JAVA环境配置

  1. $ java -version
  2. java version "1.8.0_121"
  3. Java(TM) SE Runtime Environment (build 1.8.0_121-b13)
  4. Java HotSpot(TM) 64-Bit Server VM (build 25.121-b13, mixed mode)

mac查看Java的安装位置信息

  1. $  /usr/libexec/java_home
  2. /Library/Java/JavaVirtualMachines/jdk1.8.0_121.jdk/Contents/Home

SSH配置

文件和目录的权限千万别设置成chmod 777.这个权限太大了，不安全

  1. cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
  2. chmod 0600 ~/.ssh/authorized_keys

如果没有ssh公钥,执行下面命令

  1. sudo ssh-keygen -t rsa -C"baxiang_aliyun"

开启远程登录
Mac的Hadoop开发环境搭建 - 图1
测试远程登录是否开启

  1. ssh localhost

安装hadoop

最终的安装目录在/usr/local/Cellar/hadoop/ 安装的版本是3.1.2

  1. $ brew install hadoop
  2. Updating Homebrew...
  3. ==> Downloading https://www.apache.org/dyn/closer.cgi?path=hadoop/common/hadoop-3.1.2/hadoop-3.1.2.tar.gz
  4. ==> Downloading from http://mirrors.tuna.tsinghua.edu.cn/apache/hadoop/common/hadoop-3.1.2/hadoop-3.1.2.tar.gz
  5. ######################################################################## 100.0%
  6. 🍺  /usr/local/Cellar/hadoop/3.1.2: 21,686 files, 774.1MB, built in 10 minutes 1 second

配置

需要修改配置文件都在/usr/local/Cellar/hadoop/3.1.2/libexec/etc/hadoop这个目录下

  1. $ vim hadoop-env.sh
  2. $ vim core-site.xml
  3. $ vim hdfs-site.xml

hadoop-env.sh

配置JAVA_HOME
Mac的Hadoop开发环境搭建 - 图2

将/usr/libexec/java_home查到的 Java 路径，记得去掉注释 #。

  1. export JAVA_HOME=/Library/Java/JavaVirtualMachines/jdk1.8.0_121.jdk/Contents/Home

core-site.xml

修改core-site.xml 文件参数,配置NameNode的主机名和端口号

  1. <configuration>
  2.     <property>
  3.         <name>hadoop.tmp.dir</name>
  4.         <value>/usr/local/Cellar/hadoop/hdfs/tmp</value>
  5.         <description>A base for other temporary directories</description>
  6.     </property>
  7.     <property>
  8.         <name>fs.default.name</name>
  9.         <value>hdfs://localhost:9000</value>
  10.     </property>
  11. </configuration>

hdfs-site.xml

变量dfs.replication指定了每个HDFS数据库的复制次数。 通常为3, 由于我们只有一台主机和一个伪分布式模式的DataNode，将此值修改为1

<configuration>
 <property>
 <name>dfs.replication</name>
 <value>1</value>
 </property>
</configuration>

格式化

格式化hdfs操作只要第一次才使用，否则会造成数据全部丢失
hdfs namenode -format
Mac的Hadoop开发环境搭建 - 图3

启动 NameNode 和 DataNode：

$ cd /usr/local/Cellar/hadoop/3.1.2/sbin
$ ./start-dfs.sh
Starting namenodes on [localhost]
Starting datanodes
Starting secondary namenodes [baxiangs-Mac-mini.local]
baxiangs-Mac-mini.local: Warning: Permanently added 'baxiangs-mac-mini.local,192.168.1.115' (ECDSA) to the list of known hosts.
2019-08-04 01:25:14,753 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable

http://localhost:9870/
Mac的Hadoop开发环境搭建 - 图4

$ jps
5665 DataNode
5890 Jps
5797 SecondaryNameNode
1274 Launcher
317
5566 NameNode

YARN服务

./start-yarn.sh

关闭YARN服务

./stop-yarn.sh

启动成功后，我们在浏览器中输入http://localhost:8088/cluster
Mac的Hadoop开发环境搭建 - 图5

启动/关闭Hadoop服务(等效上面两个)
./start-all.sh
./stop-all.sh

执行wordcount

创建文件

$ hadoop fs -mkdir  /test
$ echo "hello hdfs" >> demo.txt
$ hadoop fs -put demo.txt /test/demo.txt
$ $ hadoop fs -cat /test/demo.txt

执行mr统计wordcount

$ hadoop jar /usr/local/Cellar/hadoop/3.1.2/libexec/share/hadoop/mapreduce/sources/hadoop-mapreduce-examples-3.1.2-sources.jar org.apache.hadoop.examples.WordCount /test/demo.txt /test/out/word-out

查看执行结果

$hadoop fs -cat  /test/out/word-out/part-r-00000
2019-08-07 22:46:46,974 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
hdfs    1
hello    1

安装spark

=> Downloading https://www.apache.org/dyn/closer.lua?path=spark/spark-2.4.3/spark-2.4.3-bin-hadoop2.7.tgz
==> Downloading from http://45.252.224.79/files/623300000DD89759/mirror.bit.edu.cn/apache/spark/spark-2.4.3/spark-2.4.3-bin-hadoop2.7.tgz
######################################################################## 100.0%
🍺  /usr/local/Cellar/apache-spark/2.4.3: 1,059 files, 248.4MB, built in 29 seconds