How to install Hadoop?

The guides I followed when I had 12.04 were:

• Hadoop 2 Deployment on Ubuntu
• Michael Noll's Hadoop (Single Node)
• MyLearning Hadoop on Ubuntu 12.04 (Single Node)

I was actually opposed to the MyLearning one because the first thing it recommended was Oracle Java 7 instead of OpenJDK 7, but I had some issues with OpenJDK 7 when trying this out so I had to go with Oracle.

The guide is mostly straight forward and here it is:

1. Install Java

   sudo add-apt-repository ppa:webupd8team/java  
   sudo apt-get update && sudo apt-get upgrade  
   sudo apt-get install oracle-java7-installer  
   

2. Create Hadoop user

   sudo addgroup hadoop  
   sudo adduser --ingroup hadoop hduser
   

   Where hduser is the Hadoop user you want to have.

3. Configuring SSH

   su - hduser   
   ssh-keygen -t rsa -P ""
   cat .ssh/id_rsa.pub >> .ssh/authorized_keys
   

   To be sure that SSH installation went well, you can open a new terminal and try to create ssh session using hduser by the following command:

   ssh localhost
   

   reinstall ssh if localhost does not connect (you may need to add hduser to sudo as below step)

   sudo apt-get install openssh-server
   

4. Edit Sudoers

   sudo visudo
   

5. Add at the end the line to add hduser into sudoers

   hduser ALL=(ALL:ALL) ALL
   

   To save press CTRL+X, type Y and press ENTER

6. Disable IPv6

   sudo gedit /etc/sysctl.conf
   

   or

   sudo nano /etc/sysctl.conf
   

   Copy the following lines at the end of the file:

   #disable ipv6  
   net.ipv6.conf.all.disable_ipv6 = 1  
   net.ipv6.conf.default.disable_ipv6 = 1   
   net.ipv6.conf.lo.disable_ipv6 = 1  
   

   If you face a problem telling you, you don't have permissions, just run the previous command with the root account (In case sudo is not enough. For me it was)

7. Now reboot.

   You can also do sudo sysctl -p but I rather reboot.

   After rebooting, check to make sure IPv6 is off:

   cat /proc/sys/net/ipv6/conf/all/disable_ipv6
   

   it should say 1. If it says 0, you missed something.

8. Installing Hadoop

   There are several ways of doing this, the one the Guide suggests is to download from the Apache Hadoop site and decompress the file in your hduser home folder. Rename the extracted folder to hadoop.

   The other way is to use a PPA that was tested for 12.04:

   sudo add-apt-repository ppa:hadoop-ubuntu/stable  
   sudo apt-get update && sudo apt-get upgrade  
   sudo apt-get install hadoop  
   

   NOTE: The PPA may work for some and for others will not. The one I tried was to download from the official site because I did not know about the PPA.

9. Update $HOME/.bashrc

   You will need to update the .bashrc for hduser (and for every user you need to administer Hadoop). To open .bashrc file, you will need to open it as root:

   sudo gedit /home/hduser/.bashrc  
   

   or

   sudo nano /home/hduser/.bashrc  
   

   Then you will add the following configurations at the end of .bashrc file

   # Set Hadoop-related environment variables   
   export HADOOP_HOME=/home/hduser/hadoop  
   
   # Set JAVA_HOME (we will also configure JAVA_HOME directly for Hadoop later on)`
   export JAVA_HOME=/usr/lib/jvm/java-7-oracle  
   

   Now, if you have OpenJDK7, it would look something like this:

   export JAVA_HOME=/usr/lib/java-7-openjdk-amd64  
   

   The thing to watch out for in here is the folder where the Java resides with the AMD64 version. If the above does not work, you can try looking in that particular folder or setting the Java that will be in used with:

   sudo update-alternatives --config java  
   

   Now for some helpful alias:

   # Some convenient aliases and functions for running Hadoop-related commands  
   unalias fs &> /dev/null   
   alias fs="hadoop fs"    
   unalias hls &> /dev/null  
   alias hls="fs -ls"  
   
   # Add Hadoop bin/ directory to PATH  
   export PATH=$PATH:$HADOOP_HOME/bin  
   

10. Configuring Hadoop

    The following are configuration files we can use to do the proper configuration. Some of the files you will be using with Hadoop are (More information in this site):

    start-dfs.sh - Starts the Hadoop DFS daemons, the namenode and datanodes. Use this before start-mapred.sh

    stop-dfs.sh - Stops the Hadoop DFS daemons.

