hands-on-lab

Goal

The goal of this lab is to set up a webserver.

The webserver is to fetch a "package" from the ctxs-lab-content S3 bucket. The webserver needs to be publicly accessible for HTTP and RDP. To manage the webserver we will create an Ansible master. The ansible master needs be available for public SSH.

Setting up a core Ansible server:

With this exercise, we'll be using Terraform to automatically spin up a pre-configured Ansible AMI to expedite to help us hit the ground running.

Running your first playbook:

So, let's run a playbook to configure the Ansible server. This is a very simple playbook meant to fetch files called modules from GitHub and place them in the library folder.

First connect to the Ansible server via SSH

ssh -i <path to DevOps-lab.pem> ec2-user@<public IP for the Ansible server>

Now that you are connected to the Ansible server

• cd ansible\library

• rm -f *

• cd ~/ansible

• ansible-playbook localhost.yml

... and voila, you just ran your first playbook

checking the results:

To visualize the changes resulting from running the playbook, while logged into the Ansible server run:

• cd ~/ansible/library

• ls

Provisioning a Windows instance to run a static website

This goal of this excercise is to provision an instance from a template, install IIS and fetch a package from S3 which contains the website assets. Playbooks run as part of this lab are very similar to the steps used to deploy the ShareFile webapp and api servers running in the AWS cloud.

Step 1: Configuring the Ansible server to work in your VPC

• cd ~/ansible/roles/common/vars

Use your favorite text editor to edit defaults.yml, provide the values for your VPC to the variables subnet_id, windows_security_groups and region

Step 2: Deploying a simple static website

• cd ~/ansible

• ansible-playbook sample.yml -e "tag=<tag to identify your webserver in the AWS console>"

• wait until the playbook completes, it will take a little while...

If it all worked as expected, from your workstation, open your favorite web browser and enter the provisioned web server public IP to browse the automatically deployed web site.

Step 3: Re-deploy

To demonstrate the power of Ansible, let's put the machine in an 'unknown state' by deleting the folder where the website resides.

• First RDP into the web server and delete the webapp folder under "c:\inetpub\wwwroot"

• Next run the playbook again by running the command ansible-playbook sample.yml -e "tag=<tag to identify your webserver in the AWS console>" from the Ansible server prompt

Notice how the playbook runs much faster as Ansible determines what elements are the desired and therefore do not need to be touched.

Step 4: Debugging

On the Ansible server type the command

• export ANSIBLE_KEEP_REMOTE_FILES=1

• run the playbook again by running the command ansible-playbook sample.yml -e "tag=<tag to identify your webserver in the AWS console>"

• RDP into the webserver and from the PowerShell Admin prompt nagivate to c:\Users\ansible\AppData\Local\Temp

• Open any of the ansible-tmp-* files in Notepad to get a glimpse of the PowerShell scripts that have run during the deployment

AMI Map

Use these AMIs. (Terraform has them pre-loaded.)

                US-EAST-1       US-WEST-1      US-WEST-2
Ansible AMI     ami-923cbc85    ami-2589cf45   ami-8f68abef
Webhost AMI     ami-9ce8688b    ami-4197d121   ami-e067a480