hadoop源码编译及安装配置

参考博客原址：https://blog.csdn.net/liweihope/article/details/89605340

前言

直接从官网上下载的hadoop-2.6.0-cdh5.7.0.tar.gz部署后，不支持压缩，生产上是不可接受的，可以用hadoop checknative命令查看hadoop所支持的本地库：

#这个是直接下载的hadoop包部署的不支持压缩
[hadoop@hadoop001 ~]$ hadoop checknative
19/04/27 10:22:33 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
Native library checking:
hadoop:  false 
zlib:    false 
snappy:  false 
lz4:     false 
bzip2:   false 
openssl: false 
19/04/27 10:22:34 INFO util.ExitUtil: Exiting with status 1
[hadoop@hadoop001 ~]$ 

   
   

   
   

• Unix System
• JDK 1.7+
• Maven 3.0 or later
• Findbugs 1.3.9 (if running findbugs)
• ProtocolBuffer 2.5.0
• CMake 2.6 or newer (if compiling native code), must be 3.0 or newer on Mac
• Zlib devel (if compiling native code)
• openssl devel ( if compiling native hadoop-pipes )
• Internet connection for first build (to fetch all Maven and Hadoop dependencies)

组件名称	版本
centos	centos6.6
Hadoop	Hadoop-2.6.0-cdh5.7.0-src.tar.gz
jdk	jdk-7u80-linux-x64.tar.gz
maven	apache-maven-3.3.9-bin.tar.gz
protobuf	protobuf-2.5.0.tar.gz

需要注意的是：编译的JDK版本必须是1.7版本，1.8的JDK会导致编译失败。
jdk版本可以安装好多个，需要哪个版本的时候在环境变量里面把注释去掉，把其它版本的注释掉就可以了。
大致过程：把以上软件下载好之后，需要部署安装jdk、maven、protobuf、以及相关的依赖，然后用maven编译Hadoop源码，编译之后，部署Hadoop，验证是否可行。

安装依赖库

root用户：

[root@hadoop001 ~]# yum install -y svn ncurses-devel
[root@hadoop001 ~]# yum install -y gcc gcc-c++ make cmake
[root@hadoop001 ~]# yum install -y openssl openssl-devel svn ncurses-devel zlib-devel libtool
[root@hadoop001 ~]# yum install -y snappy snappy-devel bzip2 bzip2-devel lzo lzo-devel lzop autoconf automake cmake

 
 

 
 
1
2
3
4

上传软件到系统上

以前已经存在hadoop用户，没有的话创建一下
hadoop用户

#创建三个文件夹：软件、maven仓库、源码
[hadoop@hadoop001 ~]$ mkdir soft maven_repo source
[hadoop@hadoop001 ~]$ cd soft/
[hadoop@hadoop001 soft]$ rz
[hadoop@hadoop001 soft]$ ll
total 202192
-rw-r--r-- 1 hadoop hadoop   8491533 Apr 27 09:19 apache-maven-3.3.9-bin.tar.gz
-rw-r--r-- 1 hadoop hadoop  42610549 Apr 27 09:19 hadoop-2.6.0-cdh5.7.0-src.tar.gz
-rw-r--r-- 1 hadoop hadoop 153530841 Apr 27 09:18 jdk-7u80-linux-x64.tar.gz
-rw-r--r-- 1 hadoop hadoop   2401901 Apr 27 09:18 protobuf-2.5.0.tar.gz

 
 

 
 