    start-mapred.sh - Starts the Hadoop Map/Reduce daemons, the jobtracker and tasktrackers.

    stop-mapred.sh - Stops the Hadoop Map/Reduce daemons.

    start-all.sh - Starts all Hadoop daemons, the namenode, datanodes, the jobtracker and tasktrackers. Deprecated; use start-dfs.sh then start-mapred.sh

    stop-all.sh - Stops all Hadoop daemons. Deprecated; use stop-mapred.sh then stop-dfs.sh

    But before we start using them, we need to modify several files in the /conf folder.

    hadoop-env.sh

    Look for the file hadoop-env.sh, we need to only update the JAVA_HOME variable in this file:

    sudo gedit /home/hduser/hadoop/conf/hadoop-env.sh
    

    or

    sudo nano /home/hduser/hadoop/conf/hadoop-env.sh
    

    or in the latest versions it will be in

    sudo nano /etc/hadoop/conf.empty/hadoop-env.sh
    

    or

    sudo nano /etc/hadoop/hadoop-env.sh
    

    Then change the following line:

    # export JAVA_HOME=/usr/lib/j2sdk1.5-sun
    

    To

    export JAVA_HOME=/usr/lib/jvm/java-7-oracle
    

    Note: if you get Error: JAVA_HOME is not set Error while starting the services, you forgot to uncomment the previous line (just remove #).

    core-site.xml

    Now we need to create a temp directory for Hadoop framework. If you need this environment for testing or a quick prototype (e.g. develop simple hadoop programs for your personal test ...), I suggest to create this folder under /home/hduser/ directory, otherwise, you should create this folder in a shared place under shared folder (like /usr/local ...) but you may face some security issues. But to overcome the exceptions that may caused by security (like java.io.IOException), I have created the tmp folder under hduser space.

    To create this folder, type the following command:

    sudo mkdir /home/hduser/tmp   
    

    Please note that if you want to make another admin user (e.g. hduser2 in hadoop group), you should grant him a read and write permission on this folder using the following commands:

    sudo chown hduser:hadoop /home/hduser/tmp  
    sudo chmod 755 /home/hduser/tmp  
    

    Now, we can open hadoop/conf/core-site.xml to edit the hadoop.tmp.dir entry. We can open the core-site.xml using text editor:

    sudo gedit /home/hduser/etc/hadoop/core-site.xml  
    

    or

    nano /home/hduser/etc/hadoop/core-site.xml
    

    Then add the following configurations between <configure> xml elements:

    <property>
      <name>hadoop.tmp.dir</name>
      <value>/home/hduser/tmp</value>
      <description>A base for other temporary directories.</description>
    </property>
    
    <property>
      <name>fs.default.name</name>
      <value>hdfs://localhost:54310</value>
      <description>The name of the default file system.  A URI whose
      scheme and authority determine the FileSystem implementation.  The
      uri's scheme determines the config property (fs.SCHEME.impl) naming
      the FileSystem implementation class.  The uri's authority is used to
      determine the host, port, etc. for a filesystem.</description>
    </property>
    

    Now edit mapred-site.xml

    <property>
      <name>mapred.job.tracker</name>
      <value>localhost:54311</value>
      <description>The host and port that the MapReduce job tracker runs
      at.  If "local", then jobs are run in-process as a single map
      and reduce task.
      </description>
    </property>
    

    Now edit hdfs-site.xml

    <property>
      <name>dfs.replication</name>
      <value>1</value>
      <description>Default block replication.
      The actual number of replications can be specified when the file is created.
      The default is used if replication is not specified in create time.
      </description>
    </property> 
    

11. Formatting NameNode

    Now you can start working on the Node. First format:

    ~/hduser/hadoop/bin/hadoop namenode -format
    

    or

    ./home/hduser/hadoop/bin/hadoop namenode -format
    

    You should format the NameNode in your HDFS. You should not do this step when the system is running. It is usually done once at first time of your installation.

