -C /soft
#ln -s /soft/jdk1.8.0_131 /soft/jdk 创建软连接
配置环境变量
#vim /etc/profile
JAVA_HOME=/soft/jdk
PATH=$PATH:$JAVA_HOME/bin
#source /etc/profile -
配置hadoop
#scp hadoop-2.7.3.tar.gz hadoop-1:/soft/
#scp hadoop-2.7.3.tar.gz hadoop-2:/soft/
#scp hadoop-2.7.3.tar.gz hadoop-3:/soft/
#scp hadoop-2.7.3.tar.gz hadoop-4:/soft/
#scp hadoop-2.7.3.tar.gz hadoop-5:/soft/
所有服务器
#tar xf /soft/hadoop-2.7.3.tar.gz -C /soft
#ln -s /soft/hadoop-2.7.3 /soft/hadoop 创建软连接
修改环境变量
#vim /etc/profile
HADOOP_HOME=/soft/hadoop
PATH=$PATH:$HADOOP_HOME/bin:$HADOOP_HOME/sbin
#source /etc/profile
修改hadoop-env.sh
# vim /soft/hadoop/etc/hadoop/hadoop-env.sh
修改hadoop-env.sh 修改hadoop的jAVA_HOME
export JAVA_HOME=/soft/jdk
3. 本地模式
- 本地就是单机模式，hadoop默认安装完就是单机模式
- hdfs默认使用本地的文件系统
- hdfs dfs -ls / 查看本地文件系统 和linux的ls /一样
3.1 测试单词统计
#hdfs dfs -mkdir /input
#cd /input/
#echo “hello word” > file1.txt
#echo “hello hadoop” > file2.txt
#echo “hello mapreduce” >> file2.txt
#cd /soft/hadoop/
#hadoop jar share/hadoop/mapreduce/hadoop-mapreduce-examples-2.7.3.jar wordcount /input /output
#hdfs dfs -ls /output/
Found 2 items
-rw-r--r-- 1 root root 0 2017-06-21 11:56 /output/_SUCCESS
-rw-r--r-- 1 root root 48 2017-06-21 11:56 /output/part-r-00000
4. 伪分布式
4.1 伪分布式介绍
Hadoop 可以在单节点上以伪分布式的方式运行，Hadoop 进程以分离的 Java 进程来运行，节点既作为 NameNode 也作为 DataNode，同时，读取的是 HDFS 中的文件。
Hadoop 的配置文件位于 /soft/hadoop/etc/hadoop/ 中，伪分布式需要修改2个配置文件 core-site.xml 和 hdfs-site.xml 。Hadoop的配置文件是 xml 格式，每个配置以声明 property 的 name 和 value 的方式来实现。
4.2 伪分布式搭建
-
配置伪分布式
#mkdir /data/hadoop
#cd /soft/hadoop/etc/
#mv hadoop local
#cp -r local pseudo
#ln -s pseudo hadoop
#cd hadoop -
修改core-site.xml配置文件
#vim core-site.xml
[core-site.xml配置如下] <?xml version="1.0"?>
<configuration>
<property>
<name>hadoop.tmp.dir</name>
<value>file:/data/hadoop/tmp</value>
<description>Abase for other temporary directories.</description>
</property>
<property>
<name>fs.defaultFS</name>
<value>hdfs://localhost/</value>
</property>
</configuration> -
修改hdfs-site.xml配置文件
#vim hdfs-site.xml
[hdfs-site.xml配置如下]
<?xml version="1.0"?>
<configuration>
<property>
<name>dfs.replication</name>
<value>1</value>
</property>
<property>
<name>dfs.namenode.name.dir</name>
<value>file:/data/hadoop/tmp/dfs/name</value>
</property>
<property>
<name>dfs.datanode.data.dir</name>
<value>file:/data/hadoop/tmp/dfs/data</value>
</property>
</configuration> hadoop 的运行方式是由配置文件决定的（运行 Hadoop 时会读取配置文件）
因此如果需要从伪分布式模式切换回非分布式模式，需要删除 core-site.xml 中的配置项。
此外，伪分布式虽然只需要配置 fs.defaultFS 和 dfs.replication 就可以运行（官方教程如此）
不过若没有配置 hadoop.tmp.dir 参数，则默认使用的临时目录为 /tmp/hadoo-hadoop，而这个目录在重启时有可能被系统清理掉，导致必须重新执行 format 才行。所以我们进行了设置，同时也指定 dfs.namenode.name.dir 和 dfs.datanode.data.dir，否则在接下来的步骤中可能会出错。
YARN 是从 MapReduce 中分离出来的，负责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，YARN 的更多介绍在此不展开，有兴趣的可查阅相关资料 -
修改mapred-site.xml配置文件
#cp mapred-site.xml.template mapred-site.xml
#vim mapred-site.xml
[mapred-site.xml配置如下]
<?xml version="1.0"?>
<configuration>
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property>
</configuration> -
修改yarn-site.xml配置文件
#vim yarn-site.xml
[yarn-site.xml配置如下]
<?xml version="1.0"?>
