搭建CDH 阿里云 (Step 1: 启动hdfs)
准备动作
- 在阿里云申请三台机器(Centos)
- 安装Oracle JDK
在阿里云申请三台机器(Centos)
PS. 内存要大于4g，因为namenode启动的时候内存需要
在每一台机器上配置域名
vim /etc/hosts
172.24.218.96 worker1
172.24.218.97 worker2
172.24.218.98 worker3
跟着cloudera官网文档走
安装CDH有两种方式:
- Cloudera Mangaer
- Command Line
之后说的第二种方式
安装CDH好处
用Service的方式启动组件，而不用脚本启动
安装CDH前 先安装Oracle JDK
文档地址
JDK 下载地址
PS. 下载x64版本的，如果选的系统是64位
下载 jdk后解压在 /usr/java/default 目录下
添加环境变量
[root@izhp333sqrlsy3kr46m4wuz java]# vim /etc/profile
export JAVA_HOME=/usr/java/default
export PATH=$PATH:$JAVA_HOME/bin
[root@izhp333sqrlsy3kr46m4wuz java]# source /etc/profile
验证 java 是否安装成功
[root@izhp333sqrlsy3kr46m4wuz java]# java -version
java version "1.8.0_162"
Java(TM) SE Runtime Environment (build 1.8.0_162-b12)
Java HotSpot(TM) 64-Bit Server VM (build 25.162-b12, mixed mode)
安装CDH
Step 1: 添加 cdh的 yum源
添加 cdh的 yum源，有两种方式(我用的第一种):
- 下载cloudera-cdh5.repo 放入 /etc/yum.repos.d/
cd /etc/yum.repos.d/
wget https://archive.cloudera.com/cdh5/redhat/7/x86_64/cdh/cloudera-cdh5.repo
sudo yum clean all
PS. 修改/etc/yum.conf的 keepcache 0->1 可以让下载的rpm文件保存
- Creating a Local Yum Repository
将cloudera-cdh5.repo 下的所有rpm包 下载到服务器，用web-service方式提供yum安装
相当于 yum 下载cdh相关 可以走内网
文档地址
Step 2: add a repository key
sudo rpm --import https://archive.cloudera.com/cdh5/redhat/7/x86_64/cdh/RPM-GPG-KEY-cloudera
Step 3: 安装Zookeeper
因为要布置 NameNode HA，所以需要 Zookeeper 环境
文档地址
布置Zk的机器上安装 (集群选奇数台且数量>=3)
sudo yum install zookeeper -y
sudo yum install zookeeper-server -y
mkdir -p /var/lib/zookeeper
chown -R zookeeper /var/lib/zookeeper/
vim /etc/zookeeper/conf/zoo.cfg
# 添加
server.1=worker1:2888:3888
server.2=worker2:2888:3888
server.3=worker3:2888:3888
# worker1
sudo service zookeeper-server init --myid=1
# worker2
sudo service zookeeper-server init --myid=2
… 以此类推
# 启动 zookeeper
sudo service zookeeper-server start
# 测试是否启动成功
zookeeper-client -server worker1:2181
Step 4: 安装cdh服务
安装Resourcemanager+namenode
选定两台master机器安装(HA)
sudo yum install hadoop-yarn-resourcemanager -y
sudo yum install hadoop-hdfs-namenode -y
除了Resourcemanager 安装 Nodemanager
PS. (因为只有3台机器，所以在Resourcemanager上也安装了Nodemanager)
一般两台主节点， 其余的都是工作节点
sudo yum install hadoop-yarn-nodemanager hadoop-hdfs-datanode hadoop-mapreduce -y
找一个 worker节点 安装
sudo yum install hadoop-mapreduce-historyserver hadoop-yarn-proxyserver -y
客户端 安装
sudo yum install hadoop-client
Step 5: 修改配置文件
vim hdfs-site.xml
<property>
<name>dfs.nameservices</name>
<value>xiwu-cluster</value>
</property>
<property>
<name>dfs.ha.namenodes.xiwu-cluster</name>
<value>nn1,nn2</value>
</property>
<property>
<name>dfs.namenode.rpc-address.xiwu-cluster.nn1</name>
<value>worker1:8020</value>
</property>
<property>
<name>dfs.namenode.rpc-address.xiwu-cluster.nn2</name>
<value>worker2:8020</value>
</property>
<property>
<name>dfs.namenode.http-address.xiwu-cluster.nn1</name>
<value>worker1:50070</value>
</property>
<property>
<name>dfs.namenode.http-address.xiwu-cluster.nn2</name>
<value>worker2:50070</value>
</property>
<property>
<name>dfs.namenode.shared.edits.dir</name>
