|
|
--hadoop文件位置：
log目录 :
| 2 | /var/log/hadoop-mapreduce |
安装目录：
启动命令目录：
配置文件目录：
环境变量：
让/etc/profile文件修改后立即生效 ,可以使用如下命令:
node配置：
资源库位置 ：
1.本地cdh4 搭建/
1、安装httpd服务,作用：用browser查看本地库
--安装httpd
--启动httpd
vi /etctpdtpd.conf (配置文件)
--关闭防火墙
--浏览器查看
http://ip地址
2、安装createrepo
--Createrepo用于创建本地源。
| 1 | sudo yum install yum-utils createrepo |
3、搭建本地源
--下载cloudera-manager.repo,cloudera-cdh4.repo到/etc/yum.repos.d/目录下
| 2 | wget http://archive.cloudera.com/cm4/redhat/6/x86_64/cm/cloudera-manager.repo |
| 3 | wget http://archive.cloudera.com/cdh4/redhat/6/x86_64/cdh/cloudera-cdh4.repo |
--将远程源的RPM包同步到本地的httpd服务目录
| 2 | reposync -r cloudera-manager |
| 3 | reposync -r cloudera-cdh4 |
--生成或刷新：RPMS，和一个放依赖关系等信息的文件夹repodata。
| 1 | cd /var/www/html/cloudera-manager |
| 3 | cd /var/www/html/cloudera-cdh4 |
至此，CDH4本地源搭建完成。
4:本地源的使用
--修改cloudera-manager.repo中的baseurl修改指向本地局域网的服务器
将baseurl=http://archive.cloudera.com/cm4/redhat/6/x86_64/cm/4/ 改为：
baseurl=http://172.16.2.52/cloudera-manager/
--修改cloudera-cdh4.repo中的baseurl修改指向本地局域网的服务器
将baseurl=http://archive.cloudera.com/cdh4/redhat/6/x86_64/cdh/4/ 改为：
baseurl=http://172.16.2.52/cloudera-cdh4/
baseurl的地址是服务器地址+放置RPM包的目录名，结构如下图：里面有一个放RPM包的文件夹RPMS，和一个放依赖关系等信息的文件夹repodata。
2.cloudera manager安装
--安装
| 1 | sudo yum install cloudera-manager-daemons |
| 2 | sudo yum install cloudera-manager-server |
| 3 | sudo yum install cloudera-manager-server-db |
| 4 | sudo yum install cloudera-manager-agent |
--启动
| 1 | service cloudera-scm-server start |
| 2 | service cloudera-manager-server-db start |
| 3 | service cloudera-scm-agent start |
3.cdh4 安装
//安装 装CDH4
| 1 | yum install hadoop-0.20-mapreduce-jobtracker hadoop-0.20-mapreduce-tasktracker |
| 2 | yum install hadoop-hdfs-namenode |
| 3 | yum install hadoop-hdfs-datanode |
| 4 | yum install hadoop-0.20-conf-pseudo |
| 5 | sudo -u hdfs hdfs namenode -format |
--启动hdfs
| 1 | for x in `cd /etc/init.d ; ls hadoop-hdfs-*` ; do sudo service $x start ; done |
4..安装 hive
参考资料：http://heylinux.com/archives/2456.html
--安装hive
| 1 | sudo yum install hive hive-metastore hive-server |
--安装mysql
| 1 | sudo yum install mysql mysql-server mysql-devel |
--安装mysql连接驱动
| 1 | sudo yum install mysql-connector-java |
| 2 | ln -s /usr/share/java/mysql-connector-java.jar /usr/lib/hive/lib/mysql-connector-java.jar |
--设置root密码
| 1 | sudo /usr/bin/mysql_secure_installation |
根据提示输入密码
| 02 | 02 Enter current password for root (enter for none): |
| 03 | 03 OK, successfully used password, moving on... |
| 05 | 05 Set root password? [Y/n] y |
| 06 | 06 New password:hadoophive |
| 07 | 07 Re-enter new password:hadoophive |
| 08 | 08 Remove anonymous users? [Y/n] Y |
| 10 | 10 Disallow root login remotely? [Y/n] N |
| 12 | 12 Remove test database and access to it [Y/n] Y |
| 14 | 14 Reload privilege tables now? [Y/n] Y |
--进入mysql 依次执行
| 01 | mysql -u root -p hadoophive |
| 02 | CREATE DATABASE metastore; |
| 04 | SOURCE /usr/lib/hive/scripts/metastore/upgrade/mysql/hive-schema-0.10.0.mysql.sql; |
| 05 | CREATE USER 'hive'@'%' IDENTIFIED BY 'hadoophive'; |
| 06 | CREATE USER 'hive'@'localhost' IDENTIFIED BY 'hadoophive'; |
| 07 | GRANT ALL PRIVILEGES ON metastore.* TO 'hive'@'%'; |
| 08 | GRANT ALL PRIVILEGES ON metastore.* TO 'hive'@'localhost'; |
--启动hive
--修改配置文件hive-site.xml
sudo vim /etc/hive/conf/hive-site.xml参考 hive 配置
| 03 | <!-- Hive Configuration can either be stored in this file or in the hadoop configuration files --> |
| 04 | <!-- that are implied by Hadoop setup variables. --> |
| 05 | <!-- Aside from Hadoop setup variables - this file is provided as a convenience so that Hive --> |
| 06 | <!-- users do not have to edit hadoop configuration files (that may be managed as a centralized --> |
| 09 | <!-- Hive Execution Parameters --> |
| 12 | <name>javax.jdo.option.ConnectionURL</name> |