<configuration>
<property>
<name>yarn.resourcemanager.hostname</name>
<value>localhost</value>
</property>
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
</configuration> -
修改slaves配置文件
#vim slaves
[slaves配置如下]
localhost -
格式化hdfs分布式文件系统
#hadoop namenode -format
[root@hadoop-1 hadoop]# hadoop namenode -format 省略--------
17/05/15 09:29:01 INFO util.ExitUtil: Exiting with status 0
17/05/15 09:29:01 INFO namenode.NameNode: SHUTDOWN_MSG:
/************************************************************
SHUTDOWN_MSG: Shutting down NameNode at hadoop-1/172.16.1.207
************************************************************/ -
启动hadoop服务
#start-all.sh
[root@hadoop-1 hadoop]# start-all.sh
This script is Deprecated. Instead use start-dfs.sh and start-yarn.sh
Starting namenodes on [localhost]
The authenticity of host 'localhost (::1)' can't be established.
ECDSA key fingerprint is da:38:db:62:7e:97:52:6e:11:1b:81:93:1b:a4:b4:e6.
Are you sure you want to continue connecting (yes/no)? yes
localhost: Warning: Permanently added 'localhost' (ECDSA) to the list of known hosts.
localhost: starting namenode, logging to /soft/hadoop-2.7.3/logs/hadoop-root-namenode-hadoop-1.out
localhost: starting datanode, logging to /soft/hadoop-2.7.3/logs/hadoop-root-datanode-hadoop-1.out
Starting secondary namenodes [0.0.0.0]
The authenticity of host '0.0.0.0 (0.0.0.0)' can't be established.
ECDSA key fingerprint is da:38:db:62:7e:97:52:6e:11:1b:81:93:1b:a4:b4:e6.
Are you sure you want to continue connecting (yes/no)? yes
0.0.0.0: Warning: Permanently added '0.0.0.0' (ECDSA) to the list of known hosts.
0.0.0.0: starting secondarynamenode, logging to /soft/hadoop-2.7.3/logs/hadoop-root-secondarynamenode-hadoop-1.out
starting yarn daemons
starting resourcemanager, logging to /soft/hadoop-2.7.3/logs/yarn-root-resourcemanager-hadoop-1.out
localhost: starting nodemanager, logging to /soft/hadoop-2.7.3/logs/yarn-root-nodemanager-hadoop-1.out
9. **判断是否启动成功**
```
启动完成后，可以通过命令 jps 来判断是否成功启动，若成功启动则会列出如下进程:
[root@hadoop-1 hadoop]# jps
14784 NameNode
15060 SecondaryNameNode
14904 DataNode
15211 ResourceManager
15628 Jps
15374 NodeManager
如果 SecondaryNameNode 没有启动，请运行 sbin/stop-dfs.sh 关闭进程，
然后再次尝试启动尝试）。如果没有 NameNode 或 DataNode ，那就是配置不成功，
请仔细检查之前步骤，或通过查看启动日志排查原因
```
10. **登陆WEB查看**
打开http://192.168.10.145:50070
##4.2 伪分布式单词统计
1. **在本地创建目录和分析的log**
```
#mkdir /input
#cd /input
#echo "hello world" > file1.log
#echo "hello world" > file2.log
#echo "hello hadoop" > file3.log
#echo "hello hadoop" > file4.log
#echo "map" > file5.log
```
2. **在hdfs创建目录和上传本地log**
```
#hdfs dfs -mkdir -p /input/
#hdfs dfs -ls /
Found 1 items
drwxr-xr-x - root supergroup 0 2017-06-22 10:29 /input
#hdfs dfs -put file* /input/
# hdfs dfs -ls /input/
Found 5 items
-rw-r--r-- 1 root supergroup 12 2017-06-22 10:32 /input/file1.log
-rw-r--r-- 1 root supergroup 12 2017-06-22 10:32 /input/file2.log
-rw-r--r-- 1 root supergroup 13 2017-06-22 10:32 /input/file3.log
-rw-r--r-- 1 root supergroup 13 2017-06-22 10:32 /input/file4.log
-rw-r--r-- 1 root supergroup 4 2017-06-22 10:32 /input/file5.log
```
3. **使用自带jar包进行单词统计**
```