12. Starting Hadoop Cluster

    You will need to navigate to hadoop/bin directory and run the ./start-all.sh script.

    cd ~/hduser/hadoop/bin/  
    start-all.sh  
    

    If you have a different version from the one shown in the guides (Which you will most likely have if doing this with the PPA or a newer version) then try it this way:

    cd ~/hduser/hadoop/bin/   
    start-dfs.sh  
    start-mapred.sh   
    

    This will start a Namenode, Datanode, Jobtracker and a Tasktracker on your machine.

13. Checking if Hadoop is running

    There is a nice tool called jps. You can use it to ensure that all the services are up. In your hadoop bin folder type:

    jps
    

    It should show you all Hadoop related processes.

    NOTE: Since this was done around 6 months ago for me, if there is any part not working let me know.

Hadoop Using Juju (A Juju Charm for Hadoop)

Taken from Charming Hadoop

I will assume the following is already set up:

• You have a server with Juju already set
• You have access to the server (Locally or remotely)
• You have Juju configured and ready to start adding charms
• You are using 12.04 (This is because I tested all of this with 12.04)
• You have already configure ~/.juju/environments.yaml with the information regarding the server you will be using including the PPA origin.

Ok now follow this steps to have a Hadoop service running:

1. Bootstrap the environment for Hadoop

   juju bootstrap
   

   Wait until it finishes then check to see if it is connecting correctly:

   juju status
   

2. Deploy Hadoop (Master and Slave)

   juju deploy --constraints="instance-type=m1.large" hadoop hadoop-master   
   
   juju deploy --constraints="instance-type=m1.medium" -n 5 hadoop hadoop-slave
   

3. Create Relations

   juju add-relation hadoop-master:namenode hadoop-slave:datanode  
   
   juju add-relation hadoop-master:jobtracker hadoop-slave:tasktracker
   

4. Expose Hadoop (Since you already deploy and created relations the service should be running)

   juju expose hadoop-master
   

   And check status to see if it working correctly:

   juju status hadoop-master
   

Up to now you have a running Hadoop. There are many more things you can do that can be found in the link provided or in the official Juju Charm for Hadoop

For up to date JuJu Charms (Setups, Step by Step guide and more) you can visit: JuJu Charms and make your own JuJu Environment and see how each file is setup and how each service connects.

I successfully installed Hadoop by setting the path of JAVA_HOME as usr/lib/jvm/java-6-openjdk-amd64.

Derived from @Luis Alvarado's answer, here is my version for Ubuntu 14.04 and Hadoop 2.5.1

In brief

1. Install Java
2. Prepare an executive user for Hadoop hduser
3. Switch to hduser from now on
4. Allow hduser to remote via ssh with pass-phrase-less
5. Disable IPv6
6. Download and config Hadoop package
7. Prepare system path $HADOOP_HOME and $JAVA_HOME
8. Config Hadoop's services
9. Start Hadoop's services

Done. Good luck!

Detail steps

Install Java

Download and install

$ sudo add-apt-repository ppa:webupd8team/java    
$ sudo apt-get update && sudo apt-get upgrade    
$ sudo apt-get install oracle-java7-installer

Make sure you have Java7 installed

$ which java
$ ls -l /usr/bin/java
$ ls -l /etc/alternatives/java

we should have java point to /usr/lib/jvm/java-7-oracle/jre/bin/java

Prepare an executive user for Hadoop hduser

Create user hduser in group hadoop

$ sudo addgroup hadoop  
$ sudo adduser --ingroup hadoop hduser

Grant hduser the sudo privilege

Edit sudo

$ sudo visudo

Add to the end this line

hduser ALL=(ALL:ALL) ALL

Switch to hduser from now on

$ su - hduser

Allow hduser to remote via ssh with pass-phrase-less

Install openssh

$ sudo apt-get install openssh-server

Generate RSA public/private key for SSH connection; passphrase is empty as parameter -P ""

$ ssh-keygen -t rsa -P ""
$ cat .ssh/id_rsa.pub >> .ssh/authorized_keys

Make sure hduser can ssh remote locally without a password

$ ssh localhost

Disable IPv6

Edit the configuration file

$ sudo nano /etc/sysctl.conf

Copy to the end

#disable ipv6
net.ipv6.conf.all.disable_ipv6 = 1
net.ipv6.conf.default.disable_ipv6 = 1   
net.ipv6.conf.lo.disable_ipv6 = 1

