本文是在CentOS7.4 下進行CDH6集羣的徹底離線部署。CDH5集羣與CDH6集羣的部署區別比較大。html
說明:本文內容全部操做都是在root用戶下進行的。java
首先一些安裝CDH6集羣的必須文件要先在外網環境先下載好。node
CM6 RPM:https://archive.cloudera.com/...
須要下載該連接下的全部RPM文件,保存到cloudera-repos
目錄下。mysql
ASC文件:https://archive.cloudera.com/...
同時還須要下載一個asc文件,一樣保存到cloudera-repos
目錄下:linux
[root@node01 upload]# tree cloudera-repos/ cloudera-repos/ ├── allkeys.asc ├── cloudera-manager-agent-6.3.0-1281944.el7.x86_64.rpm ├── cloudera-manager-daemons-6.3.0-1281944.el7.x86_64.rpm ├── cloudera-manager-server-6.3.0-1281944.el7.x86_64.rpm ├── cloudera-manager-server-db-2-6.3.0-1281944.el7.x86_64.rpm ├── enterprise-debuginfo-6.3.0-1281944.el7.x86_64.rpm └── oracle-j2sdk1.8-1.8.0+update181-1.x86_64.rpm
要求使用5.1.26以上版本的jdbc驅動,可點擊這裏直接下載mysql-connector-java-5.1.47.tar.gz
c++
注意:不要嘗試使用FTP搭建CM的YUM庫!sql
首先安裝httpd
和createrepo
:數據庫
yum -y install httpd createrepo
啓動httpd
服務並設置開機自啓動:apache
systemctl start httpd systemctl enable httpd
而後進入到前面準備好的存放Cloudera Manager RPM包的目錄cloudera-repos
下:json
cd /data6/upload/cloudera-repos/
生成RPM元數據:
createrepo . chmod 777 -R cloudera-repos
而後將cloudera-repos
目錄移動到httpd的html目錄下:
mv cloudera-repos /var/www/html/
確保能夠經過瀏覽器查看到這些RPM包:
接着在建立cm6的repo文件(每一個節點都須要配置):
cd /etc/yum.repos.d vim cloudera-manager.repo
添加以下內容:
[cloudera-manager] name=Cloudera Manager 6.3.0 baseurl=http://node01/cloudera-repos/ gpgcheck=0 enabled=1
保存,退出,而後執行yum clean all && yum makecache
命令:
[root@master02 ~]# yum clean all && yum makecache Loaded plugins: fastestmirror, langpacks Cleaning repos: ChinaUnicom-Packages cloudera-manager Cleaning up everything Maybe you want: rm -rf /var/cache/yum, to also free up space taken by orphaned data from disabled or removed repos Loaded plugins: fastestmirror, langpacks ChinaUnicom-Packages | 3.6 kB 00:00:00 cloudera-manager | 2.9 kB 00:00:00 (1/7): ChinaUnicom-Packages/group_gz | 156 kB 00:00:00 (2/7): ChinaUnicom-Packages/filelists_db | 3.1 MB 00:00:00 (3/7): ChinaUnicom-Packages/primary_db | 3.1 MB 00:00:00 (4/7): ChinaUnicom-Packages/other_db | 1.2 MB 00:00:00 (5/7): cloudera-manager/filelists_db | 118 kB 00:00:00 (6/7): cloudera-manager/other_db | 1.0 kB 00:00:00 (7/7): cloudera-manager/primary_db | 8.6 kB 00:00:00 Determining fastest mirrors Metadata Cache Created
這一步只須要在CM Server節點上操做。
執行下面的命令:
# 安裝openjdk8 yum install oracle-j2sdk1.8 # 安裝 cm manager(只需在server節點安裝) yum install cloudera-manager-daemons cloudera-manager-agent cloudera-manager-server
將會須要不少依賴包,因此說仍是有必要搭一個局域網內yum源的,或者手動安裝rpm包
Cloudera Manager Server安裝完成後,進入到本地Parcel存儲庫目錄:
cd /opt/cloudera/parcel-repo
