Spark 1.6.1 cluster mode installation on ubuntu 14.04

posted on Nov 20th, 2016

Apache Spark

Apache Spark is an open source cluster computing framework. Originally developed at the University of California, Berkeley's AMPLab, the Spark codebase was later donated to the Apache Software Foundation, which has maintained it since. Spark provides an interface for programming entire clusters with implicit data parallelism and fault-tolerance.

Pre Requirements

1) A machine with Ubuntu 14.04 LTS operating system installed.

2) Apache Spark 1.6.1 Software (Download Here)

3) Scala 2.10.5 Software (Download Here)

NOTE

Spark is not a replacement of Hadoop. Spark is a part of the hadoop eco system. Spark can use Hadoop's distributed file system (HDFS) and also submit jobs on YARN. In order to make use of hadoop's components, you need to install Hadoop first then spark (How to install Hadoop on Ubuntu 14.04). The downloaded spark must be compatible with hadoop version. Please notice hadoop version before downloading spark.

Spark Cluster Mode Installation on Ubuntu 14.04

Spark Cluster Mode Installation

This post descibes how to install and configure spark clusters ranging from a few nodes to extremely large clusters. To play with Spark, you may first want to install it on a single machine (see, Standalone Mode Setup).

Spark Cluster Mode Installation on Ubuntu 14.04

On All machines - (Sparkmaster, Sparkslave1, Sparkslave2, Sparkslave3)

Installation Steps

Step 1 - Update. Open a terminal (CTRL + ALT + T) and type the following sudo command. It is advisable to run this before installing any package, and necessary to run it to install the latest updates, even if you have not added or removed any Software Sources.

$ sudo apt-get update

Step 2 - Installing Java 7.

$ sudo apt-get install openjdk-7-jdk

Step 3 - Install open-ssh server. It is a cryptographic network protocol for operating network services securely over an unsecured network. The best known example application is for remote login to computer systems by users.

$ sudo apt-get install openssh-server

Step 4 - Create a Group. We will create a group, configure the group sudo permissions and then add the user to the group. Here 'hadoop' is a group name and 'hduser' is a user of the group.

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

Step 5 - Configure the sudo permissions for 'hduser'.

$ sudo visudo

Since by default ubuntu text editor is nano we will need to use CTRL + O to edit.

ctrl+O

Add the permissions to sudoers.

hduser ALL=(ALL) ALL

Use CTRL + X keyboard shortcut to exit out. Enter Y to save the file.

ctrl+x

Step 6 - Edit /etc/hosts file.

$ sudo gedit /etc/hosts

/etc/hosts file. Add all machines IP address and hostname. Save and close.

10.0.0.1	sparkmaster
10.0.0.2	sparkslave1
10.0.0.3	sparkslave2
10.0.0.4	sparkslave3

Step 7 - Change the ownership and permissions of the directory /usr/local/spark. Here 'hduser' is an Ubuntu username.

$ sudo chown -R hduser /usr/local/spark
$ sudo chmod -R 755 /usr/local/spark

Step 8 - Creating scala directory.

$ sudo mkdir /usr/local/scala

Step 9 - Change the ownership and permissions of the directory /usr/local/scala. Here 'hduser' is an Ubuntu username.

$ sudo chown -R hduser /usr/local/scala
$ sudo chmod -R 755 /usr/local/scala

Step 10 - Creating /app/spark/tmp directory.

$ sudo mkdir /app/spark/tmp

Step 11 - Change the ownership and permissions of the directory /app/spark/tmp. Here 'hduser' is an Ubuntu username.

$ sudo chown -R hduser /app/spark/tmp
$ sudo chmod -R 755 /app/spark/tmp

Step 12 - Switch User, is used by a computer user to execute commands with the privileges of another user account.

$ su hduser

Step 13 - Generating a new SSH public and private key pair on your local computer is the first step towards authenticating with a remote server without a password. Unless there is a good reason not to, you should always authenticate using SSH keys.

$ ssh-keygen -t rsa -P ""

Step 14 - Now you can add the public key to the authorized_keys

$ cat $HOME/.ssh/id_rsa.pub >> $HOME/.ssh/authorized_keys

Step 15 - Adding hostname to list of known hosts. A quick way of making sure that 'hostname' is added to the list of known hosts so that a script execution doesn't get interrupted by a question about trusting computer's authenticity.

$ ssh hostname 

Only on Sparkmaster Machine

Step 16 - Change the directory to /home/hduser/Desktop , In my case the downloaded spark-1.6.1-bin-hadoop2.6.tgz file is in /home/hduser/Desktop folder. For you it might be in /downloads folder check it.

$ cd /home/hduser/Desktop/

Step 17 - Untar the spark-1.6.1-bin-hadoop2.6.tgz file.

$ tar xzf spark-1.6.1-bin-hadoop2.6.tgz

Step 18 - Move the contents of spark-1.6.1-bin-hadoop2.6 folder to /usr/local/spark

$ mv spark-1.6.1-bin-hadoop2.6/* /usr/local/spark

Step 19 - Untar the scala-2.10.5.tgz file. In my case the downloaded scala-2.10.5.tgz file is in /home/hduser/Desktop folder. For you it might be in /downloads folder check it.

