1. 環境准備
1.1. 處理Linux環境
a) 確保各主機擁有root權限
主要是為了避免安裝過程中的權限問題
b) 關閉防火牆
[root@abc] systemctl stop firewalld [root@abc] systemctl disable firewalld
c) 確保如下端口是開放的
https://www.cloudera.com/documentation/enterprise/6/6.1/topics/cm_ig_ports.html
d) 關閉SELinux
[root@abc] getenforce
如果輸出為Disabled，則不需要修改配置，如果為enforcing，則需要做如下配置
[root@abc] vim /etc/selinux/config #----------------------------------- SELINUX=disabled #----------------------------------- :wq [root@abc] setenforce 0
e) 調整交換內存閥值
[root@abc] sysctl vm.swappiness=10 && echo "vm.swappiness=10" >> /etc/sysctl.conf
f) 關閉大內存頁配置
[root@abc] echo never > /sys/kernel/mm/transparent_hugepage/defrag && echo never > /sys/kernel/mm/transparent_hugepage/enabled [root@abc] chmod +x /etc/rc.d/rc.local [root@abc] echo "echo never > /sys/kernel/mm/transparent_hugepage/defrag" >> /etc/rc.local [root@abc] echo "echo never > /sys/kernel/mm/transparent_hugepage/enabled" >> /etc/rc.local
1.2. 配置HOSTNAME
主機規划：
| IP |
主機名 |
配置 |
服務 |
| 10.6.222.85 |
txznwltesthadoop1 |
8c16g |
ClouderaManagerServer、ClouderaManagerAgent、ClouderaManagerService、HistoryServer(Spark)、JobHistoryServer(YARN)、HDFSBalancer |
| 10.6.222.86 |
txznwltesthadoop2 |
8c16g |
ClouderaManagerAgent、NameNode、ResourceManager、HMaster |
| 10.6.222.87 |
txznwltesthadoop3 |
24c48g |
ClouderaManagerAgent、DataName、NodeManager、HRegionServer、SparkGateway |
| 10.6.222.88 |
txznwltesthadoop4 |
24c48g |
ClouderaManagerAgent、DataName、NodeManager、HRegionServer、SparkGateway |
| 10.6.222.89 |
txznwltesthadoop5 |
24c48g |
ClouderaManagerAgent、DataName、NodeManager、HRegionServer、SparkGateway |
| 10.6.222.90 |
txznwltesthadoop6 |
8c16g |
ClouderaManagerAgent、ZooKeeper、KafkaBroker |
| 10.6.222.91 |
txznwltesthadoop7 |
8c16g |
ClouderaManagerAgent、ZooKeeper、KafkaBroker |
| 10.6.222.92 |
txznwltesthadoop8 |
8c16g |
ClouderaManagerAgent、ZooKeeper、KafkaBroker |
[root@abc] vim /etc/hostname #---------------------------------------- txznwltesthadoop1 # 對應的主機設置對應的主機名 #---------------------------------------- :wq [root@abc] vim /etc/sysconfig/network #---------------------------------------- HOSTNAME=txznwltesthadoop1 # 對應的主機設置對應的主機名 #---------------------------------------- :wq [root@abc] hostname txznwltesthadoop1 # 對應的主機設置對應的主機名 [root@abc] vim /etc/hosts #---------------------------------------- 10.6.222.85 txznwltesthadoop1 10.6.222.86 txznwltesthadoop2 10.6.222.87 txznwltesthadoop3 10.6.222.88 txznwltesthadoop4 10.6.222.89 txznwltesthadoop5 10.6.222.90 txznwltesthadoop6 10.6.222.91 txznwltesthadoop7 10.6.222.92 txznwltesthadoop8 #---------------------------------------- :wq
注意：這里必須設置，並且主機名不能有下划線
1.3. 配置SSH免密
使用ssh-copy-id進行免密配置
[root@abc] ssh-keygen -t rsa # 不需要設置密碼，直接全部回車 [root@abc] ssh-copy-id -p 60028 root@10.6.222.85
或者直接在其他機器上配置公鑰
# 在主機器上生成RSA密鑰對 [root@abc] ssh-keygen -t rsa # 在需要背面密的機器上做如下操作 [root@abc] mkdir ~/.ssh && chmod 700 ~/.ssh && cd ~/.ssh [root@abc] echo "公鑰<id_rsa.pub>內容" > authorized_keys && chmod 600 authorized_keys
1.4. JDK版本要求
JDK把版本必須符合以下： https://docs.cloudera.com/documentation/enterprise/6/6.1/topics/cm_intro_primer.html
注：需要先卸載與OpenJDK相關的所有包
1.5. NTP服務器配置
需要在所有節點上配置ntpd服務
