1. 启动三个主机镜像

容器规划

	hadoop-namenode	hadoop-second-namenode	hadoop-datanode
hdfs	NameNode/DataNode	SecondaryNameNode/DataNode	DataNode
yarn	NodeManager	NodeManager	ResourceManager

• 创建hdfs的网卡，便于容器间通信

#创建网卡
docker network create hdfs

#查看网卡
docker network ls

• 启动namenode

docker run -d -it -p 37764 -p 50090 -p 38832 -p 44026 -p 50020 -p 50475 -p 50075 -p 50010 -p 50470 -p 50070 -p 9000 --network hdfs --hostname hadoop-namenode --name hadoop-namenode centos:6.6 /bin/bash

• 启动second namenode(其兼职datanode)

docker run -d -it -p 37764 -p 50090 -p 38832 -p 44026 -p 50020 -p 50475 -p 50075 -p 50010 -p 50470 -p 50070 -p 9000 --network hdfs --hostname hadoop-second-namenode --name hadoop-second-namenode centos:6.6 /bin/bash

• 启动datanode

docker run -d -it -p 37764 -p 50090 -p 38832 -p 44026 -p 50020 -p 50475 -p 50075 -p 50010 -p 50470 -p 50070 -p 9000 --network hdfs --hostname hadoop-datanode --name hadoop-datanode centos:6.6 /bin/bash

2. ssh免密登录

• 安装三个容器的ssh客户端和服务端, 以namenode为例

# 安装ssh服务依赖
yum install -y openssh-server
# 安装ssh客户端依赖
yum install -y openssh-clients
# 设置ssh密码
echo 'root:123456' | chpasswd
sed -i 's/PermitRootLogin prohibit-password/PermitRootLogin yes/' /etc/ssh/sshd_config
sed 's@session\s*required\s*pam_loginuid.so@session optional pam_loginuid.so@g' -i /etc/pam.d/sshd
echo "export VISIBLE=now" >> /etc/profile
# 启动ssh服务
/etc/init.d/sshd start

• 登录namenode

# 进入容器内部
docker exec -it hadoop-namenode /bin/bash

# 生成共钥和私钥
ssh-keygen -t rsa 

# 将共钥拷贝到second namenode和datanode
ssh-copy-id localhost
ssh-copy-id 0.0.0.0
ssh-copy-id hadoop-namenode
ssh-copy-id hadoop-second-namenode
ssh-copy-id hadoop-datanode

• 登录hadoop-second-namenode

# 进入容器内部
docker exec -it hadoop-second-namenode /bin/bash

# 生成共钥和私钥
ssh-keygen -t rsa 

# 将共钥拷贝到namenode和datanode
ssh-copy-id localhost
ssh-copy-id hadoop-namenode
ssh-copy-id hadoop-second-namenode
ssh-copy-id hadoop-datanode

• 登录hadoop-datanode

# 进入容器内部
docker exec -it hadoop-datanode /bin/bash

# 生成共钥和私钥
ssh-keygen -t rsa 

# 将共钥拷贝到namenode和second namenode
ssh-copy-id localhost
ssh-copy-id hadoop-namenode
ssh-copy-id hadoop-second-namenode
ssh-copy-id hadoop-datanode

3. 下载hadoop安装包并进行配置

• 下载hadoop二进制包 我选的版本是2.8.5的

cd /home
yum install -y tar wget which
wget http://mirror.bit.edu.cn/apache/hadoop/common/hadoop-2.8.5/hadoop-2.8.5.tar.gz
tar -zxvf hadoop-2.8.5.tar.gz
rm -rf hadoop-2.8.5.tar.gz 

• yum安装jdk

yum -y install java-1.8.0-openjdk
yum -y install java-1.8.0-openjdk-devel

• 配置环境变量 ~/.bashrc
  JAVA_HOME根据实际版本填写，因为我安装的版本可能和你不一样。

export JAVA_HOME=/usr/lib/jvm/java-1.8.0-openjdk-1.8.0.242.b07-1.el6_10.x86_64
export CLASSPATH=$CLASSPATH:$JAVA_HOME/lib
export PATH=$JAVA_HOME/bin:$PATH:$HOME/bin
export HADOOP_HOME=/home/hadoop-2.8.5
export PATH=$HADOOP_HOME/sbin:$HADOOP_HOME/bin:$PATH
export CLASSPATH=$HADOOP_HOME/lib:$CLASSPATH

source ~/.bashrc

• 修改core-site.xml

hadoop-namenode这台机器fs.defaultFS写0.0.0.0:9000其他节点写hadoop-namenode:9000!!!

<configuration>
<property>
    <name>fs.defaultFS</name>
    <value>hdfs://0.0.0.0:9000</value>
</property>
<property>
    <name>hadoop.tmp.dir</name>
    <value>/tmp/hadoop-2.8.5</value>
</property>
<property>
    <name>hadoop.proxyuser.hadoop.hosts</name>
    <value>*</value>
</property>
<property>
    <name>hadoop.proxyuser.hadoop.groups</name>
    <value>*</value>
</property>
</configuration>

• 修改hdfs-site.xml

<configuration>
    <property>
        <name>dfs.replication</name>
        <value>3</value>
    </property>
    <property>
        <name>dfs.namenode.secondary.http-address</name>
        <value>hadoop-second-namenode:50090</value>
    </property>
    <property>
        <name>dfs.datanode.data.dir</name>
        <value>/tmp/hadoop-2.8.5</value>
    </property>
</configuration>

• 修改slaves

hadoop-namenode
hadoop-second-namenode
hadoop-datanode

• 修改yarn-site.xml

<configuration>
<property>
    <name>yarn.nodemanager.aux-services</name>
    <value>mapreduce_shuffle</value>
</property>
<property>
    <name>yarn.resourcemanager.hostname</name>
    <value>hadoop-datanode</value>
</property>
</configuration>

• 修改mapred-site.xml

<configuration>
    <property>
        <name>mapreduce.framework.name</name>
    <value>yarn</value>
    </property>
</configuration>

4. 启动集群

进入hadoop-namenode容器

hdfs namenode -format
start-dfs.sh

然后查看hadoop-namenode端口情况比如我的

570b3fd120ae        centos:6.6          "/bin/bash"         40 minutes ago      Up 40 minutes       0.0.0.0:32945->9000/tcp, 0.0.0.0:32944->37764/tcp, 0.0.0.0:32943->38832/tcp, 0.0.0.0:32942->44026/tcp, 0.0.0.0:32941->50010/tcp, 0.0.0.0:32940->50020/tcp, 0.0.0.0:32939->50070/tcp, 0.0.0.0:32938->50075/tcp, 0.0.0.0:32937->50090/tcp, 0.0.0.0:32936->50470/tcp, 0.0.0.0:32935->50475/tcp   hadoop-namenode

所以通过http://ip:32939 访问namenode

5. 坑记录

• docker中环境变量要写在~/.bashrc， 写在/etc/profile和～/.bash_profile重新进入容器是会失效的
• docker容器重新启动后，/etc/init.d/sshd start 需要重新执行别忘了。