1
2
3
4
5
6
7
8
9
10

安装JDK

root用户：
解压安装包，安装目录必须是/usr/java,安装后记得修改拥有者为root

[root@hadoop001 ~]# tar -zxvf /home/hadoop/soft/jdk-7u80-linux-x64.tar.gz -C /usr/java/
[root@hadoop001 ~]# cd /usr/java
[root@hadoop001 java]# ll
total 169220
drwxr-xr-x 8 uucp  143      4096 Apr 11  2015 jdk1.7.0_80
drwxr-xr-x 8 root root      4096 Apr 11  2015 jdk1.8.0_45
-rw-r--r-- 1 root root 173271626 Jan 31 00:34 jdk-8u45-linux-x64.gz
[root@hadoop001 java]# chown -R root:root jdk1.7.0_80
[root@hadoop001 java]# ll
total 169220
drwxr-xr-x 8 root root      4096 Apr 11  2015 jdk1.7.0_80
drwxr-xr-x 8 root root      4096 Apr 11  2015 jdk1.8.0_45
-rw-r--r-- 1 root root 173271626 Jan 31 00:34 jdk-8u45-linux-x64.gz
[root@hadoop001 java]# 

 
 

 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14

添加环境变量
因为我的环境之前安装了jdk1.8版本，所以在这里要把这1.8的给注释掉就好了。

[root@hadoop001 java]# vi /etc/profile
#添加如下环境变量
#export JAVA_HOME=/usr/java/jdk1.8.0_45
export JAVA_HOME=/usr/java/jdk1.7.0_80
export JRE_HOME=$JAVA_HOME/jre
export CLASSPATH=.:$JAVA_HOME/lib:$JER_HOME/lib:$CLASSPATH
export PATH=$JAVA_HOME/bin:$JER_HOME/bin:$PATH
#测试java是否安装成功
[root@hadoop001 java]# source /etc/profile
[root@hadoop001 java]# java -version
java version "1.7.0_80"
Java(TM) SE Runtime Environment (build 1.7.0_80-b15)
Java HotSpot(TM) 64-Bit Server VM (build 24.80-b11, mixed mode)
[root@hadoop001 java]# 

 
 

 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14

安装maven

hadoop用户：

[root@hadoop001 java]# su - hadoop
[hadoop@hadoop001 ~]$ tar -zxvf ~/soft/apache-maven-3.3.9-bin.tar.gz -C ~/app/

 
 

 
 
1
2

添加环境变量

#修改haoop用户的环境变量
[hadoop@hadoop001 ~]$ vi ~/.bash_profile
#添加或修改如下内容，注意MAVEN_OPTS设置了maven运行的内存，防止内存太小导致编译失败
export MAVEN_HOME=/home/hadoop/app/apache-maven-3.3.9
export MAVEN_OPTS="-Xms1024m -Xmx1024m"
export PATH=$MAVEN_HOME/bin:$PATH
#测试maven是否安装成功
[hadoop@hadoop001 ~]$ source ~/.bash_profile 
[hadoop@hadoop001 ~]$ which mvn
~/app/apache-maven-3.3.9/bin/mvn
[hadoop@hadoop001 ~]$ 

 
 

 
 
1
2
3
4
5
6
7
8
9
10
11

配置maven

[hadoop@hadoop001 protobuf-2.5.0]$ vim ~/app/apache-maven-3.3.9/conf/settings.xml
#配置maven的本地仓库位置
<localRepository>/home/hadoop/maven_repo/repo</localRepository>
#添加阿里云中央仓库地址，注意一定要写在<mirrors></mirrors>之间，否则后面编译会报错
<mirror>
     <id>nexus-aliyun</id>
     <mirrorOf>central</mirrorOf>
     <name>Nexus aliyun</name>
     <url>http://maven.aliyun.com/nexus/content/groups/public</url>
</mirror>

 
 

 
 
1
2
3
4
5
6
7
8
9
10

小知识点：网络慢可能导致mvn第一次编译时下载需要超长的时间甚至编译失败。如果有别人或者自己已经建立好的本地仓库，可以把maven本地仓库添加到/home/hadoop/maven_repo/中去。