[root@abc] yum -y install ntp [root@abc] vim /etc/ntp.conf #------------------------------------------------------- # 添加如下的server配置 server time1.aliyum.com prefer server time2.aliyum.com prefer server time3.aliyum.com prefer server time4.aliyum.com prefer server time5.aliyum.com prefer server time6.aliyum.com prefer server time7.aliyum.com prefer server 0.centos.pool.ntp.org iburst server 1.centos.pool.ntp.org iburst server 2.centos.pool.ntp.org iburst server 3.centos.pool.ntp.org iburst #------------------------------------------------------- :wq [root@abc] systemctl enable ntpd.service && systemctl start ntpd.service
1.6. 其他環境
1. 必須安裝python2.7，CentOS7自帶可以忽略
2. 最好保證設備上只有一個網卡同時工作
3. 如果/tmp目錄單獨掛在了一個分區或者邏輯卷，則大小必須大於5G
2. 安裝Cloudera Manager
2.1. Cloudera Manager部分的安裝
YUM方式安裝Cloudera Manager
[root@abc] wget https://archive.cloudera.com/cm6/6.1.1/redhat7/yum/cloudera-manager.repo -P /etc/yum.repos.d/ [root@abc] rpm --import https://archive.cloudera.com/cm6/6.1.1/redhat7/yum/RPM-GPG-KEY-cloudera [root@abc] yum list | grep cloudera [root@abc] yum install -y oracle-j2sdk1.8 cloudera-manager-daemons cloudera-manager-agent cloudera-manager-server [root@abc] echo "" >> /etc/profile && \ echo "export JAVA_HOME=/usr/local/jdk" >> /etc/profile && \ echo "export CLASSPATH=.:\$JAVA_HOME/lib/dt.jar:\$JAVA_HOME/lib/tools.jar" >> /etc/profile && \ echo "export PATH=\$PATH:\$JAVA_HOME/bin" >> /etc/profile && \ source /etc/profile [root@abc] java -version
注：由於安裝包太大，所以不建議進行網絡模式安裝，最好提前下載好RPM包，然后使用下面的本地RPM方式進行安裝；
本地RPM包進行安裝
### 所有節點執行如下操作 [root@abc] wget https://archive.cloudera.com/cm6/6.1.1/redhat7/yum/RPMS/x86_64/cloudera-manager-daemons-6.1.1-853290.el7.x86_64.rpm -P /opt/src/ [root@abc] wget https://archive.cloudera.com/cm6/6.1.1/redhat7/yum/RPMS/x86_64/cloudera-manager-agent-6.1.1-853290.el7.x86_64.rpm -P /opt/src/ [root@abc] wget https://archive.cloudera.com/cm6/6.1.1/redhat7/yum/RPMS/x86_64/ oracle-j2sdk1.8-1.8.0+update181-1.x86_64.rpm -P /opt/src/ [root@abc] yum -y localinstall /opt/src/oracle-j2sdk1.8-1.8.0+update181-1.x86_64.rpm –nogpgcheck [root@abc] ln -s /usr/java/jdk1.8.0_181-cloudera /usr/local/jdk [root@abc] echo "" >> /etc/profile && \ echo "export JAVA_HOME=/usr/local/jdk" >> /etc/profile && \ echo "export CLASSPATH=.:\$JAVA_HOME/lib/dt.jar:\$JAVA_HOME/lib/tools.jar" >> /etc/profile && \ echo "export PATH=\$PATH:\$JAVA_HOME/bin" >> /etc/profile && \ source /etc/profile [root@abc] java -version [root@abc] yum -y localinstall /opt/src/cloudera-manager-daemons-6.1.1-853290.el7.x86_64.rpm --nogpgcheck [root@abc] yum -y localinstall /opt/src/cloudera-manager-agent-6.1.1-853290.el7.x86_64.rpm –nogpgcheck ### 僅Master主控節點需要指向如下操作 [root@abc] wget https://archive.cloudera.com/cm6/6.1.1/redhat7/yum/RPMS/x86_64/cloudera-manager-server-6.1.1-853290.el7.x86_64.rpm -P /opt/src/ [root@abc] yum -y localinstall /opt/src/cloudera-manager-server-6.1.1-853290.el7.x86_64.rpm --nogpgcheck
注：注意安裝順序，包之間是有依賴的
2.2. 預先准備CDH Parcel相關文件