<value>[qjournal://worker1:8485](qjournal://worker1:8485);worker2:8485;worker3:8485/xiwu-cluster</value>
</property>
<property>
<name>dfs.journalnode.edits.dir</name>
<value>/data/1/dfs/jn</value>
</property>
<property>
<name>dfs.client.failover.proxy.provider.xiwu-cluster</name>
<value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
</property>
<property>
<name>dfs.ha.fencing.methods</name>
<value>sshfence(hdfs)</value>
</property>
<property>
<name>dfs.ha.fencing.ssh.connect-timeout</name>
<value>30000</value>
</property>
<property>
<name>dfs.ha.fencing.ssh.private-key-files</name>
<value>/var/lib/hadoop-hdfs/.ssh/id_rsa</value>
</property>
<property>
<name>dfs.ha.automatic-failover.enabled</name>
<value>true</value>
</property>
<property>
<name>dfs.permissions.superusergroup</name>
<value>hadoop</value>
</property>
<property>
<name>dfs.namenode.name.dir</name>
<value>file:///data/1/dfs/nn,file:///nfsmount/dfs/nn</value>
</property>
<property>
<name>dfs.datanode.data.dir</name>
<value>file:///data/1/dfs/dn,file:///data/2/dfs/dn,file:///data/3/dfs/dn,file:///data/4/dfs/dn</value>
</property>
<property>
<name>dfs.webhdfs.enabled</name>
<value>true</value>
</property>
vim core-site.xml
<property>
<name>fs.defaultFS</name>
<value>[hdfs://xiwu-cluster](hdfs://xiwu-cluster)</value>
</property>
<property>
<name>ha.zookeeper.quorum</name>
<value>worker1:2181,worker2:2181,worker3:2181</value>
</property>
将上面的两个配置同步到所有集群机器上
配置 HA 两台机器的hdfs用户免密登陆
<property>
<name>dfs.ha.fencing.methods</name>
<value>sshfence(hdfs)</value>
</property>
先选一台master
su hdfs
ssh-keygen -t rsa
cd ~/.ssh
# pwd: /var/lib/hadoop-hdfs/.ssh
cp id_rsa.pub authorized_keys
# 将 ~/.ssh 目录 同步到另一台master上
PS. 注意目录权限 chown -R hdfs:hdfs .ssh
journalnode (选3台机器 一般是worker)
sudo mkdir -p /data/1/dfs/jn
sudo chown -R hdfs:hdfs /data/1/dfs/jn
sudo yum install hadoop-hdfs-journalnode -y
sudo service hadoop-hdfs-journalnode start -y
namenode
sudo mkdir -p /data/1/dfs/nn /nfsmount/dfs/nn
sudo chmod 700 /data/1/dfs/nn /nfsmount/dfs/nn
sudo chown -R hdfs:hdfs /data/1/dfs/nn /nfsmount/dfs/nn
# 格式化
sudo -u hdfs hdfs namenode -format
datanode
sudo mkdir -p /data/1/dfs/dn /data/2/dfs/dn /data/3/dfs/dn /data/4/dfs/dn
sudo chown -R hdfs:hdfs /data/1/dfs/dn /data/2/dfs/dn /data/3/dfs/dn /data/4/dfs/dn
zkfc (在namenode HA上控制)
sudo yum install hadoop-hdfs-zkfc -y
sudo service hadoop-hdfs-zkfc start
hdfs zkfc -formatZK
Step6: 启动服务
Primary NameNode
sudo service hadoop-hdfs-namenode start
Standby NameNode
sudo -u hdfs hdfs namenode -bootstrapStandby
sudo service hadoop-hdfs-namenode start
DataNode
sudo service hadoop-hdfs-datanode start
ResourceManager
sudo service hadoop-yarn-resourcemanager start
NodeManager
sudo service hadoop-yarn-nodemanager start
MapReduce JobHistory
sudo service hadoop-mapreduce-historyserver start
检测HDFS 是否启动成功
sudo -u hdfs hadoop fs -mkdir /tmp
sudo -u hdfs hadoop fs -chmod -R 1777 /tmp
PS. 关闭开机自动启动
chkconfig --list
chkconfig hadoop-yarn-nodemanager off
chkconfig hadoop-yarn-resourcemanager off