主机列表
-
192.168.1.11
-
192.168.1.12
-
192.168.1.13
操作系统（必须）
- CentOS6
1.设置hostname，并配置host解析
192.168.1.11主机执行：
cat >> /etc/sysconfig/network <<EOF
HOSTNAME=hadoop1
EOF
echo hadoop1 >/proc/sys/kernel/hostname
cat >> /etc/hosts <<EOF
192.168.1.11 hadoop1
192.168.1.12 hadoop2
192.168.1.13 hadoop3
EOF
192.168.1.12主机执行：
cat >> /etc/sysconfig/network <<EOF
HOSTNAME=hadoop2
EOF
echo hadoop2 >/proc/sys/kernel/hostname
cat >> /etc/hosts <<EOF
192.168.1.11 hadoop1
192.168.1.12 hadoop2
192.168.1.13 hadoop3
EOF
192.168.1.13主机执行：
cat >> /etc/sysconfig/network <<EOF
HOSTNAME=hadoop3
EOF
echo hadoop3 >/proc/sys/kernel/hostname
cat >> /etc/hosts <<EOF
192.168.1.11 hadoop1
192.168.1.12 hadoop2
192.168.1.13 hadoop3
EOF
2.安装必要软件
yum install -y openssh-clients yum-utils createrepo yum-plugin-priorities
3.实现主机互信（hadoop1节点执行）
hadoop1节点执行以下命令，生成ssh密钥
ssh-keygen -t rsa -P "" -f ~/.ssh/id_rsa
分发密钥完成互信（依次输入各节点密码）
ssh-copy-id hadoop1
ssh-copy-id hadoop2
ssh-copy-id hadoop3
4.调整系统参数（三个节点均执行）
# 调整文件句柄数
echo "* soft nofile 655350" >> /etc/security/limits.conf
echo "* hard nofile 655350" >> /etc/security/limits.conf
echo "* soft nproc 65535" >> /etc/security/limits.conf
echo "* hard nproc 65535" >> /etc/security/limits.conf
ulimit -n 655350
# 关闭防火墙
service iptables stop
chkconfig iptables off
# 关闭selinux
setenforce 0
sed -i "s#SELINUX=enforcing#SELINUX=disabled#g" /etc/selinux/config
5.安装oracle jdk1.8（所有节点）
安装oracle jdk1.8,并配置软链接
mkdir -p /usr/java
ln -s /opt/java /usr/java/jdk1.8
6.安装mysql，初始化用户
创建scm_server用户
mysql -uroot -proot <<EOF
create user 'scm_server'@'%' identified by 'scm_server';
create database scm_server_db DEFAULT CHARACTER SET utf8 COLLATE utf8_general_ci;
grant all privileges on scm_server_db.* to 'scm_server'@'%' with grant option;
create user 'hive'@'%' identified by 'hive';
create database hive DEFAULT CHARACTER SET utf8 COLLATE utf8_general_ci;
grant all privileges on hive.* to 'hive'@'%' with grant option;
create user 'oozie'@'%' identified by 'oozie';
create database oozie DEFAULT CHARACTER SET utf8 COLLATE utf8_general_ci;
grant all privileges on oozie.* to 'oozie'@'%' with grant option;
flush privileges;
exit
EOF
7.安装配置ntp
hadoop1节点执行：
yum install -y ntp
service ntpd start
chkconfig ntpd on
sed -i "s;restrict default kod nomodify notrap nopeer noquery;#restrict default kod nomodify notrap nopeer noquery;g" /etc/ntp.conf
sed -i "s;restrict -6 default kod nomodify notrap nopeer noquery;#restrict -6 default kod nomodify notrap nopeer noquery;g" /etc/ntp.conf
sed -i "s#restrict -6 ::1#restrict ::1#g" /etc/ntp.conf
sed -i "s;server 0.centos.pool.ntp.org iburst;#server 0.centos.pool.ntp.org iburst;g" /etc/ntp.conf
sed -i "s;server 1.centos.pool.ntp.org iburst;#server 1.centos.pool.ntp.org iburst;g" /etc/ntp.conf
sed -i "s;server 2.centos.pool.ntp.org iburst;#server 2.centos.pool.ntp.org iburst;g" /etc/ntp.conf
sed -i "s;server 3.centos.pool.ntp.org iburst;#server 3.centos.pool.ntp.org iburst;g" /etc/ntp.conf