安装protobuf

hadoop用户：
解压

[hadoop@hadoop001 ~]$ tar -zxvf ~/soft/protobuf-2.5.0.tar.gz -C ~/app/
[hadoop@hadoop001 protobuf-2.5.0]$ ls
aclocal.m4    config.sub        depcomp                       install-sh   Makefile.am          python
autogen.sh    configure         editors                       INSTALL.txt  Makefile.in          README.txt
CHANGES.txt   configure.ac      examples                      java         missing              src
config.guess  CONTRIBUTORS.txt  generate_descriptor_proto.sh  ltmain.sh    protobuf-lite.pc.in  vsprojects
config.h.in   COPYING.txt       gtest                         m4           protobuf.pc.in

 
 

 
 
1
2
3
4
5
6
7

编译protobuf

#./configure  --prefix是指定编译好的包要存放的路径，软件要安装的路径
#make 进行编译，make install进行安装
[hadoop@hadoop001 protobuf-2.5.0]$ ./configure  --prefix=/home/hadoop/app/protobuf-2.5.0
[hadoop@hadoop001 protobuf-2.5.0]$ make
[hadoop@hadoop001 protobuf-2.5.0]$ make install

 
 

 
 
1
2
3
4
5

添加环境变量

[hadoop@hadoop001 protobuf-2.5.0]$ vi ~/.bash_profile 
#追加如下两行内容
export PROTOBUF_HOME=/home/hadoop/app/protobuf-2.5.0
export PATH=$PROTOBUF_HOME/bin:$PATH
#测试是否生效，若出现libprotoc 2.5.0则为生效
[hadoop@hadoop001 protobuf-2.5.0]$ source ~/.bash_profile 
[hadoop@hadoop001 protobuf-2.5.0]$ protoc --version
libprotoc 2.5.0
[hadoop@hadoop001 protobuf-2.5.0]$ 

 
 

 
 
1
2
3
4
5
6
7
8
9

编译hadoop

hadoop用户：
解压

[hadoop@hadoop001 ~]$ tar -zxvf ~/soft/hadoop-2.6.0-cdh5.7.0-src.tar.gz -C ~/source/
#进入hadoop的源码目录
[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ ls
BUILDING.txt       hadoop-client          hadoop-mapreduce1-project  hadoop-project       LICENSE.txt
cloudera           hadoop-common-project  hadoop-mapreduce-project   hadoop-project-dist  NOTICE.txt
dev-support        hadoop-dist            hadoop-maven-plugins       hadoop-tools         pom.xml
hadoop-assemblies  hadoop-hdfs-project    hadoop-minicluster         hadoop-yarn-project  README.txt

 
 

 
 
1
2
3
4
5
6
7

编译hadoop使其支持压缩：mvn clean package -Pdist,native -DskipTests -Dtar

#进入hadoop的源码目录
[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ cd ~/source/hadoop-2.6.0-cdh5.7.0/
[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ mvn clean package -Pdist,native -DskipTests -Dtar
#进行编译，第一次编译会下载很多依赖的jar包，快慢由网速决定，需耐心等待
#我是把别人的本地仓库拷贝到我的本地仓库，所以很快

 
 

 
 
1
2
3
4
5

若报异常，主要信息如下（无异常跳过）(这个我没有碰到，是其他人碰到，学习一下)：

[FATAL] Non-resolvable parent POM for org.apache.hadoop:hadoop-main:2.6.0-cdh5.7.0: Could not transfer artifact com.cloudera.cdh:cdh-root:pom:5.7.0 from/to cdh.repo (https://repository.cloudera.com/artifactory/cloudera-repos): Remote host closed connectio
#分析：是https://repository.cloudera.com/artifactory/cloudera-repos/com/cloudera/cdh/cdh-root/5.7.0/cdh-root-5.7.0.pom文件下载不了，但是虚拟机确实是ping通远程的仓库，很是费解为什么。
#解决方案：前往本地仓库到目标文件目录，然后 通过wget 文件，来成功获取该文件，重新执行编译命令，或者执行4.5的可选步骤，将需要的jar直接放到本地仓库

