Chef components

Ruby is a simple readable language.

Recipes will be having something called as Resources.

In chef,we called it as a resources

In Ansible,we called It as a module.

(for every tool,the name which you called is different).

My things: (Manual steps).

1)    Installing Tomcat

a)    Yum install tomcat

        In chef è  what resource can I use  to achieve this ?

       (we can use :  

       package

       yum

       (A resource means something which will do action for us).

       If you samething on the Ubuntu server:

b)      Apt-get install tomcat7

2)    Creating a directory    /var/log/tomcat

       Mkdir /var/logs/tomcat

3)    Starting tomcat services.

       Service tomcat enable

       Service tomcat start

      (these are manually executed)

      End steps ,which I need to do is the deployment on the redhat machine.

      Something,we need to make this unit services,reusable we use the resources:

For example:

Write a file, a=tomcat  b=/var/logs/tomcat

Yum install a

Mkdir b

Service a start

Note:

If you want to change to httpd in place of tomcat we can change to the a = httpd

For example:

You want to install nginx change a = nginx

Every tool has its own way of defining the variables.

In chef,it is popularly called it as attributes and databags.

If the resources are not present,what to do?

Recipes rely primarily on resources, resources described a desired state of an element in the infrastructure.

What is Chef?

"Chef  turns infrastructure into code.With Chef,you can automate how you build, deploy and manage your infrastructure.Your  infrastructure becomes as versionable,testable and repeatable as application code.

Chef relies on reusable definitions knows as recipes are instructions for configuring web servers, databases and load balancers.Together,recipes describe what your infrastructure consists of and how each  part of your infrastructure  should be deployed, configured and managed.

•           Packages to be installed
•           Files to be created
•           Directories to be created 
•           Services to be started 
•           Config files to be updated
•           Commands to be executed

Common Chef Terminology

Recipes:

   Fundamental configuration element within an organization.

Cookbooks:

   Defines a scenario and is the fundamental unit of configuration and policy distribution.

Convergence:

   Occurs when chef-client configures the system/node based off the information collected from chef-server.

Configuration Drift:

   Occurs when the node state does not reflect  the updated state if polices/configurations on the chef-server.

Resources:

    A  statement  of configuration policy within a recipe 

    Describes the desired state of an element in the infrastructure and steps needed to configure.

when we install python.

For operating system point of view,we see here:

Apt-get install python

Apt-get install python-pip

(that means the provider is apt-get )

when it comes to the awscli language:

pip install awscli

(pip is the provider).

Provider is the thing,which does the work.

If anyone asks to do the installation on ruby:

There is something called the rubygem

(who does this work,those are called as a provider).

In the first two statements the provider is apt-get

 And the last one the provider is the pip

Chef-repo is the chef repository,what it contains,it contains:it is the area where it contains chef scripts.

Why it is special:because it contains details about the chef-server.

When you create an account in  hosted chef,you will be creating an organsation.

(it is an isolation,you create across team).

For every organsation you will have the chef-repo.

Note:for two organsations,you don’t have a same chef-repo.

Learn how to write chef recipes?

Always dealing with the topmost item:

The topmost item is cookbook

                                          Recipes

                                              Resources

For that,we are using a tool called the knife.

The timer starts  the moment you install chef (because the chef node will  interact with the chef server for every 30 minutes,to apply the new policies).

Note:

First thing,ensure that,all of the servers are installing chef-client at sametime.

(so that,for example you install the chef-client at 7.00 am and the convergence will happen at the 7:30,so ensure that the chef-client installation on the chef nodes are installed at the same time).

Whenever you execute knife command,you have to be in chef-repo:

download the starter kit from the chef-server > starter kit >  chef-repo > this is the directory structure.

Whatever the cookbook that is given to you by default,it is the starter cookbook.

The problem here is,you did not install chefdk:

you have to install chefdk in your local workstation,here localworkstation is the VM.

For example:

Install chefdk on windows:

this is the thing,how to install chefdk on the windows machine.

before writing anything inside the recipes,we have to know about the chef resources

just type chef resources in the google

Note:

What we have to pass -------à is the attributes and values.

What has to be done ----à is the action.

Example: Install the tree on  my node:

Going to use the resource package (which is in the chef documentation).

Note:

When you upload to the chef server,we upload it as a cookbook.

When you add it to the run_list,we add it as a recipe.

When your adding to the run_list : your going to tell in this manner:

Recipe[cookbookname::recipename(it is a filename,with out .rb)]

Recipe[helloworld::default]

(we have to remember,we have to add the recipes in the run_list).

apt-get install  tree

type  'name'  do

   attribute

   action

end

package  'tree' do

   action :install

end

package 'tomcat' do

   action :install

end

package 'apache2'  do

   action  :install

end

service 'httpd'  do

   action  :start

end

Now your assigning recipe,to be a part of convergence:

recipe [cookbookname::recipename]

recipe [helloworld::default1]

knife node  run_list  add  <NODE>  "recipe[motd]"