Ubuntu12.04安装Hive(二)

来源:互联网 发布:软件生命周期v模型 编辑:程序博客网 时间:2024/06/16 10:21

This post is about the installation of Hive on Ubuntu. The environment of my system is ubuntu12.04, hadoop1.0.3, hive 0.9.0. Supposing the hadoop is installed.

The details of installation for hive describe as follows:

step 1: download hive-0.9.0.tar.gz from the url: http://hive.apache.org/index.html

step 2: uncompress the above file using linux command in the shell.

[plain] view plaincopy
  1. $ sudo tar -xzf hive-0.10.0.tar.gz  

step 3: copy the uncompressed files to the directory /home/hadoop/.

[html] view plaincopy
  1. $ sudo cp -r /home/eboss/hive /usr/local/hadoop/
  2. $ mvhive-0.10.0  hive

step 4: entry the configure directory of hive and copy the configure file.

[html] view plaincopy
  1. $ cd /usr/local/hadoop/hive/conf  
  2.   
  3. $ sudo cp hive-env.sh.template hive-env.sh  

step 5: modify the configure file in the last line as follows:

[html] view plaincopy
  1. $ sudo gedit /usr/local/hadoop/hive/conf/hive-env.sh  
  2.   
  3. #add the below configure information  
  4.   
  5. export Hadoop_HOME=/usr/local/hadoop/
  6.   
  7. export HIVE_HOME=/usr/local/hadoop/hive 
  8.   
  9. export PATH=$PATH:/usr/local/hadoop/hive/bin  

step 6: change the directory of hive to the hadoop user.

[html] view plaincopy
  1. $ sudo chown -R hadoop:hadoop hive  

step 7: start hive and test hive

[html] view plaincopy
  1. $ bin/hive or bin/hive start or ./hive  
  2.   
  3. $ show tables;  
  4.   
  5. $ create database hive;  
  6.   
  7. $ use hive;  
  8.   
  9. $ create table test(id int,name string);  

step 8: modify hive-exec-log4j.properties.template

[html] view plaincopy
  1.  mv conf/hive-exec-log4j.properties.template  conf/hive-exec-log4j.properties
  2. 如果不修改此文件的话 启动hive时会出现警告信息的

step 8: 配置hive支持mysql数据的导入

[html] view plaincopy
  1.  要把mysql-connector-java-5.1.22-bin.jar 复制到 hive/lib目录下
  2. 如果不修改此文件的话 启动hive时会出现警告信息的







444