將第一部分下載的CDH parcels文件上傳至該目錄下,而後執行修改sha文件:
mv /data6/upload/parcels/* /opt/cloudera/parcel-repo/ mv CDH-6.3.0-1.cdh6.3.0.p0.1279813-el7.parcel.sha1 CDH-6.3.0-1.cdh6.3.0.p0.1279813-el7.parcel.sha
而後執行下面的命令修改文件全部者:
chown -R cloudera-scm:cloudera-scm /opt/cloudera/parcel-repo
最終/opt/cloudera/parcel-repo
目錄內容以下:
├── CDH-6.3.0-1.cdh6.3.0.p0.1279813-el7.parcel ├── CDH-6.3.0-1.cdh6.3.0.p0.1279813-el7.parcel.sha └── manifest.json
MySQL的安裝在環境準備部分中已經有說明,這裏就跳過MySQL安裝了。
CDH官方給的有一份推薦的MySQL的配置內容:
[mysqld] datadir=/var/lib/mysql socket=/var/lib/mysql/mysql.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=/var/lib/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 = 4G innodb_thread_concurrency = 8 innodb_flush_method = O_DIRECT innodb_log_file_size = 512M [mysqld_safe] log-error=/var/log/mysqld.log pid-file=/var/run/mysqld/mysqld.pid sql_mode=STRICT_ALL_TABLES
從前面下載好的mysql-connector-java-5.1.47.tar.gz
包中解壓出mysql-connector-java-5.1.47-bin.jar
文件,將mysql-connector-java-5.1.47-bin.jar
文件上傳至CM Server節點上的/usr/share/java/
目錄下並重命名爲mysql-connector-java.jar
(若是/usr/share/java/
目錄不存在,須要手動建立):
tar zxvf mysql-connector-java-5.1.47.tar.gz mkdir -p /usr/share/java/ cp mysql-connector-java-5.1.47-bin.jar /usr/share/java/mysql-connector-java.jar
根據所須要安裝的服務參照下表建立對應的數據庫以及數據庫用戶,數據庫必須使用utf8編碼,建立數據庫時要記錄好用戶名及對應密碼:
服務名 | 數據庫名 | 用戶名 |
---|---|---|
Cloudera Manager Server | scm | scm |
Activity Monitor | amon | amon |
Reports Manager | rman | rman |
Hue | hue | hue |
Hive Metastore Server | metastore | hive |
Sentry Server | sentry | sentry |
Cloudera Navigator Audit Server | nav | nav |
Cloudera Navigator Metadata Server | navms | navms |
Oozie | oozie | oozie |
建立數據庫及對應用戶:
# scm CREATE DATABASE scm DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; GRANT ALL ON scm.* TO 'scm'@'%' IDENTIFIED BY 'scm'; # amon CREATE DATABASE amon DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; GRANT ALL ON amon.* TO 'amon'@'%' IDENTIFIED BY 'amon'; # rman CREATE DATABASE rman DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; GRANT ALL ON rman.* TO 'rman'@'%' IDENTIFIED BY 'rman'; # hue CREATE DATABASE hue DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; GRANT ALL ON hue.* TO 'hue'@'%' IDENTIFIED BY 'hue'; # hive CREATE DATABASE metastore DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; GRANT ALL ON metastore.* TO 'hive'@'%' IDENTIFIED BY 'hive'; # sentry CREATE DATABASE sentry DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; GRANT ALL ON sentry.