集群规划
安装Zookeeper
- 解压:tar -zxvf zookeeper-3.4.14.tar.gz
- 配置环境变量:
vim /etc/profile
export ZOOKEEPER=/opt/zookeeper-3.4.14
export PATH=$ZOOKEEPER/bin:$PATH
source /etc/profile
监测配置输入zk,按下Tab,会有提示信息
- 配置zookeeper
如果不存在则复制配置文件
cp zoo_sample.cfg zoo.cfg
配置
vim /opt/zookeeper-3.4.14/bin/zoo.cfg
dataDir=/opt/zookeeper-3.4.14/tmp
dataLogDir=/opt/zookeeper-3.4.14/tmpLog
最后追加,.1.2.3是zookeeper选取leader的顺序,还有一个默认顺序数据多的作为leader,2888,3888外部传递接口和 zookeeper集群通信接口
server.1=192.168.1.191:2888:3888
server.2=192.168.1.192:2888:3888
server.3=192.168.1.193:2888:3888
配置完成后创建目录:
mkdir /opt/zookeeper-3.4.14/tmp
mkdir /opt/zookeeper-3.4.14/tmpLog
在目录中追加:
192.168.1.191追加:
echo 1 > /opt/zookeeper-3.4.14/tmp/myid
192.168.1.192追加
echo 2 > /opt/zookeeper-3.4.14/tmp/myid
192.168.1.193追加:
echo 3 > /opt/zookeeper-3.4.14/tmp/myid
分发到其他机器:
scp -r /opt/zookeeper-3.4.14 node03:/opt/
注意:分发到其他机器不能忘记配置环境变量
- 每台都启动
192.168.1.191:zkServer.sh start
192.168.1.192:zkServer.sh start
192.168.1.193:zkServer.sh start
免密钥
配置node02到node01免密钥,同时node01到node02,node03,node0都要免密钥配置相同
配置node02到node01免密钥
生成密钥:ssh-keygen -t rsa -P '' -f ~/.ssh/id_rsa
追加给自己:cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
添加权限:chmod 0600 ~/.ssh/authorized_keys
将公钥给其他机器:scp id_rsa.pub node01:/root/.ssh/node02.pub
其他机器追加公钥:cat ~/.ssh/node02.pub >> ~/.ssh/authorized_keys
配置Hadoop的HA
配置Hadoop的HA如果没有使用过Hadoop看上篇Hadoop集群配置
(注意:需要安装psmisc:yum install psmisc)
- 拷贝上一次,保存副本,为轻量级没有zk的hadoop:cp -r hadoop-3.1.2 hadoop-3.1.2-full
- 更改hadoop配置
hdfs-site.xml ,注意HA模式没有Secondary
<property>
<name>dfs.replication</name>
<value>2</value>
</property>
<property>
<name>dfs.name.dir</name>
<value>/opt/hadoop-3.1.2/hatmp/dfs/name</value>
</property>
<property>
<name>dfs.data.dir</name>
<value>/opt/hadoop-3.1.2/hatmp/dfs/data</value>
</property>
<property>
<name>dfs.permissions</name>
<value>false</value>
</property>
<!--zk配置-->
<property>
<name>dfs.nameservices</name>
<value>mycluster</value>
</property>
<property>
<name>dfs.ha.namenodes.mycluster</name>
<value>nn1,nn2</value>
</property>
<property>
<name>dfs.namenode.rpc-address.mycluster.nn1</name>
<value>node01:8020</value>
</property>
<property>
<name>dfs.namenode.rpc-address.mycluster.nn2</name>
<value>node02:8020</value>
</property>
<property>
<name>dfs.namenode.http-address.mycluster.nn1</name>
<value>node01:50070</value>
</property>
<property>
<name>dfs.namenode.http-address.mycluster.nn2</name>
<value>node02:50070</value>
</property>
<property>
<name>dfs.namenode.shared.edits.dir</name>
<value>qjournal://node01:8485;node02:8485;node03:8485/mycluster</value>
</property>
<property>
<name>dfs.journalnode.edits.dir</name>
<value>/opt/hadoop-3.1.2/ha/jn</value>
</property>
<property>
<name>dfs.client.failover.proxy.provider.mycluster</name>
<value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