# hadoop jar /soft/hadoop/share/hadoop/mapreduce/hadoop-mapreduce-examples-2.7.3.jar wordcount /input /output
17/05/15 09:48:38 INFO client.RMProxy: Connecting to ResourceManager at localhost/127.0.0.1:8032
17/05/15 09:48:39 INFO input.FileInputFormat: Total input paths to process : 5
17/05/15 09:48:40 INFO mapreduce.JobSubmitter: number of splits:5
17/05/15 09:48:40 INFO mapreduce.JobSubmitter: Submitting tokens for job: job_1494855010567_0001
17/05/15 09:48:40 INFO impl.YarnClientImpl: Submitted application application_1494855010567_0001
17/05/15 09:48:40 INFO mapreduce.Job: The url to track the job: http://hadoop-1:8088/proxy/application_1494855010567_0001/
17/05/15 09:48:40 INFO mapreduce.Job: Running job: job_1494855010567_0001
17/05/15 09:48:48 INFO mapreduce.Job: Job job_1494855010567_0001 running in uber mode : false
17/05/15 09:48:48 INFO mapreduce.Job: map 0% reduce 0%
17/05/15 09:48:59 INFO mapreduce.Job: map 20% reduce 0%
17/05/15 09:49:00 INFO mapreduce.Job: map 80% reduce 0%
17/05/15 09:49:01 INFO mapreduce.Job: map 100% reduce 0%
17/05/15 09:49:06 INFO mapreduce.Job: map 100% reduce 100%
17/05/15 09:49:06 INFO mapreduce.Job: Job job_1494855010567_0001 completed successfully
17/05/15 09:49:06 INFO mapreduce.Job: Counters: 50
File System Counters
FILE: Number of bytes read=114
FILE: Number of bytes written=711875
FILE: Number of read operations=0
FILE: Number of large read operations=0
FILE: Number of write operations=0
HDFS: Number of bytes read=539
HDFS: Number of bytes written=31
HDFS: Number of read operations=18
HDFS: Number of large read operations=0
HDFS: Number of write operations=2
Job Counters
Killed map tasks=1
Launched map tasks=5
Launched reduce tasks=1
Data-local map tasks=5
Total time spent by all maps in occupied slots (ms)=48562
Total time spent by all reduces in occupied slots (ms)=4413
Total time spent by all map tasks (ms)=48562
Total time spent by all reduce tasks (ms)=4413
Total vcore-milliseconds taken by all map tasks=48562
Total vcore-milliseconds taken by all reduce tasks=4413
Total megabyte-milliseconds taken by all map tasks=49727488
Total megabyte-milliseconds taken by all reduce tasks=4518912
Map-Reduce Framework
Map input records=5
Map output records=9
Map output bytes=90
Map output materialized bytes=138
Input split bytes=485
Combine input records=9
Combine output records=9
Reduce input groups=4
Reduce shuffle bytes=138
Reduce input records=9
Reduce output records=4
Spilled Records=18
Shuffled Maps =5
Failed Shuffles=0
Merged Map outputs=5
GC time elapsed (ms)=1662
CPU time spent (ms)=2740
Physical memory (bytes) snapshot=1523605504
Virtual memory (bytes) snapshot=12609187840
Total committed heap usage (bytes)=1084227584
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=54
File Output Format Counters
Bytes Written=31
```
4. ** 查看结果**
```
[root@hadoop-1 ~]# hdfs dfs -cat /output/*
hadoop 2
hello 4
map 1
world
[root@hadoop-1 ~]# hdfs dfs -get /output/* .
[root@hadoop-1 ~]# ls
file1.log file2.log file3.log file4.log file5.log part-r-00000 _SUCCESS
[root@hadoop-1 ~]# cat part-r-00000
hadoop 2
hello 4
map 1