* TO 'sentry'@'%' IDENTIFIED BY 'sentry'; # nav CREATE DATABASE nav DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; GRANT ALL ON nav.* TO 'nav'@'%' IDENTIFIED BY 'nav'; # navms CREATE DATABASE navms DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; GRANT ALL ON navms.* TO 'navms'@'%' IDENTIFIED BY 'navms'; # oozie CREATE DATABASE oozie DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci; GRANT ALL ON oozie.* TO 'oozie'@'%' IDENTIFIED BY 'oozie'; # flush FLUSH PRIVILEGES;
Cloudera Manager Server包含一個配置數據庫的腳本。
/opt/cloudera/cm/schema/scm_prepare_database.sh mysql scm scm
/opt/cloudera/cm/schema/scm_prepare_database.sh mysql -h <mysql-host-ip> --scm-host <cm-server-ip> scm scm
[root@master02 ~]# /opt/cloudera/cm/schema/scm_prepare_database.sh mysql -h 10.172.54.51 --scm-host 10.172.54.52 scm scm Enter SCM password: JAVA_HOME=/usr/java/jdk1.8.0_181-cloudera Verifying that we can write to /etc/cloudera-scm-server Creating SCM configuration file in /etc/cloudera-scm-server Executing: /usr/java/jdk1.8.0_181-cloudera/bin/java -cp /usr/share/java/mysql-connector-java.jar:/usr/share/java/oracle-connector-java.jar:/usr/share/java/postgresql-connector-java.jar:/opt/cloudera/cm/schema/../lib/* com.cloudera.enterprise.dbutil.DbCommandExecutor /etc/cloudera-scm-server/db.properties com.cloudera.cmf.db. [ main] DbCommandExecutor INFO Successfully connected to database. All done, your SCM database is configured correctly!
systemctl start cloudera-scm-server
而後等待Cloudera Manager Server啓動,可能須要稍等一下子,能夠經過命令tail -f /var/log/cloudera-scm-server/cloudera-scm-server.log
去監控服務啓動狀態。
當看到INFO WebServerImpl:com.cloudera.server.cmf.WebServerImpl: Started Jetty server.
日誌打印出來後,說明服務啓動成功,能夠經過瀏覽器訪問Cloudera Manager WEB界面了。
打開瀏覽器,訪問地址:http://<server_host>:7180
,默認帳號和密碼都爲admin:
首先是Cloudera Manager的歡迎頁面,點擊頁面右下角的【繼續】按鈕進行下一步:
### 接受條款
勾選接受條款,點擊【繼續】進行下一步:
### 版本選擇
這裏我就選擇免費版了:
選擇版本之後會出現第二個歡迎界面,不過這個是安裝集羣的歡迎頁:
這一步是要搜索並選擇用於安裝CDH集羣的主機,在主機名稱後面的輸入框中輸入各個節點的hostname,中間使用英文逗號分隔開,而後點擊搜索,在結果列表中勾選要安裝CDH的節點便可:
這裏選擇自定義,填寫上面使用httpd搭建好的Cloudera Manager YUM 庫URL:
若是咱們以前的【配置本地Parcel存儲庫】步驟操做無誤的話,這裏會自動選擇【使用Parcel】,並加載出CDH版本,確認無誤後點擊【繼續】:
所以,不須要本身手動安裝 Cloudera Manager Agent了
這一步驟我就再也不勾選安裝JDK了,由於我在環境準備部分已經安裝過了。取消勾選,而後繼續:
用於配置集羣主機之間的SSH登陸,填寫root用戶的密碼,根據集羣配置填寫合適的【同時安裝數量】值便可:
到這一步會自動進行節點Agent的安裝,稍等一下子,便可安裝完成:
這一步一樣是自動安裝,分配步驟的速度主要取決於網絡環境,耐心等待便可:
等待檢查完成便可:
Cloudera 建議將 /proc/sys/vm/swappiness
設置爲最大值 10。當前設置爲 30。使用 sysctl
命令在運行時更改該設置並編輯 /etc/sysctl.conf
,以在重啓後保存該設置。您能夠繼續進行安裝,但 Cloudera Manager 可能會報告您的主機因爲交換而運行情況不良。
臨時修改:
sysctl vm.swappiness=10 cat /proc/sys/vm/swappiness
這裏咱們的修改已經生效,可是若是咱們重啓了系統,又會變成30.