echo "server 127.127.1.0" >> /etc/ntp.conf
echo "fudge 127.127.1.0 stratum 10" >> /etc/ntp.conf
echo "disable monitor" >> /etc/ntp.conf
echo "restrict default nomodify" >> /etc/ntp.conf
hadoop2、hadoop3节点执行：
yum install -y ntp
sed -i "s;restrict default kod nomodify notrap nopeer noquery;#restrict default kod nomodify notrap nopeer noquery;g" /etc/ntp.conf
sed -i "s;restrict -6 default kod nomodify notrap nopeer noquery;#restrict -6 default kod nomodify notrap nopeer noquery;g" /etc/ntp.conf
sed -i "s#restrict -6 ::1#restrict ::1#g" /etc/ntp.conf
sed -i "s;server 0.centos.pool.ntp.org iburst;#server 0.centos.pool.ntp.org iburst;g" /etc/ntp.conf
sed -i "s;server 1.centos.pool.ntp.org iburst;#server 1.centos.pool.ntp.org iburst;g" /etc/ntp.conf
sed -i "s;server 2.centos.pool.ntp.org iburst;#server 2.centos.pool.ntp.org iburst;g" /etc/ntp.conf
sed -i "s;server 3.centos.pool.ntp.org iburst;#server 3.centos.pool.ntp.org iburst;g" /etc/ntp.conf
echo "server hadoop1" >> /etc/ntp.conf
echo "disable monitor" >> /etc/ntp.conf
echo "restrict default nomodify" >> /etc/ntp.conf
ntpdate hadoop1
8.下载mysql驱动（hadoop1节点）
下载mysql-connector-java-5.1.48.tar.gz上传至hadoop1:/tmp下
执行以下命令，调整名称
mkdir -p /usr/share/java
tar zxvf mysql-connector-java-5.1.48.tar.gz
cp mysql-connector-java-5.1.48/mysql-connector-java-5.1.48.jar /usr/share/java/mysql-connector-java.jar
1.创建离线镜像源目录（所有节点）
mkdir -p /cm
下载cloudera-manager下的所有文件，并上传至/cm下
2.配置cm源（所有节点）
创建repo源文件
cd /cm && createrepo ./
cat > /etc/yum.repos.d/cm.repo <<EOF
[cloudera-manager]
name=cm
baseurl=file:///cm
gpgkey=file:///cm/RPM-GPG-KEY-cloudera
enable = 1
gpgcheck = 1
EOF
/cm目录最终结构如下：
/cm
├── cloudera-manager-agent-5.4.0-1.cm540.p0.165.el6.x86_64.rpm
├── cloudera-manager-daemons-5.4.0-1.cm540.p0.165.el6.x86_64.rpm
├── cloudera-manager-server-5.4.0-1.cm540.p0.165.el6.x86_64.rpm
├── cloudera-manager-server-db-2-5.4.0-1.cm540.p0.165.el6.x86_64.rpm
├── enterprise-debuginfo-5.4.0-1.cm540.p0.165.el6.x86_64.rpm
├── jdk-6u31-linux-amd64.rpm
├── oracle-j2sdk1.7-1.7.0+update67-1.x86_64.rpm
├── repodata
│ ├── 3a8b6a8a03c3846eadd0f0d8df2ef1142e6e32d21ce7e4e58a304ad3bef8b5b7-primary.sqlite.bz2
│ ├── 853ca50d5d1f076b5f53cd06ed4d74c62ee729af1a86e3caa1bd39aaf6e68cf7-other.sqlite.bz2
│ ├── a6b67b1228bbb6791eb66fd52cfc2044a681a9444e1a1aa044111029b6f4760c-filelists.xml.gz
│ ├── b05946fbbf3fec9e107640249a183e7109d0e336ef23fcbe199b3dd1743f84f3-other.xml.gz
│ ├── c2c48ea8c58913116c14e8ec853d2fd2731bee779edafc81dec0d60771709f17-filelists.sqlite.bz2
│ ├── fd1f07dacbe9d5e3be1e7f7930fbb6eac4d29a75c172c2a31e6e18baa56b5fee-primary.xml.gz
│ └── repomd.xml
└── RPM-GPG-KEY-cloudera
3.安装cloudera-manager-server(hadoop1节点)
yum install cloudera-manager-daemons cloudera-manager-server -y
4.安装cloudera-manager-agent(hadoop1、hadoop2、hadoop3节点)
yum install cloudera-manager-agent -y
5.修改cloudera-manager-agent配置文件(hadoop1、hadoop2、hadoop3节点)
