精品-大数据CDH集群安装部署
一、系统环境
操作系统:CentOS 7.2 x64
Clouder Manager:5.14.4
CDH:5.14.4
二、机器规划
准备4个节点,2个master节点,3个datanode节点, 3个zk节点。
master:8核16G ;10.16.30.212; bigdataop-test-30-212
datanode:8核16G; 10.16.30.212-215; bigdataop-test-30-212/215
三、集群基础准备工作(非常重要*****)
1、配置hosts
1)hostname建议小写,如果开启kerberos大写认证报错 2)hosts配置为集群所有主机映射
[root@localhost ~]# echo "hostname" > /etc/hostname
2、免秘钥登录(略...)
在规划的namenode节点中,使用ops账号建立免秘钥,且有权限执行sudo -i切换到root账户
ssh-keygen #建立公钥和私钥
ssh-copy-id + IP #将公钥copy到集群其他节点
3、ntp时钟同步(略...) *** hadoop强依赖
[root@localhost ~]# cat /etc/ntp.conf
driftfile /var/lib/ntp/drift
restrict default kod nomodify notrap nopeer noquery
restrict -6 default kod nomodify notrap nopeer noquery
restrict 127.0.0.1
restrict -6 ::1
server 192.168.10.100
includefile /etc/ntp/crypto/pw
keys /etc/ntp/keys
[root@localhost ~]# ntpdate 192.168.10.100
[root@localhost ~]# systemctl enable ntpd
[root@localhost ~]# systemctl restart ntpd
4、安装系统依赖库
[root@localhost ~]# yum -y install bind-utils libxslt cyrus-sasl-plain cyrus-sasl-gssapi portmap fuse-libs /lib/lsb/init-functions httpd mod_ssl openssl-devel python-psycopg2 MySQL-python fuse
5、关闭防火墙
[root@localhost ~]# systemctl stop firewalld
[root@localhost ~]# systemctl disabled firewalld
6、关闭selinux(略...)
7、安装jdk8(略...)
8、关闭swap
(1) 第一步 关闭swap分区:
[root@localhost ~]# swapoff -a
(2) 第二步修改配置文件 - /etc/fstab
删除swap相关行 /mnt/swap swap swap defaults 0 0 这一行或者注释掉这一行
(3) 第三步确认swap已经关闭
[root@localhost ~]# free -m
(4) 第四步调整 swappiness 参数
[root@localhost ~]# echo 0 > /proc/sys/vm/swappiness # 临时生效
[root@localhost ~]# vim /etc/sysctl.conf # 永久生效
#修改 vm.swappiness 的修改为 0
vm.swappiness=0
[root@localhost ~]# sysctl -p # 使配置生效
9、透明大叶修改
[root@localhost ~]# echo 'never' >/sys/kernel/mm/transparent_hugepage/enabled
[root@localhost ~]# echo 'never' >/sys/kernel/mm/transparent_hugepage/defrag
[root@localhost ~]# echo "echo 'never' > /sys/kernel/mm/transparent_hugepage/enabled" >> /etc/rc.d/rc.local
[root@localhost ~]# echo "echo 'never' > /sys/kernel/mm/transparent_hugepage/defrag" >> /etc/rc.d/rc.local
10、磁盘格式化(可选项,根据服务器操作)
#!/bin/bash
i=1
for disk in `ls -l /dev/sd* |awk '{print substr($10,0,8)}'|grep -v sda|uniq`
do
parted ${disk} << EXIT
mklabel gpt
mkpart primary 0 -1
ignore
quit
EXIT
mkfs.ext4 ${disk}1
echo "****************${disk}1 Fdisked!Waithing For 10 second***********"
sleep 1s
mkdir -p /mnt/hdfs/$i
echo "${disk}1 /data/${i} ext4 defaults,noatime 0 0 ">>/etc/fstab
i=$(($i + 1))
done
cat /etc/fstab
mount -a
四、安装说明
注意:离线安装,要先下载对应的CM和对应的CDH
cm和cdh的下载地址:
http://archive.cloudera.com ## url后面根据cm/cdh
五、构建本地yum源:
5.1 安装web服务
略搭建本地http或nginx服务...