永久修改:
在/etc/sysctl.conf
文件裏添加以下參數:
vm.swappiness=10
或者:
echo 'vm.swappiness=10'>> /etc/sysctl.conf
已啓用透明大頁面壓縮,可能會致使重大性能問題。請運行echo never > /sys/kernel/mm/transparent_hugepage/defrag
和echo never > /sys/kernel/mm/transparent_hugepage/enabled
以禁用此設置,而後將同一命令添加到 /etc/rc.local
等初始化腳本中,以便在系統重啓時予以設置。
安裝上面的提示執行便可;
這裏我選擇自定義服務,Zookeeper, HDFS,Yarn:
能夠先安裝基礎組件,而後用到啥在安裝啥若是全部服務都安裝,可能安裝過程當中會出現不少問題
CDH會自動給出一個角色分配,若是以爲不合理,咱們能夠手動調整一下,注意角色分配均衡:
Java HotSpot(TM) 64-Bit Server VM warning: ignoring option PermSize=128m; support was removed in 8.0 Exception in thread "main" java.lang.UnsatisfiedLinkError: Could not load library. Reasons: [Can't load library: /run/cloudera-scm-agent/process/68-cloudera-mgmt-SERVICEMONITOR/libleveldbjni-1.8.so, Can't load library: /run/cloudera-scm-agent/process/68-cloudera-mgmt-SERVICEMONITOR/libleveldbjni.so, no leveldbjni64-1.8 in java.library.path, no leveldbjni-1.8 in java.library.path, no leveldbjni in java.library.path, /run/cloudera-scm-agent/process/68-cloudera-mgmt-SERVICEMONITOR/libleveldbjni-64-1-4792575431304239050.8: libstdc++.so.6: : , /tmp/libleveldbjni-64-1-6079277982211108711.8: libstdc++.so.6: : ]
出錯緣由:當前節點的glibc升級有關。既然不存在leveldbjni的庫,那便給他安裝一個。
安裝leveldbjni庫的方式很是有趣:
1) 首先下載leveldbjni-all-1.8.jar
2)解壓該jar包,在META-INFnativelinux64目錄下找到libleveldbjni.so文件
3) 將libleveldbjni.so文件上傳到1中java.library.path中
若是由於其餘緣由,須要卸載Cloudera Manager,在各節點執行以下步驟便可。
systemctl stop cloudera-scm-server systemctl stop cloudera-scm-agent yum -y remove 'cloudera-manager-*' yum clean all umount cm_processes umount /var/run/cloudera-scm-agent/process rm -Rf /usr/share/cmf /var/lib/cloudera* /var/cache/yum/cloudera* /var/log/cloudera* /var/run/cloudera* rm -rf /tmp/.scmpreparenode.lock rm -Rf /var/lib/flume-ng /var/lib/hadoop* /var/lib/hue /var/lib/navigator /var/lib/oozie /var/lib/solr /var/lib/sqoop* /var/lib/zookeeper rm -rf /var/lib/hadoop-* /var/lib/impala /var/lib/solr /var/lib/zookeeper /var/lib/hue /var/lib/oozie /var/lib/pgsql /var/lib/sqoop2 /data/dfs/ /data/impala/ /data/yarn/ /dfs/ /impala/ /yarn/ /var/run/hadoop-*/ /var/run/hdfs-*/ /usr/bin/hadoop* /usr/bin/zookeeper* /usr/bin/hbase* /usr/bin/hive* /usr/bin/hdfs /usr/bin/mapred /usr/bin/yarn /usr/bin/sqoop* /usr/bin/oozie /etc/hadoop* /etc/zookeeper* /etc/hive* /etc/hue /etc/impala /etc/sqoop* /etc/oozie /etc/hbase* /etc/hcatalog systemctl stop mariadb yum -y remove mariadb-* rm -rf /var/lib/mysql rm -rf /var/log/mysqld.log rm -rf /usr/lib64/mysql rm -rf /usr/share/mysql rm -rf /opt/cloudera
ParcelUpdateService:com.cloudera.parcel.components.ParcelDownloaderImpl: Unable to retrieve remote parcel repository manifest
