So finally we get to do Hadoop installation for single node.
-SSH server: Master communicate with the slaves with the SSH. So we need to get SSH server running. We need to generate public/private key pair. Than we take that key, that we generate on the master and send it to all our nodes.
-Java 6 or greater
-Download and install hadoop
-add paths to bashrc
-Configure hadoop environment
(env.sh)-> java.home/disable ipv6
-Configure.XML ( core -site ! mapred-site)
-Launch Hadoop Daemons
Now i will show you installation
Lets do SSH to the name node
lets start going with our prerequisites. The first prerequisite is to generate public-private key.
No problem, we just need to type the command
now it will ask for password. we are keeping the password blank
There we go, we have our key generated.
Now lets get JAVA installed.
We need to type the command
# sudo apt-get install openjdk-7-jdk
It will ask that do you want to download 218MB package, just say yes. So Java Installed !!
Now we shift focus on Hadoop.
First we download the hadoop from the website. We can download using command line or browser. Once its downloaded, we move to next course of action that to unzip the .tar file
#tar -zxvf hadoop-1.2.0-bin.tar.gz
now we copy this to user local area and name it hadoop
next is to add paths to bashrc and configure env.sh
#Sudo vi $HOME/.bashrc
go to the bottom of the file and insert the following entries and save
then run the command
This will run the command
now to configure the env.sh. We type the command
#sudo vi /usr/local/hadoop/conf/hadoop-env.sh
we go to 7th line from the top and remove the #
and edit the line to replace with our type of java
and disable IPV6 on the same file
Now we have to configure those XML files.
#Sudo vi /usr/local/hadoop/conf/core-site.xml
#Sudo vi /usr/local/hadoop/conf/mapred-site.xml
this is the last step.
#hadoop namenode -format
then we say
Now we start Hadoop daemon and we are done