$ tar xzf scala-2.10.5.tgz

Step 20 - Move the contents of scala-2.10.5 folder to /usr/local/scala

$ mv scala-2.10.5/* /usr/local/scala

Step 21 - Edit $HOME/.bashrc file by adding the spark and scala path.

$ sudo gedit $HOME/.bashrc

$HOME/.bashrc file. Add the following lines

export SCALA_HOME=/usr/local/scala
export SPARK_HOME=/usr/local/spark
export PATH=$SPARK_HOME/bin:$JAVA_HOME/bin:$SCALA_HOME/bin:$PATH

Step 22 - Reload your changed $HOME/.bashrc settings

$ source $HOME/.bashrc

Step 23 - Change the directory to /usr/local/spark/conf

$ cd /usr/local/spark/conf

Step 24 - Copy the spark-env.sh.template to spark-env.sh

$ cp spark-env.sh.template spark-env.sh

Step 25 - Edit spark-env.sh file

$ sudo gedit spark-env.sh

Step 26 - Add the below lines to spark-env.sh file. Save and Close.

export SCALA_HOME=/usr/local/scala
export JAVA_HOME=/usr/lib/jvm/java-7-openjdk-amd64
export SPARK_WORKER_MEMORY=1g
export SPARK_WORKER_INSTANCES=2
export SPARK_MASTER_IP=10.0.0.1
export SPARK_MASTER_PORT=7077
export SPARK_WORKER_DIR=/app/spark/tmp

Step 27 - Copy the spark-defaults.conf.template to spark-defaults.conf

$ cp spark-defaults.conf.template spark-defaults.conf

Step 28 - Edit spark-defaults.conf file

$ sudo gedit spark-defaults.conf

Step 29 - Add the below line to spark-defaults.conf file. Save and Close.

spark.master                     spark://10.0.0.1:7077

Step 30 - Copy the slaves.template to slaves

$ cp slaves.template slaves

Step 31 - Edit slaves file.

$ sudo gedit slaves

Step 32 - Add the below line to slaves file. Save and Close.

10.0.0.2
10.0.0.3
10.0.0.4

Step 33 - ssh-copy-id is a small script which copy your ssh public-key to a remote host; appending it to your remote authorized_keys.

$ ssh-copy-id -i $HOME/.ssh/id_rsa.pub hduser@10.0.0.2

Step 34 - ssh is a program for logging into a remote machine and for executing commands on a remote machine. Check remote login works or not.

$ ssh 10.0.0.2

Step 35 - Exit from remote login.

$ exit 

Same steps 33, 34 and 35 for other machines (Sparkslave2, Sparkslave3).

$ ssh-copy-id -i $HOME/.ssh/id_rsa.pub hduser@10.0.0.3
$ ssh 10.0.0.3
$ exit 

$ ssh-copy-id -i $HOME/.ssh/id_rsa.pub hduser@10.0.0.4
$ ssh 10.0.0.4
$ exit 

Step 36 - Secure copy or SCP is a means of securely transferring computer files between a local host and a remote host or between two remote hosts. Here we are transferring configured spark files from master to slave nodes.

$ scp -r /usr/local/spark/* hduser@10.0.0.2:/usr/local/spark
$ scp -r /usr/local/spark/* hduser@10.0.0.3:/usr/local/spark
$ scp -r /usr/local/spark/* hduser@10.0.0.4:/usr/local/spark

Step 37 - Here we are transferring scala files from master to slave nodes.

$ scp -r /usr/local/scala/* hduser@10.0.0.2:/usr/local/scala
$ scp -r /usr/local/scala/* hduser@10.0.0.3:/usr/local/scala
$ scp -r /usr/local/scala/* hduser@10.0.0.4:/usr/local/scala

Step 38 - Here we are transferring configured .bashrc file from master to slave nodes.

$ scp -r $HOME/.bashrc hduser@10.0.0.2:$HOME/.bashrc
$ scp -r $HOME/.bashrc hduser@10.0.0.3:$HOME/.bashrc
$ scp -r $HOME/.bashrc hduser@10.0.0.4:$HOME/.bashrc

Step 39 - Change the directory to /usr/local/spark/sbin

$ cd /usr/local/spark/sbin

Step 40 - Start Master and all Worker Daemons.

$ ./start-all.sh

Step 41 - The JPS (Java Virtual Machine Process Status Tool) tool is limited to reporting information on JVMs for which it has the access permissions.

$ jps

Once the spark is up and running check the web-ui of the components as described below

http://10.0.0.1:8080/

Only on slave machines - (Sparkslave1, Sparkslave2, and Sparkslave3)

Step 42 - The JPS (Java Virtual Machine Process Status Tool) tool is limited to reporting information on JVMs for which it has the access permissions.

$ jps

Only on Sparkmaster Machine

Step 43 - Stop Master and all Worker Daemons.

$ ./stop-all.sh

Please share this blog post and follow me for latest updates on

facebook             google+             twitter             feedburner

Previous Post                                                                                          Next Post

Labels : Spark Standalone Mode Installation   Spark with YARN Configuration   Spark WordCount Java Example   Spark spark-submit Script Usage   Spark Shell Usage   Spark Shell Scala Example   Spark WordCount Scala Example