Make sure IPv6 is off by a reboot or call

$ sudo sysctl -p 

Then call

$ cat /proc/sys/net/ipv6/conf/all/disable_ipv6

It should say 1 which means OK ^^

Download and config Hadoop package

Download Hadoop 2.5.1 packages from Apache Hadoop site

The direct URL for this package is this link

http://www.eu.apache.org/dist/hadoop/core/hadoop-2.5.1/hadoop-2.5.1.tar.gz

So let's download to hduser 's home folder, extract it, and rename it to hadoop

$ wget http://www.eu.apache.org/dist/hadoop/core/hadoop-2.5.1/hadoop-2.5.1.tar.gz
$ tar -xvzf hadoop-2.5.1.tar.gz
$ mv hadoop-2.5.1 hadoop

Make sure we have Hadoop stored in hduser home

$ ls /home/hduser/hadoop

Prepare system path $HADOOP_HOME and $JAVA_HOME

Edit hduser 's .bashrc file

$ nano .bashrc

Put to the end values for $HADOOP_HOME and $JAVA_HOME

# Set Hadoop-related environment variables   
export HADOOP_HOME=/home/hduser/hadoop  

# Set JAVA_HOME (we will also configure JAVA_HOME directly for Hadoop later on)
export JAVA_HOME=/usr/lib/jvm/java-7-oracle 

Add the Hadoop binary folders to system $PATH

export PATH=$HADOOP_HOME/bin:$HADOOP_HOME/sbin:$PATH

Open a new terminal, log in as hduser, and make sure you have $HADOOP_HOME with available commands

$ echo $HADOOP_HOME
$ which start-all.sh
$ which start-dfs.sh
$ which start-yarn.sh

We should see the full path of those names.

Config Hadoop's services

Each component in Hadoop is configured using an XML file.

• Common properties go in core-site.xml

• HDFS properties go in hdfs-site.xml

• MapReduce properties go in mapred-site.xml

These files are all located in folder $HADOOP_HOME/etc/hadoop

Define, again, JAVA_HOME in hadoop-env.sh by edit the line

export JAVA_HOME=/usr/lib/jvm/java-7-oracle

Define Hadoop temp folder and file system name in core-site.xml at

<configuration>
  ...
  <property>
    <name>hadoop.tmp.dir</name>
    <value>/home/hduser/tmp</value>
    <description>A base for other temporary directories.</description>
  </property>

  <property>
    <name>fs.default.name</name>
    <value>hdfs://localhost:54310</value>
    <description>The name of the default file system.  A URI whose
    scheme and authority determine the FileSystem implementation.  The
    uri's scheme determines the config property (fs.SCHEME.impl) naming
    the FileSystem implementation class.  The uri's authority is used to
    determine the host, port, etc. for a filesystem.</description>
  </property>
  ...
</configuration>

We need to prepare this temp folder as configured at /home/hduser/tmp

$ cd /home/hduser
$ mkdir tmp
$ chown hduser:hadoop tmp
$ chmod 755 tmp

Define file system 's block replication in hdfs-site.xml

<configuration>
  ...
  <property>
    <name>dfs.replication</name>
    <value>1</value>
    <description>Default block replication.
    The actual number of replications can be specified when the file is created.
    The default is used if replication is not specified in create time.
    </description>
  </property>
  ...
</configuration>

Define map-reduce job in mapred-site.xml

<configuration>
  ...
  <property>
    <name>mapred.job.tracker</name>
    <value>localhost:54311</value>
    <description>The host and port that the MapReduce job tracker runs
    at.  If "local", then jobs are run in-process as a single map
    and reduce task.
    </description>
  </property>
  ...
</configuration>

Format name node

$ hdfs namenode -format

Start Hadoop service

Call

$ start-dfs.sh && start-yarn.sh

these two commands are located at $HADOOP_HOME/sbin which we have added to system $PATH before.

Make sure Hadoop services are started properly

$ jps

we should see

To be able to install sun-java with the apt-get command, you need to add a line to a file called sources.list. This file can be found in /etc/apt/sources.list.

Open the file using this command:

sudo nano /etc/apt/sources.list

Then at the very end of (bottom) of that file, you copy/paste the line:

deb http://us.archive.ubuntu.com/ubuntu/ hardy multiverse

Now press Ctrl+X to exit, and y for saving.

Now type the command:

sudo apt-get update 

And when that is done, you can successfully run the command:

sudo apt-get install sun-java6-jdk