 
 

 
 
1
2
3

编译成功后会有提示：

[INFO] ------------------------------------------------------------------------
[INFO] BUILD SUCCESS
[INFO] ------------------------------------------------------------------------

 
 

 
 
1
2
3

查看编译后的包：hadoop-2.6.0-cdh5.7.0.tar.gz

[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ ll /home/hadoop/source/hadoop-2.6.0-cdh5.7.0/hadoop-dist/target/
total 563956
drwxrwxr-x 2 hadoop hadoop      4096 Apr 27 12:42 antrun
drwxrwxr-x 3 hadoop hadoop      4096 Apr 27 12:42 classes
-rw-rw-r-- 1 hadoop hadoop      1998 Apr 27 12:42 dist-layout-stitching.sh
-rw-rw-r-- 1 hadoop hadoop       690 Apr 27 12:42 dist-tar-stitching.sh
drwxrwxr-x 9 hadoop hadoop      4096 Apr 27 12:42 hadoop-2.6.0-cdh5.7.0
-rw-rw-r-- 1 hadoop hadoop 191842632 Apr 27 12:42 hadoop-2.6.0-cdh5.7.0.tar.gz
-rw-rw-r-- 1 hadoop hadoop      7314 Apr 27 12:42 hadoop-dist-2.6.0-cdh5.7.0.jar
-rw-rw-r-- 1 hadoop hadoop 385574240 Apr 27 12:43 hadoop-dist-2.6.0-cdh5.7.0-javadoc.jar
-rw-rw-r-- 1 hadoop hadoop      4855 Apr 27 12:42 hadoop-dist-2.6.0-cdh5.7.0-sources.jar
-rw-rw-r-- 1 hadoop hadoop      4855 Apr 27 12:42 hadoop-dist-2.6.0-cdh5.7.0-test-sources.jar
drwxrwxr-x 2 hadoop hadoop      4096 Apr 27 12:42 javadoc-bundle-options
drwxrwxr-x 2 hadoop hadoop      4096 Apr 27 12:42 maven-archiver
drwxrwxr-x 3 hadoop hadoop      4096 Apr 27 12:42 maven-shared-archive-resources
drwxrwxr-x 3 hadoop hadoop      4096 Apr 27 12:42 test-classes
drwxrwxr-x 2 hadoop hadoop      4096 Apr 27 12:42 test-dir

 
 

 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

伪分布式部署

由于我之前已经部署过hadoop-2.6.0-cdh5.7.0，所以这里先把它名字修改一下

hadoop@hadoop001 app]$ mv  hadoop-2.6.0-cdh5.7.0 hadoop-2.6.0-cdh5.7.0-20190427bak

 
 

 
 
1

解压安装包

[hadoop@hadoop001 target]$ pwd
/home/hadoop/source/hadoop-2.6.0-cdh5.7.0/hadoop-dist/target
[hadoop@hadoop001 target]$ tar -zxvf hadoop-2.6.0-cdh5.7.0.tar.gz -C ~/app

 
 

 
 
1
2
3

配置环境变量
(因为我之前已经配置过环境变量，所以这里不需要配置，没有配的话，可以这样配一下)

[hadoop@hadoop001 ~]$ vi ~/.bash_profile 
export HADOOP_HOME=/home/hadoop/app/hadoop-2.6.0-cdh5.7.0
export PATH=$HADOOP_HOME/bin:$HADOOP_HOME/sbin:$PATH
[hadoop@hadoop001 ~]$ source ~/.bash_profile 
[hadoop@hadoop001 ~]$ which hadoop
~/app/hadoop-2.6.0-cdh5.7.0/bin/hadoop

 
 

 
 