[root@abc] wget https://archive.cloudera.com/cdh6/6.1.1/parcels/manifest.json -P /opt/src/ [root@abc] wget https://archive.cloudera.com/cdh6/6.1.1/parcels/CDH-6.1.1-1.cdh6.1.1.p0.875250-el7.parcel -P /opt/src/ # 如果 /opt/cloudera/parcel-repo/ 目錄存在，則不需要創建 [root@abc] mkdir /opt/cloudera/parcel-repo/ [root@abc] cp /opt/src/parcel/manifest.json /opt/cloudera/parcel-repo/ [root@abc] cp /opt/src/parcel/CDH-6.1.1-1.cdh6.1.1.p0.875250-el7.parcel /opt/cloudera/parcel-repo/ [root@abc] sha1sum /opt/src/parcel/CDH-6.1.1-1.cdh6.1.1.p0.875250-el7.parcel [root@abc] echo "生成的SHA1字符串" >> /opt/cloudera/parcel-repo/CDH-6.1.1-1.cdh6.1.1.p0.875250-el7.parcel.sha [root@abc] chown cloudera-scm:cloudera-scm -R /opt/cloudera/parcel-repo/
2.3. 安裝MySQL
MySQL版本為：mysql-5.7.24-el7-x86_64
[root@abc] wget https://cdn.mysql.com//Downloads/MySQL-5.7/mysql-5.7.24-el7-x86_64.tar.gz -P /opt/src/ [root@abc] groupadd mysql && useradd mysql -g mysql [root@abc] tar -zxvf /opt/src/mysql-5.7.24-el7-x86_64.tar.gz -C /opt/app/ [root@abc] ln -s /opt/app/mysql-5.7.24-el7-x86_64 /usr/local/mysql [root@abc] mkdir -p /opt/data/mysql/{data,logs} [root@abc] chown mysql:mysql -R /opt/app/mysql-5.7.24-el7-x86_64 /usr/local/mysql /opt/data/mysql
配置MySQL
[root@abc] cp /etc/my.cnf /etc/my_sample.cnf [root@abc] vim /etc/my.cnf #----------------------------------------------------------------------------- [client] port=3306 [mysql] socket=/opt/data/mysql/mysqld.sock [mysqld] basedir=/usr/local/mysql datadir=/opt/data/mysql/data socket=/opt/data/mysql/mysqld.sock transaction-isolation = READ-COMMITTED # Disabling symbolic-links is recommended to prevent assorted security risks; # to do so, uncomment this line: symbolic-links = 0 key_buffer_size = 32M max_allowed_packet = 32M thread_stack = 256K thread_cache_size = 64 query_cache_limit = 8M query_cache_size = 64M query_cache_type = 1 max_connections = 550 #expire_logs_days = 10 #max_binlog_size = 100M #log_bin should be on a disk with enough free space. #Replace '/var/lib/mysql/mysql_binary_log' with an appropriate path for your #system and chown the specified folder to the mysql user. log_bin=/opt/data/mysql/mysql_binary_log #In later versions of MySQL, if you enable the binary log and do not set #a server_id, MySQL will not start. The server_id must be unique within #the replicating group. server_id=1 binlog_format = mixed read_buffer_size = 2M read_rnd_buffer_size = 16M sort_buffer_size = 8M join_buffer_size = 8M # InnoDB settings innodb_file_per_table = 1 innodb_flush_log_at_trx_commit = 2 innodb_log_buffer_size = 64M innodb_buffer_pool_size = 2G innodb_thread_concurrency = 8 innodb_flush_method = O_DIRECT innodb_log_file_size = 512M [mysqld_safe] log-error=/opt/data/mysql/logs/mysqld_safe.log pid-file=/opt/data/mysql/logs/mysqld_dafe.pid sql_mode=STRICT_ALL_TABLES #----------------------------------------------------------------------------- :wq
初始化數據庫
[root@abc] ./bin/mysqld --initialize --user=mysql --basedir=/usr/local/mysql/ --datadir=/opt/data/mysql/data
# 注意打印出的日志里面的 root 用戶的初始密碼，沒有密碼不能登錄，之后會修改密碼
配制成SystemCTL服務
[root@abc] vim /usr/lib/systemd/system/mysqld.service #------------------------------------------------------------ [Unit] Description=mysql After=network.target remote-fs.target nss-lookup.target [Service] User=mysql Group=mysql Type=forking ExecStart=/usr/local/mysql/support-files/mysql.server start ExecReload=/usr/local/mysql/support-files/mysql.server restart ExecStop=/usr/local/mysql/support-files/mysql.server stop #PrivateTmp=true [Install] WantedBy=multi-user.target #------------------------------------------------------------ :wq ##注意：啟動、重啟、停止命令全部要求使用絕對路徑 [root@abc] chmod +x /usr/lib/systemd/system/mysqld.service [root@abc] systemctl daemon-reload & systemctl enable mysqld.service & systemctl start mysqld.service
