Puppet is a Configuration Management tool that is used for deploying, configuring and managing servers. It performs the following functions:
  • Defining distinct configurations for each and every host, and continuously checking and confirming whether the required configuration is in place and is not altered (if altered Puppet will revert back to the required configuration) on the host.
  • Dynamic scaling-up and scaling-down of machines.
  • Providing control over all your configured machines, so a centralized (master-server or repo-based) change gets propagated to all, automatically.
Puppet uses a Master Slave architecture in which the Master and Slave communicate through a secure encrypted channel with the help of SSL

Configuration Management

System Administrators usually perform repetitive tasks such as installing servers, configuring those servers, etc. They can automate this task, by writing scripts, but it is a very hectic job when you are working on a large infrastructure.
To solve this problem, Configuration Management was introduced. Configuration Management is the practice of handling changes systematically so that a system maintains its integrity over time. Configuration Management (CM) ensures that the current design and build state of the system is known, good & trusted; and doesn’t rely on the tacit knowledge of the development team. It allows access to an accurate historical record of system state for project management and audit purposes. Configuration Management overcame the following challenges:
  • Figuring out which components to change when requirements change.
  • Redoing an implementation because the requirements have changed since the last implementation.
  • Reverting to a previous version of the component if you have replaced with a new but flawed version.
  • Replacing the wrong component because you couldn’t accurately determine which component needed replacing                                                         
 WHY WE NEED CONFIGURATION MANAGEMENT TOOLS?
  • As a system administrator our job is a tedious one with repetitive tasks. As infrastructure grows these tedious tasks will take a big chunk of our time. This is where automation comes in to picture. Many people started using scripting languages like Shell scripting, Perl scripting, Python scripting, Ruby, PHP or even new language Go to automate repeated tasks. But our infrastructure growth did not stopped to just couple of 100 machines and growing exponentially with the advent of Cloud. This is where we require centralized configuration management tools like Puppet and Ansible. These tools allows us to configure any server or even multiple servers remotly and with ease. This is a kind of big brother to fabric module in Python. Below are some advantages.
  • Saves the time,
  • Repetitive task removal at a large scale.
  • Can work in homogeneous environments.
  • With puppet like tools we can build entire infrastructure without manual intervention in hours time which normally takes weeks or even months to set-up with huge financial inputs.

  • WHAT ARE DISADVANTAGES OF SCRIPTING?
    • A scripting language can be used to automate simple to medium tasks but if you want to build complete infrastructure we have to take help from configuration management tools which can talk to different parts of an infrastructure. As these tools implemented different modules to talk to different parts of IT infra structure.

  • WHY WE REQUIRE PUPPET?
    • Though cfengine is the first configuration tool it did not capture market. Puppet did that and most companies uses it now. But from my experience puppet is good for devops and Ansible is good for system administrators.

  • WHO DEVELOPED PUPPET?
    • Luke Kanies is a system admin who thought scripting is limiting factor to automate system admin tasks and developed puppet as an open source project.

  • WHOM PUPPET IS MEANT FOR?
    • Puppet is meant for all the people who like to automate infrastructure tasks(devops) or orchestration tasks(sysops).

  • WHERE WE CAN/CAN NOT USE PUPPET?
    • Ok, is puppet/Ansible only solution for system administration? The answer is no. There are still some tasks which we have to depend on manual system administration or scripting.



Install Ansible


This blog will guide you to install Ansible in your CentOS machine in two easy steps. 
You will also learn how to setup your node machines and I will show you an example of how to install Nginx in your node machine from your control machine. 

Install Ansible on CentOS 6.8

Step 1: Set EPEL Repository
EPEL (Extra Packages for Enterprise Linux) is open source and free community based repository project from Fedora team which provides high quality add-on software packages for Linux distribution including RHEL (Red Hat Enterprise Linux), CentOS, and Scientific Linux.
Ansible package is not available in the default yum repositories, so we will enable EPEL repository for CentOS 6.8 using the command below:
#sudo rpm -ivh http://dl.fedoraproject.org/pub/epel/6/i386/epel-release-6-8.noarch.rpm

This will download all the necessary packages which will be required to install Ansible.
Step 2: Install Ansible
Now that your EPEL repository has been added, all you have to do now is install Ansible using the command below:
#yum install ansible -y
Installation completed, simple right
Now, if you want to check out your version of Ansible which you have installed, you can use the command below:
#ansible –version