1
2
3
4
5
6

配置ssh

为什么要配置ssh，因为hadoop在启动的时候要用到ssh。

[hadoop@hadoop001 ~]$ rm -rf ~/.ssh
[hadoop@hadoop001 ~]$ ssh-keygen 
[hadoop@hadoop001 ~]$ cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
[hadoop@hadoop001 ~]$ chmod 600 ~/.ssh/authorized_keys
#测试ssh是否成功，用户第一ssh会提示输入是否连接，yes。成功显示时间
[hadoop@hadoop001 ~]$ ssh hadoop001 date

 
 

 
 
1
2
3
4
5
6

修改配置文件

编辑hadoop-env.sh文件相关配置

[hadoop@hadoop001 hadoop]$ pwd
/home/hadoop/app/hadoop-2.6.0-cdh5.7.0/etc/hadoop
[hadoop@hadoop001 hadoop]$ vi hadoop-env.sh
export JAVA_HOME=/usr/java/jdk1.7.0_80

• 1
• 2
• 3
• 4
• 5

编辑core-site.xml文件相关配置

#这里我创建了hadoop_tmp目录
[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ mkdir hadoop_tmp
[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ chmod 777 hadoop_tmp

[hadoop@hadoop001 hadoop]$ vi core-site.xml
<configuration>
<property>
<name>fs.defaultFS</name>
<value>hdfs://hadoop001:9000</value>
</property>
#这里我配置了它的临时目录设定，默认是在/tmp/hadoop目录下的，linux会定期清理/tmp目录
<property>
<name>hadoop.tmp.dir</name>
<value>/home/hadoop/app/hadoop-2.6.0-cdh5.7.0/hadoop_tmp</value>
</property>
</configuration>

• 1
• 2
• 3
• 4
• 5
• 6
• 7
• 8
• 9
• 10
• 11
• 12
• 13
• 14
• 15
• 16

编辑hdfs-site.xml文件相关配置

[hadoop@hadoop001 hadoop]$ vi hdfs-site.xml 
<configuration>
    <property>
        <name>dfs.replication</name>
        <value>1</value>
    </property>
    <property>
        <name>dfs.namenode.secondary.http-address</name>
        <value>hadoop001:50090</value>
    </property>
    <property>
        <name>dfs.namenode.secondary.https-address</name>
        <value>hadoop001:50091</value>
    <property>
</configuration>

 
 

 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

编辑mapred-site.xml文件相关配置

[hadoop@hadoop001 hadoop]$ cp mapred-site.xml.template mapred-site.xml
[hadoop@hadoop001 hadoop]$ vi mapred-site.xml
<configuration>
    <property>
        <name>mapreduce.framework.name</name>
        <value>yarn</value>
    </property>
</configuration>

 
 

 
 
1
2
3
4
5
6
7
8

编辑yarn-site.xml文件相关配置

[hadoop@hadoop001 hadoop]$ vi yarn-site.xml
<configuration>

<!-- Site specific YARN configuration properties -->
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
</configuration>

• 1
• 2
• 3
• 4
• 5
• 6
• 7
• 8
• 9

添加slaves节点

[hadoop@hadoop001 ~]$ vim ~/app/hadoop-2.6.0-cdh5.7.0/etc/hadoop/slaves
#把默认的localhost修改成hadoop001
hadoop001

 
 

 
 
1
2
3

格式化namenode

[hadoop@hadoop001 ~]$ hdfs namenode -format
......
#若出现 has been successfully formatted 则表示格式化成功
Re-format filesystem in Storage Directory /tmp/hadoop-hadoop/dfs/name ? (Y or N) Y
19/04/27 15:00:42 INFO namenode.FSImage: Allocated new BlockPoolId: BP-835614160-127.0.0.1-1556348442807
19/04/27 15:00:43 INFO common.Storage: Storage directory /tmp/hadoop-hadoop/dfs/name has been successfully formatted.
19/04/27 15:00:43 INFO namenode.NNStorageRetentionManager: Going to retain 1 images with txid >= 0
19/04/27 15:00:43 INFO util.ExitUtil: Exiting with status 0
19/04/27 15:00:43 INFO namenode.NameNode: SHUTDOWN_MSG: 
/************************************************************
SHUTDOWN_MSG: Shutting down NameNode at hadoop001/127.0.0.1
************************************************************/
[hadoop@hadoop001 hadoop]$ 