5.2 下载安装包
Cloudera Manager安装包地址:
http://archive.cloudera.com/cm5/redhat/7/x86_64/cm/5.14.0/RPMS/x86_64/cloudera-manager-agent-5.14.0-1.cm5140.p0.25.el7.x86_64.rpm
http://archive.cloudera.com/cm5/redhat/7/x86_64/cm/5.14.0/RPMS/x86_64/cloudera-manager-daemons-5.14.0-1.cm5140.p0.25.el7.x86_64.rpm
http://archive.cloudera.com/cm5/redhat/7/x86_64/cm/5.14.0/RPMS/x86_64/cloudera-manager-server-5.14.0-1.cm5140.p0.25.el7.x86_64.rpm
CDH安装包地址:
http://archive.cloudera.com/cdh5/parcels/5.14.4/CDH-5.14.4-1.cdh5.14.4.p0.3-el7.parcel
http://archive.cloudera.com/cdh5/parcels/5.14.4/CDH-5.14.4-1.cdh5.14.4.p0.3-el7.parcel.sha1
http://archive.cloudera.com/cdh5/parcels/5.14.4/manifest.json
5.3 配置repo
yum install -y yum-utils createrepo
mkdir /var/www/html/cdh/
mv cloudera* /var/www/html/cm/ && mv CDH* /var/www/html/cdh/ ## 移动rpm文件夹到web server下
[root@localhost packages]# cd /var/www/html/cdh/5
[root@localhost cdh]# createrepo . #创建repo
[root@localhost cm]# createrepo .
[root@localhost ~]# systemctl start nginx
server {
listen 80;
server_name localhost;
access_log /var/log/nginx_access.log main;
error_log /var/log/nginx_error.log warn;
root /var/www/html;
location / {
autoindex on;
}
}
5.4 所有节点添加yum源的配置 ***
[root@localhost ~]# vim /etc/yum.repos.d/cm.repo
[cloudera-manager]
name = Cloudera Manager, Version 5.14.0
baseurl = http://10.193.xx.xx/cdh/
gpgcheck = 0
5.5 验证源:yum list|grep cloudera
如果验证失败更新一下yum源
1)删除缓存rm -rf /var/cache/yum
2) yum makecache
3) yum clean all
六、集群安装部署
6.1 管理节点上安装cloudera-manager-server、cloudera-manager-daemons、cloudera-manager-agent
[root@localhost ~]# yum -y install cloudera-manager-server cloudera-manager-daemons cloudera-manager-agent
(1)安装完后再/opt/cloudera/parcel-repo目录下下载如下三个文件并授权cloudera-scm属主属组:
[root@localhost ~]# cd /opt/cloudera/parcel-repo
CDH-5.14.4-1.cdh5.14.4.p0.3-el7.parcel
CDH-5.14.4-1.cdh5.14.4.p0.3-el7.parcel.sha
manifest.json
(2)配置server
[root@localhost ~]# vim /etc/cloudera-scm-server/db.properties
com.cloudera.cmf.db.type=mysql
com.cloudera.cmf.db.host=xx.xx.xx.xx #数据库IP地址
com.cloudera.cmf.db.name=scm
com.cloudera.cmf.db.user=scm
com.cloudera.cmf.db.password=123456
com.cloudera.cmf.db.setupType=EXTERNAL
(3)启动server
systemctl start cloudera-scm-server
6.2 安装MySQL服务(用来保存集群信息很重要,后期也会保存hive,hue等等)
安装部署略...
创建目录 mkdir /usr/share/java/
下载mysql驱动器并放入到/usr/share/java/mysql-connector-java.jar
6.3 Agent节点上安装cloudera-manager-daemons、cloudera-manager-agent
[root@localhost ~]# yum -y install cloudera-manager-daemons cloudera-manager-agent
(1)配置agent
vim /etc/cloudera-scm-agent/config.ini
server_host=10.193.xx.xx ## 指向管理节点
(2)启动agent
systemctl start cloudera-scm-agent
6.4 访问验证
管理节点ip:7108

浙公网安备 33010602011771号