2019-08-27 20:35:50,469 INFO WebServerImpl:com.cloudera.server.cmf.WebServerImpl: Started Jetty server. 2019-08-27 20:35:50,600 ERROR ParcelUpdateService:com.cloudera.parcel.components.ParcelDownloaderImpl: Unable to retrieve remote parcel repository manifest java.util.concurrent.ExecutionException: java.net.UnknownHostException: archive.cloudera.com: Name or service not known at com.ning.http.client.providers.netty.future.NettyResponseFuture.abort(NettyResponseFuture.java:231) at com.ning.http.client.providers.netty.request.NettyRequestSender.abort(NettyRequestSender.java:422) at com.ning.http.client.providers.netty.request.NettyRequestSender.sendRequestWithNewChannel(NettyRequestSender.java:290) at com.ning.http.client.providers.netty.request.NettyRequestSender.sendRequestWithCertainForceConnect(NettyRequestSender.java:142) at com.ning.http.client.providers.netty.request.NettyRequestSender.sendRequest(NettyRequestSender.java:117) at com.ning.http.client.providers.netty.NettyAsyncHttpProvider.execute(NettyAsyncHttpProvider.java:87) at com.ning.http.client.AsyncHttpClient.executeRequest(AsyncHttpClient.java:506) at com.ning.http.client.AsyncHttpClient$BoundRequestBuilder.execute(AsyncHttpClient.java:229) at com.cloudera.parcel.components.ParcelDownloaderImpl.getRepositoryInfoFuture(ParcelDownloaderImpl.java:592) at com.cloudera.parcel.components.ParcelDownloaderImpl.getRepositoryInfo(ParcelDownloaderImpl.java:544) at com.cloudera.parcel.components.ParcelDownloaderImpl.syncRemoteRepos(ParcelDownloaderImpl.java:357) at com.cloudera.parcel.components.ParcelDownloaderImpl$1.run(ParcelDownloaderImpl.java:464) at com.cloudera.parcel.components.ParcelDownloaderImpl$1.run(ParcelDownloaderImpl.java:459) at com.cloudera.cmf.persist.ReadWriteDatabaseTaskCallable.call(ReadWriteDatabaseTaskCallable.java:36) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:748) Caused by: java.net.UnknownHostException: archive.cloudera.com: Name or service not known at java.net.Inet4AddressImpl.lookupAllHostAddr(Native Method) at java.net.InetAddress$2.lookupAllHostAddr(InetAddress.java:928) at java.net.InetAddress.getAddressesFromNameService(InetAddress.java:1323) at java.net.InetAddress.getAllByName0(InetAddress.java:1276) at java.net.InetAddress.getAllByName(InetAddress.java:1192) at java.net.InetAddress.getAllByName(InetAddress.java:1126) at java.net.InetAddress.getByName(InetAddress.java:1076) at com.ning.http.client.NameResolver$JdkNameResolver.resolve(NameResolver.java:28) at com.ning.http.client.providers.netty.request.NettyRequestSender.remoteAddress(NettyRequestSender.java:358) at com.ning.http.client.providers.netty.request.NettyRequestSender.connect(NettyRequestSender.java:369) at com.ning.http.client.providers.netty.request.NettyRequestSender.sendRequestWithNewChannel(NettyRequestSender.java:283) ... 15 more
不影響使用
安裝hive報錯:org.apache.hadoop.hive.metastore.HiveMetaException: Failed to retrieve schema tables from Hive Metastore DB,Not supported
[root@master01 ~]# rpm -qa|grep mysql-connector-java mysql-connector-java-5.1.25-3.el7.noarch
jdbc版本不對,要求使用5.1.26以上版本的jdbc驅動