 
 

 
 
1
2
3
4
5
6
7
8
9
10
11
12
13

启动hadoop

[hadoop@hadoop001 sbin]$ ./start-all.sh 
This script is Deprecated. Instead use start-dfs.sh and start-yarn.sh
Starting namenodes on [hadoop001]
hadoop001: starting namenode, logging to /home/hadoop/app/hadoop-2.6.0-cdh5.7.0/logs/hadoop-hadoop-namenode-hadoop001.out
hadoop001: starting datanode, logging to /home/hadoop/app/hadoop-2.6.0-cdh5.7.0/logs/hadoop-hadoop-datanode-hadoop001.out
Starting secondary namenodes [hadoop001]
hadoop001: starting secondarynamenode, logging to /home/hadoop/app/hadoop-2.6.0-cdh5.7.0/logs/hadoop-hadoop-secondarynamenode-hadoop001.out
starting yarn daemons
starting resourcemanager, logging to /home/hadoop/app/hadoop-2.6.0-cdh5.7.0/logs/yarn-hadoop-resourcemanager-hadoop001.out
hadoop001: starting nodemanager, logging to /home/hadoop/app/hadoop-2.6.0-cdh5.7.0/logs/yarn-hadoop-nodemanager-hadoop001.out
#查询出五个守护进程
[hadoop@hadoop001 sbin]$ jps
868 ResourceManager
456 NameNode
1353 DataNode
1419 Jps
720 SecondaryNameNode
961 NodeManager
[hadoop@hadoop001 sbin]$ 

 
 

 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

验证hadoop

hdfs验证:

[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ hdfs dfs -ls /
[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ hdfs dfs -put README.txt hdfs://hadoop001:9000/
[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ hdfs dfs -ls /
Found 1 items
-rw-r--r--   1 hadoop supergroup       1366 2019-04-27 15:39 /README.txt
[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ 

 
 

 
 
1
2
3
4
5
6

yarn验证:

#用这个案例验证一下，看一下结果
[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ hadoop jar ./share/hadoop/mapreduce/hadoop-mapreduce-examples-2.6.0-cdh5.7.0.jar pi 4 10

 
 

 
 
1
2

检测是否支持压缩

[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ hadoop checknative
19/04/27 16:04:13 INFO bzip2.Bzip2Factory: Successfully loaded & initialized native-bzip2 library system-native
19/04/27 16:04:13 INFO zlib.ZlibFactory: Successfully loaded & initialized native-zlib library
Native library checking:
hadoop:  true /home/hadoop/app/hadoop-2.6.0-cdh5.7.0/lib/native/libhadoop.so.1.0.0
zlib:    true /lib64/libz.so.1
snappy:  true /usr/lib64/libsnappy.so.1
lz4:     true revision:99
bzip2:   true /lib64/libbz2.so.1
openssl: true /usr/lib64/libcrypto.so
[hadoop@hadoop001 hadoop-2.6.0-cdh5.7.0]$ 

 
 

 
 
1
2
3
4
5
6
7
8
9
10
11

本文参考：
①主要参考博客：https://blog.csdn.net/qq_32641659/article/details/89074365#commentBox
②关于Hadoop的本地库参考：
http://blog.sina.com.cn/s/blog_3d9e90ad0102wqrp.html
③Protocol Buffer介绍
https://blog.csdn.net/carson_ho/article/details/70568606
④./configure、make、make install 命令详解：
https://www.cnblogs.com/tinywan/p/7230039.html
https://www.cnblogs.com/qiaopei/p/5550168.html