In the last two days, we have set up a Linux big data processing environment for a project. The system is centos 6.3. I mainly configure JDK, install Tomcat, hadoop, hbase, and zookeeper software. I am also a newbie in hadoop. I have encountered many problems in configuring this environment and checked a lot of information, here is a summary for future review.
First, you need to modify the account permissions. to install the software environment, you need to upload files and modify the permissions of some system files. Therefore, you 'd better set the permission to root.
Permission modification method: http://www.linuxidc.com/Linux/2012-03/55629.htm
There are many articles on software installation on the Internet:
Install JDK: http://www.cnblogs.com/zhoulf/archive/2013/02/04/2891608.html
Install Tomcat: http://www.cnblogs.com/zhoulf/archive/2013/02/04/2891633.html
Installing hadoop: http://www.cnblogs.com/xia520pi/archive/2012/05/16/2503949.html
Http://bjbxy.blog.51cto.com/854497/352692
Install hbase: http://blog.csdn.net/21aspnet/article/details/18776833
Installing zookeeper: http://www.cnblogs.com/ilovexiao/p/3946547.html
You need to download the correct version before installing hadoop and hbase. You can refer to the official documentation and select http://abloz.com/hbase/book.html. you can also see the configuration of multiple hbaseat the same time in the document.
When configuring hadoop and hbase, the host name, that is, the hostname, is very important. I modified the hostname on the originally configured server and encountered a series of problems, how to change all problems (heartbroken AH), and finally found in the modification of hostname need to clear the data below TMP, and then format hadoop namenode data, such as: http://www.07net01.com/storage_networking/hadoopanzhuangbushuyudaowenti_guzhangpaicha_90974_1363776317.html
You may encounter many problems when configuring the software environment. You can view the log files of hadoop, hbase, and zookeeper to find out the problem and make modifications and adjustments.
After the environment is configured, test the program and find that the data is stored on the server and then read out the garbled code. In Windows, the test is normal. This is the problem of the server encoding, you can change the server encoding to gb2312 or GBK, the specific modification method such as: http://blog.sina.com.cn/s/blog_697b96890101jdxj.html
If hadoop and hbase are not disabled before the server is restarted, after the server is restarted, hadoop enters safe mode and must be disabled before hbase and its commands can be used in Chongqing. Http://www.07net01.com/storage_networking/hadoopanzhuangbushuyudaowenti_guzhangpaicha_90974_1363776317.html
Install and configure hadoop, hbase, and zookeeper in centos