CDH Hadoop基于Cloudera Manager半在线方式安装详解
1 CM简介
Cloudera Manager是一个端到端的管理CDH集群的工具。它通过管理CDH集群（组件安装、服务管理、参数监控、问题诊断、性能监控）各个环节从而增强集群管理的质量, 增强规范性,降低运营成本。
注:CM自6.3.3之后不再支持社区免费版(Express),后续版本需付费才能下载和使用。
2 准备
2.1 环境说明
1 该安装方式依赖如下linux组件:httpd, mysql
2 安装时需要用到其它系统组件,请配置好本地源
3 安装mysql是请确保清理好系统自带的mysql.
4 CM 下载地址见下,需要下载如下的所有文件.详细目录分布见2.2.6说明.
http://archive-primary.cloudera.com/cm5/redhat/6/x86_64/cm/5.5.1/RPMS/x86_64/
http://archive-primary.cloudera.com/cm5/redhat/6/x86_64/cm/5.5.1/repodata/
http://archive-primary.cloudera.com/cm5/redhat/6/x86_64/cm/RPM-GPG-KEY-cloudera
http://archive-primary.cloudera.com/cm5/redhat/6/x86_64/cm/cloudera-manager.repo
5 CDH parcel下载地址http://archive-primary.cloudera.com/cdh5/parcels/5.5.1/
6 CM installer下载地址
http://archive-primary.cloudera.com/cm5/installer/5.5.1/cloudera-manager-installer.bin
注:
1 CM 核心安装文件截图:
2 CDH parcel相关文件截图:
3 CM installer截图:
2.1.1 虚拟机里Linux系统版本
[root@localhost ranger-0.5.0-usersync]# cat /etc/issue | grep Linux
Red Hat Enterprise Linux Server release 6.5 (Santiago)
2.1.2 JDK版本
[root@localhost native]# java -version
java version "1.7.0_67"
注:官网强调一定是1.7以上版本.
Java(TM) SE RuntimeEnvironment (build 1.7.0_67-b01)
Java HotSpot(TM) 64-Bit ServerVM (build 24.65-b04, mixed mode)
2.1.3mysql版本
[root@localhost native]# mysql -uroot -proot-e"select version()";
Warning: Using a password onthe command line interface can be insecure.
+-----------+
| version() |
+-----------+
| 5.6.14 |
+-----------+
注:
1 Mysql 驱动为mysql-connector-java-5.1.31-bin.jar
2 该jar需要放置在/usr/share/cmf/lib/内被CM安装时用到
2.2 安装准备
2.2.1 禁用防火墙
service iptables stop
chkconfig iptables off
2.2.2 禁用Selinux
vi /etc/selinux/config
#修改内容见下:
SELINUX=disabled
SELINUXTYPE=targeted
2.2.3 修改主机名
vi/etc/hosts
#加上archive.cloudera.com域名
192.168.56.101 archive.cloudera.com shenl
vi/etc/sysconfig/network
#修改主机名如下:
HOSTNAME=archive.cloudera.com
#重启电脑使得更改生效.
2.2.4 配置本地源
1) 虚拟机加载Linux ISO镜像
2) 找到rom对应设备名
[root@localhost ~]# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
sr0 11:0 1 1024M 0 rom
sr1 11:1 1 3.6G 0 rom
sr2 11:2 1 1024M 0 rom
sda 8:0 0 8G 0 disk
?..sda1 8:1 0 500M 0 part /boot
?..sda2 8:2 0 7.5G 0 part
?..VolGroup-lv_root (dm-0) 253:0 0 6.7G 0 lvm /
?..VolGroup-lv_swap (dm-1) 253:1 0 816M 0 lvm [SWAP]
3) 创建目录
[root@localhost ~]#mkdir -p /mnt/cdrom/
4) 挂载镜像
mount -t iso9660 /dev/sr1 /mnt/cdrom
5) 配置linux本地更新源(真实环境这里的应该baseurl应配置成http方式,如http://ios)
vi /etc/yum.repos.d/redhat.repo
#修改内容如下:
[rhel-source]
name=Redhat
baseurl=file:///mnt/cdrom/
enabled=1
gpgcheck=1
gpgkey=file:///mnt/cdrom//RPM-GPG-KEY-redhat-release
6) 更新更新源
yum clean all
yum makecache
yum update list
2.2.5 安装配置mysql
1) 删除系统自带的mysql
rpm -qa | grep -i mysql
MySQL-client-5.1.49-1.glibc23
rpm -e --nodeps MySQL-client-5.1.49-1.glibc23
chkconfig --list | grep -i mysql