Post Installation

After installing Ansible you need to add servers which you want to manage through Ansible. For that I have created another CentOS VM which will act as my node machine.
The first task is to setup password-less SSH authentication with nodes on my Control machine.
Step 1: Generate SSH key on the Ansible Control Machine. In order to do that, use the command below:
#ssh-keygen
As you can see in the above snapshot a public SSH key has been generated.
Step 2: Now, check the IP addresses of your nodes because you will need to specify it in the Ansible inventory later. To do that, type in the command ifconfig on your node terminal. 
Step 3: After the key is generated, the next task is to copy public key of Ansible server to its nodes. Use the command below:
#ssh-copy-id -i root@
Step 4: Now, you can use any editor to write your inventory or specify the IP addresses of your nodes grouped under ‘test-servers’ (or whatever you want your group name to be). I am using the vi editor.
use the following command:
#vi /etc/ansible/hosts
After you have added the IP address under ‘test-servers’, save the file and then exit.
Step 5: You might want to check if the IP address of your host has been added or not. To sample output of the host files using the following command:
#cat /etc/ansible/hosts
You can see the IP address of my host being displayed in the snapshot above. 
Step 6: Let us now perform a simple ping operation to test the connectivity using Ansible. To do that, just type the following command:
#ansible -m ping ‘test-servers’

Now, that you have checked the connection with your host machines, you are ready to manage them with Ansible.
Let me show you few shell command examples using Ansible.
  • Check uptime of your node machines
Uptime is a measure of the availability and working time of a computer. In order to check that use the following command:
#ansible -m command -a “uptime” ‘test-servers’
  • Check kernel version of your nodes
It can be useful to know the version number of the kernel (i.e., the core of the operating system). Use the command below:
#ansible -m command -a “uname -r” ‘test-servers’
Now, let us install Nginx into my node machine from my control machine using Ansible.

Deploy Nginx Using Ansible

Nginx is software to provide a web server. It can act as a reverse proxy server for TCP, UDP, HTTP, HTTPS, SMTP, POP3, and IMAP protocols, as well as a load balancer and an HTTP cache.
I am deploying Nginx using Ansible in just one node. You can deploy it in multiple nodes too using the same way. All you have to do is list out the IP addresses of the nodes under ‘test-servers’.
Use the following command in your control machine:
#ansible test-servers -m yum -a “name=nginx state=installed”

Now to check if it is installed in your node machine, type the following command in your node:
#ps waux | grep nginx


In this article, I will show you how to install ‘Ansible’ on RHEL/CentOS 7/6. Like other configuration management tools like puppet, chef and CFEngine, server software is installed on one machine and client machines are managed through the agent. Wherein Ansible, the nodes are managed by controlling machine (Ansible server) over SSH, so there won’t be any agent running on node machines.

Base Requirements    :  

  1. Controlling Machine , 
  2. Client Nodes , 
  3. Required Packages

Software Prerequisites      :  

  1. Operating System: RHEL/CentOS/Fedora and Ubuntu/Debian/Linux Mint
  2. Jinja2: A modern, fast and easy to use stand-alone template engine for Python.
  3. PyYAML: A YAML parser and emitter for the Python programming language.
  4. parmiko: A native Python SSHv2 channel library.
  5. httplib2: A comprehensive HTTP client library.
  6. sshpass: A non-interactive ssh password authentication.
Note: Use yum for installation because it will install all dependencies automatically.

My Controlling Machine and Nodes   : 

  • 192.168.87.145    server.gekvplabs.com    server
  • 192.168.87.146    node1.gekvplabs.com    node1    
  • 192.168.87.147    node2.gekvplabs.com    node2

Installation of  Ansible on On RHEL/CentOS :

  1.  Run yum update  for updating all existing packages
    1. yum update -y
  2.  Install EPEL repo 
    1. yum install epel-release
  3.  After configuring epel repository, you can install Ansible using following command
    1. sudo yum install ansible -y
  4. After installed successfully, you can verify the version by executing below command.
    1. ansible --version 


Ansible is a free configuration management tool, it supports managing the configurations of Unix-like and Microsoft windows systems. Ansible manages nodes over SSH or PowerShell and python to be installed on them.

Ansible is agent-less, that means no need of any agent installation on remote nodes, so it means there are no any background daemons or programs are executing for Ansible, when it’s not managing any nodes.

Ansible can handle 100’s of nodes from a single system over SSH connection and the entire operation can be handled and executed by one single command ‘ansible’. But, in some cases, where you required to execute multiple commands for a deployment, here we can build playbooks.