hadoop1、hadoop2、hadoop3节点执行：
sed -i "s#server_host=localhost#server_host=hadoop1#g" /etc/cloudera-scm-agent/config.ini
echo "listening_ip=`hostname`" >> /etc/cloudera-scm-agent/config.ini
echo "listening_hostname=`hostname`" >> /etc/cloudera-scm-agent/config.ini
6.创建CDH离线源仓储
创建目录(hadoop1节点)
mkdir -p /opt/cloudera/parcel-repo
上传以下文件至hadoop1节点/opt/cloudera/parcel-repo下
调整CDH-5.4.0-1.cdh5.4.0.p0.27-el6.parcel.sha1名称
cd /opt/cloudera/parcel-repo
mv CDH-5.4.0-1.cdh5.4.0.p0.27-el6.parcel.sha1 CDH-5.4.0-1.cdh5.4.0.p0.27-el6.parcel.sha
7.初始化CM Server数据库(hadoop1节点)
/usr/share/cmf/schema/scm_prepare_database.sh mysql scm_server_db scm_server scm_server -h 127.0.0.1
8.启动cloudera-manager-agent（hadoop1、hadoop2、hadoop3节点）
service cloudera-scm-agent start
chkconfig cloudera-scm-agent on
9.启动cloudera-manager-server(hadoop1节点)
service cloudera-scm-server start
chkconfig cloudera-scm-server on
查看日志
tail -200f /var/log/cloudera-scm-server/cloudera-scm-server.log
10.访问控制台初始化
主要hadoop1替换为实际IP地址
访问http://hadoop1:7180，账号密码: admin/admin
部署免费版本
确认部署应用，点击继续
添加部署节点，点击继续
确认cdh版本，继续
确认安装完成，点击继续
确认检测结果，点击完成
选取hbase内核hadoop安装
默认角色分配，继续
输入用户名密码数据库实例名，测试连接后点击继续
hive/hive hive oozie/oozie oozie
确认审核设置，点击继续
安装完成
主机列表
-
192.168.1.11
-
192.168.1.12
-
192.168.1.13
1.下载安装介质及脚本
互联网下载地址:
2.创建/trafodion目录，上传安装介质及脚本至该目录下
目录结构如下：
/trafodion/
├── apache-trafodion_installer-2.1.0-incubating.tar.gz
└── apache-trafodion_server-2.1.0-RH6-x86_64-incubating.tar.gz
3.解压运行安装脚本
cd /trafodion
tar zxvf apache-trafodion_pyinstaller-2.1.0-incubating.tar.gz
cd python-installer
./db_install.py
4.按提示输入相关信息
Enter HDP/CDH web manager URL:port, (full URL, if no http/https prefix, default prefix is http://):
-- 输入 http://192.168.1.11:7180
Enter HDP/CDH web manager user name [admin]:
-- 回车默认
Enter HDP/CDH web manager user password:
-- 输入 admin
Confirm Enter HDP/CDH web manager user password:
-- 输入 admin
Enter full path to Trafodion tar file:
-- 输入 /trafodion/apache-trafodion_server-2.1.0-RH6-x86_64-incubating.tar.gz
Enter directory name to install trafodion to [apache-trafodion-2.1.0]:
-- 回车默认
Enter trafodion user password:
-- 输入 trafodion
Enter number of DCS client connections per node [4]:
-- 回车默认
Enter trafodion scratch file folder location(should be a large disk),
if more than one folder, use comma seperated [$TRAF_HOME/tmp]:
-- 回车默认
Start instance after installation (Y/N) [Y]:
-- 回车默认
Enable LDAP security (Y/N) [N]:
-- 回车默认
Enable DCS High Avalability (Y/N) [N]:
-- 回车默认
Enter Hadoop admin password, default is [admin]:
-- 回车默认
Confirm result (Y/N) [N]:
-- 输入 Y
安装过程
安装完毕
5.查看trafodion状态
登录hadoop2，切换到trafodion用户执行以下语句：
sqcheck
返回如下，说明成功
*** Checking Trafodion Environment ***
Checking if processes are up.
Checking attempt: 1; user specified max: 2. Execution time in seconds: 0.
The Trafodion environment is up!
Process Configured Actual Down
------- ---------- ------ ----
DTM 2 2
RMS 4 4
DcsMaster 1 1
DcsServer 2 2
mxosrvr 8 8
RestServer 1 1