mysql 0:off 1:off 2:on 3:on 4:on 5:on 6:off
chkconfig --del mysql
whereis mysql
mysql: /usr/lib/mysql /usr/share/mysql
rm -rf /usr/lib/mysql/
rm -rf /usr/share/mysql
2) 安装Mysql服务、客户端
rpm -ivh MySQL-shared-5.6.14-1.el6.x86_64.rpm
rpm -ivh MySQL-shared-compat-5.6.14-1.el6.x86_64.rpm
rpm -ivh MySQL-server-5.6.14-1.el6.x86_64.rpm
rpm -ivh MySQL-client-5.6.14-1.el6.x86_64.rpm
3) 启动mysql服务
service mysql start
4) 修改mysql初始密码,先找到安装时的初始密码,在修改成自己的密码
[root@localhost ~]#cat /root/.mysql_secret
# The random password set for the root user at Tue Dec 2221:17:22 2015 (local time):RUmKBqcY
mysql -uroot -p RUmKBqcY
set password=password(‘root’)
注: 如果想修改mysql默认的datadir,假设到/data/mysqldata内,可参考如下方式修改:
#迁移mysql数据目录,做软连接,修改目录的所有者
mv /var/lib/mysql//data/mysqldata/mysql
ln -s /data/mysqldata/mysql/var/lib/
chown -R mysql:mysql/data/mysqldata/mysql
2.2.6 安装httpd,并按照如下的目录结构放置CM和CDH parcel的安装文件.
yum install tree
yum install httpd
service httpd start
cd /var/www/html
注:如果想修改httpd的网站默认主目录,可通过如下方式:
mkdir -p /data/www/
chown -R apache:apache/data/www/
vi /etc/httpd/conf/httpd.conf
#找到DocumentRoot 修改"/var/www/html"成"/data/www"
service httpd restart
报如下错:
Forbidden
You don't have permission to access /cm/ on this server.
vi /etc/httpd/conf/httpd.conf
#找到 <Directory"/var/www/html"> 修改成 /data/www
service httpd restart
#CM的目录安排需要按照如下方式:
[root@archive html]# tree./cm5/
./cm5/
└── redhat
└── 6
└── x86_64
└── cm
├── 5.5.1
│ ├── repodata
│ │ ├── filelists.xml.gz
│ │ ├── filelists.xml.gz.asc
│ │ ├── other.xml.gz
│ │ ├── other.xml.gz.asc
│ │ ├── primary.xml.gz
│ │ ├── primary.xml.gz.asc
│ │ ├── repomd.xml
│ │ └── repomd.xml.asc
│ └── RPMS
│ └── x86_64
│ ├──cloudera-manager-agent-5.5.1-1.cm551.p0.8.el6.x86_64.rpm
│ ├── cloudera-manager-daemons-5.5.1-1.cm551.p0.8.el6.x86_64.rpm
│ ├──cloudera-manager-server-5.5.1-1.cm551.p0.8.el6.x86_64.rpm
│ ├──cloudera-manager-server-db-2-5.5.1-1.cm551.p0.8.el6.x86_64.rpm
│ ├──enterprise-debuginfo-5.5.1-1.cm551.p0.8.el6.x86_64.rpm
│ ├── jdk-6u31-linux-amd64.rpm
│ └──oracle-j2sdk1.7-1.7.0+update67-1.x86_64.rpm
├── cloudera-manager.repo
└── RPM-GPG-KEY-cloudera
8 directories, 17 files
#CDH parcel的目录安排需要按照如下方式:
[root@archive html]#tree ./cdh5/
./cdh5/
└── parcels
└── 5.5.1
├── CDH-5.5.1-1.cdh5.5.1.p0.11-el6.parcel
├── CDH-5.5.1-1.cdh5.5.1.p0.11-el6.parcel.sha1
└── manifest.json
2 directories, 3 files
#将CM安装脚本放置/var/www/html(可放置在其它目录,这里为了统一).:
chmod +x cloudera-manager-installer.bin
注:下载repodata文件夹时可以在Window里下载好,执行wget命令
E:\ProgramFiles\wget-1.16.3-win64>wget -r -np -ndhttp://archive-primary.cloude
ra.com/cm5/redhat/6/x86_64/cm/5.5.1/repodata/
3安装
3.1安装中
1) 执行CM安装脚本
./cloudera-manager-installer.bin
#分析并解决CM安装过程中的问题,详细见 3.2安装问题:
cat/var/log/cloudera-manager-installer/2.install-oracle-j2sdk1.7.log
2) 输入CM所在的IP地址和7180端口,进行界面化安装CDH.
