CentOS 7 Hadoop installation Configuration

I use two computers for the configuration of the cluster, if it is a single machine, there may be some problems. First, set the hostname root permissions on the two computer to open the/etc/host file

Set Hostname,root permissions again to open/etc/hostname file settings

From the machine set to Slaver.hadoop

1. Install the Java JDK and configure the environment

CentOS comes with a JDK installed, and if we want to install the JDK ourselves and configure the environment, first uninstall the original JDK

In command-line mode, enter command: Yum remove Java (root user, non-root user, please add sudo before the command).

Next install the JDK and configure the environment, download the JDK rpm package online, using the command: RPM-IVH JDK-......;

To configure the JDK environment:

The root user opens the/etc/profile file at the end of the file by adding a JDK path similar to the following configuration, just pay attention to the JDK version

　　

The above section is a Java environment variable, followed by a Hadoop variable, and so on.

2. Install Hadoop and configure the environment.

Download the hadoop-2.6.0 version to the Apache website.

Extract files: tar-xvf hadoop-2.6.0.tar.gz (Unzip to the path you need to install, or set the path again)

Go to the extracted directory and find the following configuration files under the hadoop-2.6.0/etc/hadoop/path

　　

Core.site.xml:

　　

Hdfs-site.xml

　　

Mapred-site.xml.template

　　

hadoop-env.sh

　　

Slaves

192.168.1.126 slave IP

To complete this configuration, start the tutorial on the reference web.

/hadoop-2.6.0/bin/hadoop Namenode-format

/hadoop-2.6.0/sbin/start-all.sh

Complete.

Note: You need to shut down the firewall of two computers before starting: sudo systemctl stop Iptables.service

3. Installing the configuration Hadoop Eclipse plug-in

Download the Hadoop Eclipse plugin and put it in the plugins folder in the Eclipse folder to start Eclipse.

Eclipse Window->open persperctive->other->

　　

Open Map/reduce and Resource respectively. Then create a new M/R project in Eclipse, find Wordcount.java source code execution,

Set parameters for run as configuertion in Eclipse:

　　

Project directory in eclipse:

　　

Execution Result:

　　

　　

　　

CentOS 7 Hadoop installation Configuration