登錄修改密碼
[root@abc] mysql -uroot -p Enter password: -> alter user 'root'@'localhost' identified by 'root'; -> use mysql -> update user set host='%' where user='root'; -> flush privileges; -> quit;
創建Cloudera Manager需要的數據庫
[root@abc] mysql -uroot -p Enter password: -> CREATE DATABASE scm DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; -> CREATE DATABASE amon DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; -> CREATE DATABASE rman DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; -> CREATE DATABASE hue DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; -> CREATE DATABASE metastore DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; -> CREATE DATABASE sentry DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; -> CREATE DATABASE nav DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; -> CREATE DATABASE navms DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; -> CREATE DATABASE oozie DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; > CREATE DATABASE hive DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; -> quit;
主節點下載並配置MySQL驅動
[root@abc] wget https://dev.mysql.com/get/Downloads/Connector-J/mysql-connector-java-5.1.46.tar.gz -P /opt/src/ [root@abc] mkdir -p /usr/share/java/ [root@abc] tar -zxvf /opt/src/mysql-connector-java-5.1.46.tar.gz -C /opt/src/ [root@abc] cp /opt/src/mysql-connector-java-5.1.46/mysql-connector-java-5.1.46-bin.jar /usr/share/java/mysql-connector-java.jar
注：Cloudera Manager是用的MySQL驅動版本必須是5.1
初始化Cloudera Manager Database（MySQL）
[root@abc] /opt/cloudera/cm/schema/scm_prepare_database.sh mysql scm root root # 解釋: # /opt/cloudera/cm/schema/scm_prepare_database.sh 數據庫類型 數據庫名 用戶名 密碼 # Cloudera Manager Database默認名為scm，默認用戶名為scm (具體請以你建庫時的設計為准)。
2.4. 啟動Cloudera Manager
[root@abc] systemctl enable cloudera-scm-server && systemctl start cloudera-scm-server
3. 開始使用並配置Cloudera Manager
3.1. Cloudera Manager 歡迎頁
初始默認賬號和密碼為：admin/admin
注：當前安裝版本為6.2.1，這里圖片是以前安裝6.1.1時截的，差別不大。
3.2. Cloudera Manager用戶協議
3.3. 選擇Cloudera Manager版本（免費版還是商業版）
3.4. Cloudera Manager集群安裝歡迎頁
3.5. 配置集群主機節點
3.6. 選擇CDH版本並下載
注：當前安裝版本為6.2.1，直接選擇6.2.1的cdh即可，上圖為以前安裝6.1.1時截的圖。
3.7. Java安裝
注：由於之前已經安裝好了JDK所以這里不用勾選
3.8. 提供SSH登入憑據
注：私鑰文件需要將上傳服務器上的id_rsa文件。由於生成rsa密鑰對的時候沒有輸入密碼，說以這里也不需要輸入密碼。由於只有8個節點，所以同時安裝數輸入8就夠了；
3.9. 開始所有主機的相關服務安裝
注意：
1.如果只是出現上述問題，說明軟件后已經安裝成功，那么可以忽略進入下一步；
2.如果各個節點安裝詳細信息出現 Delta RPMs disabled because /usr/bin/applydeltarpm not installed.錯誤
則需要：
[root@abc] yum provides '*/applydeltarpm' && yum install deltarpm
3.在執行之前，最好自己先在從節點安裝好cloudera-manager-daemons和cloudera-manager-agent服務，這樣既可以不用等待太長時間
3.10. 開始CDH的下載和分發
如果出現：
則需要刪除問題節點的/var/lib/cloudera-scm-agent/cm_guid文件，然后重啟問題節點的cloudera-scm-agent服務
3.11. 檢查各個主機環境
如果有警告問題可按頁面提示進行操作，解決問題
3.12. 開始CDH相關服務的安裝
選擇自定義服務，這里部署HBase、HDFS、Kafka、Spark、YARN、Hive、ZooKeeper等服務
3.13. 配置Hive MetaStore數據庫
這里cloudera默認會給一個服務在集群中的分布情況，自己也可以進行修改
這里需要修改數據存儲目錄，和填寫郵箱，host monitor和service monitor存儲目錄，目錄在啟動時自己創建；
3.14. 集群搭建成功