#选择存储库时需要点击parcel旁边的更多选项,删除默认定义的路径,写上
http://192.168.56.101/cdh5/parcels/5.5.1
或者
http://archive.cloudera.com/cdh5/parcels/5.5.1
#这时如果选择此Cloudera Manager Server的匹配发行版,会报错,想见3.2安装问题之d).这里建议选择自定义存储库.见下图:
2) 按照提示一步步安装,直至集群安装成功…
注: CM在安装时会自动生成cloudera-manager.repo,请确认里面的url跟自定义存储库里的一致,建议走http协议.cloudera-manager.repo里默认下载地址见下:
https://archive.cloudera.com/cm5/redhat/6/x86_64/cm/5.5.1
3.2 安装问题
1) 正确配置CM和CDH的目录树结构
2) 正确配置parcel和自定义存储库
a) Loaded plugins: product-id,security, subscription-manager
This system is not registered to Red Hat SubscriptionManagement. You can use subscription-manager to register.
http://archive.cloudera.com/cm5/redhat/6/x86_64/cm/5/repodata/repomd.xml:[Errno 14] PYCURL ERROR 22 - "The requested URL returned error: 404 NotFound"
解决方案:重命名文件名,详细见下:
cd /var/www/html
mv ./cm5/redhat/6/x86_64/cm/5.5.1./cm5/redhat/6/x86_64/cm/5
b) .
解决方案:重命名文件名,详细见如下:
mv ./cm5/redhat/6/x86_64/cm/5 ./cm5/redhat/6/x86_64/cm/5.5.1
c) CMinstaller正常安装相关日志
[root@archive cloudera-manager-installer]# pwd
/var/log/cloudera-manager-installer
[root@archivecloudera-manager-installer]# ll
total 20
-rw-r--r-- 1 root root 0 Jan 4 15:51 0.check-selinux.log
-rw-r--r-- 1 root root 0 Jan 4 15:51 1.install-repo-pkg.log
-rw-r--r-- 1 root root 1647 Jan 4 15:51 2.install-oracle-j2sdk1.7.log
-rw-r--r-- 1 root root 2080 Jan 4 15:53 3.install-cloudera-manager-server.log
-rw-r--r-- 1 root root 0 Jan 4 15:46 3.remove-cloudera-manager-repository.log
-rw-r--r-- 1 root root 3250 Jan 4 15:53 4.install-cloudera-manager-server-db-2.log
-rw-r--r-- 1 root root 1978 Jan 4 15:54 5.start-embedded-db.log
-rw-r--r-- 1 root root 59 Jan 4 15:54 6.start-scm-server.log
d) 按照错误提示修改成b)后仍然报如下错误
BEGIN yum info cloudera-manager-agent
Loaded plugins: product-id, security, subscription-manager
This system is not registered to Red Hat Subscription Management. You can usesubscription-manager to register.
https://archive.cloudera.com/cm5/redhat/6/x86_64/cm/5.5.1/repodata/repomd.xml:[Errno 14] PYCURL ERROR 7 - "couldn't connect to host"
Trying other mirror.
Error: Cannot retrieve repository metadata (repomd.xml) for repository:cloudera-manager. Please verify its path and try again
解决方案:这时需要在 群集安装→选择存储库里选择自定义存储库,配置见下图:
e) 配置CM事件监控、报表管理数据库时的错误
解决方案: 拷贝mysql驱动到cm的相关lib目录下.
cp mysql-connector-java-5.1.31-bin.jar/usr/share/cmf/lib/
注:mysql驱动事先下载好.
f) 添加Hive服务指定mysql作为数据源时报找不到驱动的错误
解决方法:拷贝mysql驱动
cp mysql-connector-java-5.1.31-bin.jar/usr/share/java/
mv /usr/share/java/mysql-connector-java.jar
g) CM集群验证错误
解决方法:按照提示修改告警
vi/etc/sysctl.conf
vm.swappiness= 0
/sbin/sysctl-p /etc/sysctl.conf
#确认是否修改为0
cat/proc/sys/vm/swappiness
#永久修复该告警
vi/etc/rc.local
#追加如下内容到上述文件内
echo never > /sys/kernel/mm/redhat_transparent_hugepage/defrag
#临时修复该告警
echonever > /sys/kernel/mm/redhat_transparent_hugepage/defrag
4 配置
暂无
5 使用
登录CM安装IP和7180对CDH进行管理和维护.
6 卸载
暂无
7 总结
通过日志分析安装的问题.