</property>
<property>
<name>dfs.ha.fencing.methods</name>
<value>sshfence</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>/root/.ssh/id_rsa</value>
</property>
<property>
<name>dfs.ha.automatic-failover.enabled</name>
<value>true</value>
</property>
- core-site.xml的配置
<property>
<name>ha.zookeeper.quorum</name>
<value>node02:2181,node03:2181,node04:2181</value>
</property>
<property>
<name>fs.defaultFS</name>
<value>hdfs://mycluster</value>
</property>
<property>
<name>hadoop.tmp.dir</name>
<value>/opt/hadoop-3.1.2/hatmp</value>
</property>
- 将配置文件分发
scp core-site.xml hdfs-site.xml node02:/opt/hadoop-3.1.2/etc/hadoop
scp core-site.xml hdfs-site.xml node03:/opt/hadoop-3.1.2/etc/hadoop
scp core-site.xml hdfs-site.xml node04:/opt/hadoop-3.1.2/etc/hadoop
- 上面启动了zk,下面启动journalnode:hadoop-daemon.sh start journalnode
node01:hadoop-daemon.sh start journalnode
node02:hadoop-daemon.sh start journalnode
node03:hadoop-daemon.sh start journalnode - 下面格式化:hdfs namenode -format
- 启动namenode:hadoop-daemon.sh start namenode
- 将格式化复制到下面几台的namenode,也就是node02上使用命令复制:hdfs namenode -bootstrapStandby
- 使用zkfc格式化zookeeper:hdfs zkfc -formatZK
- 最后启动hadoop:start-dfs.sh
配置yarn
vim yarn-site.xml追加如下
<property>
<name>yarn.resourcemanager.ha.enabled</name>
<value>true</value>
</property>
<!--该cluster-id不能与nameService相同-->
<property>
<name>yarn.resourcemanager.cluster-id</name>
<value>myclusteryarn</value>
</property>
<!--指定2台Resource Manager (即Name Node )节点-->
<property>
<name>yarn.resourcemanager.ha.rm-ids</name>
<value>rm1,rm2</value>
</property>
<property>
<name>yarn.resourcemanager.hostname.rm1</name>
<value>node01</value>
</property>
<property>
<name>yarn.resourcemanager.hostname.rm2</name>
<value>node02</value>
</property>
<!--指定zookeeper 节点-->
<property>
<name>yarn.resourcemanager.zk-address</name>
<value>node02:2181,node03:2181,node04:2181</value>
</property>
<!--启用自动恢复,当任务进行一半,rm坏掉,就要启动自动恢复,默认是false-->
<property>
<name>yarn.resourcemanager.recovery.enabled</name>
<value>true</value>
</property>
<!--指定resourcemanager的状态信息存储在zookeeper集群,默认是存放在FileSystem里面。-->
<property>
<name>yarn.resourcemanager.store.class</name>
<value>org.apache.hadoop.yarn.server.resourcemanager.recovery.ZKRMStateStore</value>
</property>
手动启动:start-yarn.sh
关闭:stop-yarn-sh
附录
rm -rf /opt/hadoop-3.1.2/ha
rm -rf /opt/hadoop-3.1.2/hatmp
rm -rf /opt/hadoop-3.1.2/tmp
rm -rf /opt/zookeeper-3.4.14/tmp/version-2/
rm -rf /opt/zookeeper-3.4.14/tmp/zookeeper_server.pid
rm -rf /opt/hadoop-3.1.2/tmp/dfs/name/
rm -rf /opt/hadoop-3.1.2/tmp/dfs/data/
mkdir /opt/hadoop-3.1.2/tmp/dfs/name/
mkdir /opt/hadoop-3.1.2/tmp/dfs/data/
rm -rf /opt/hadoop-3.1.2/hatmp/dfs/name/
rm -rf /opt/hadoop-3.1.2/hatmp/dfs/data/
mkdir /opt/hadoop-3.1.2/hatmp/dfs/name/
mkdir /opt/hadoop-3.1.2/hatmp/dfs/data/