| 13 | <value>jdbc:mysql://localhost:3306/metastore</value> |
| 14 | <description>JDBC connect string for a JDBC metastore</description> |
| 18 | <name>javax.jdo.option.ConnectionDriverName</name> |
| 19 | <value>com.mysql.jdbc.Driver</value> |
| 20 | <description>Driver class name for a JDBC metastore</description> |
| 23 | <name>javax.jdo.option.ConnectionUserName</name> |
| 27 | <name>javax.jdo.option.ConnectionPassword</name> |
| 28 | <value>hadoophive</value> |
| 31 | <name>hive.metastore.uris</name> |
| 32 | <value>thrift://127.0.0.1:9083</value> |
| 33 | <description>IP address (or fully-qualified domain name) and port of the metastore host</description> |
--启动hive
| 1 | service hive-metastore start |
| 2 | service hive-server start |
--依次执行
| 01 | $ sudo -u hdfs hadoop fs -mkdir /user/hive |
| 02 | $ sudo -u hdfs hadoop fs -chown hive /user/hive |
| 03 | $ sudo -u hdfs hadoop fs -mkdir /tmp |
| 04 | $ sudo -u hdfs hadoop fs -chmod 777 /tmp |
| 05 | $ sudo -u hdfs hadoop fs -chmod o+t /tmp |
| 06 | $ sudo -u hdfs hadoop fs -mkdir /data |
| 07 | $ sudo -u hdfs hadoop fs -chown hdfs /data |
| 08 | $ sudo -u hdfs hadoop fs -chmod 777 /data |
| 09 | $ sudo -u hdfs hadoop fs -chmod o+t /data |
| 10 | $ sudo chown -R hive:hive /var/lib/hive |
$ sudo vim /tmp/kv1.txt
| 1 | CREATE TABLE IF NOT EXISTS pokes ( foo INT,bar STRING ) ROW FORMAT DELIMITED FIELDS TERMINATED BY "," LINES TERMINATED BY "\n"; |
| 4 | LOAD DATA LOCAL INPATH '/tmp/kv1.txt' OVERWRITE INTO TABLE pokes; |
5.安装impala
参考资料：http://heylinux.com/archives/2456.html
--下载资源包
| 1 | sudo wget http://archive.cloudera.com/impala/redhat/6/x86_64/impala/cloudera-impala.repo |
--同步资源库
| 2 | reposync -r cloudera-impala |
| 3 | cd /var/www/html/cloudera-impala |
--修改cloudera-impala.repo
将baseurl=http://archive.cloudera.com/impala/redhat/6/x86_64/impala/1/改为：
baseurl=http://172.16.2.52/cloudera-impala/
--安装
| 2 | sudo yum install impala-shell |
| 3 | sudo yum install impala-server |
| 4 | sudo yum install impala-state-store |
--添加配置/etc/hadoop/conf/hdfs-site.xml如下
sudo vim /etc/hadoop/conf/hdfs-site.xml
| 02 | <name>dfs.client.read.shortcircuit</name> |
| 06 | <name>dfs.domain.socket.path</name> |
| 07 | <value>/var/run/hadoop-hdfs/dn._PORT</value> |
| 10 | <name>dfs.client.file-block-storage-locations.timeout</name> |
| 14 | <name>dfs.datanode.hdfs-blocks-metadata.enabled</name> |
--复制文件到/etc/impala/conf/
| 1 | sudo cp -rpa /etc/hadoop/conf/core-site.xml /etc/impala/conf/ |
| 2 | sudo cp -rpa /etc/hadoop/conf/hdfs-site.xml /etc/impala/conf/ |
| 3 | sudo cp -rpa /etc/hive/conf/hive-site.xml /etc/impala/conf/ |
--重新启动datanode 及 impala
| 1 | sudo service hadoop-hdfs-datanode restart |
| 2 | sudo service impala-state-store restart |
| 3 | sudo service impala-server restart |
--进入impala命令行
5.安装hbase（伪分布）
--安装zookper
| 1 | yum install zookeeper-server |
--启动zookper
| 1 | service zookeeper-server start |
--安装hbase，hbase-master，hbase-regionserver
| 2 | sudo yum install hbase-master |
| 3 | sudo yum install hbase-regionserver |
--修改配置
sudo vim /etc/security/limits.conf
--修改配置
$sudo vim /etc/pam.d/common-session
| 1 | session required pam_limits.so |
--修改hadoop配置
sudo vim /etc/hadoop/conf/hdfs-site.xml
| 2 | <name>dfs.datanode.max.xcievers</name> |
--修改hbase配置/etc/hbase/conf/hbase-site.xml
| 03 | <name>hbase.rootdir</name> |
| 04 | <value>hdfs://myhost:8020/hbase</value> |
| 07 | <name>hbase.cluster.distributed</name> |
--创建hbase工作目录
| 1 | sudo -u hdfs hadoop fs -mkdir /hbase |
| 2 | sudo -u hdfs hadoop fs -chown hbase /hbase |
--hdfs重启
| 1 | sudo /etc/init.d/hadoop-hdfs-namenode restart |
| 2 | sudo /etc/init.d/hadoop-hdfs-datanode restart |
--启动
| 1 | sudo service hbase-master start |
| 2 | sudo service hbase-regionserver start |
--进入hbase
|
|
QQ群⑧：
窥视卡
雷达卡
发表于 2016-12-10 10:24:28
提升卡
置顶卡
沉默卡
喧嚣卡
变色卡
千斤顶
显身卡