Dynamic Risk-Based Updates Using Python and Excel

Learn how to elevate your Ansible update strategy by creating a dynamic, risk-based inventory using Excel 📊 and Python 🐍. This article walks you through replacing static hosts files with a flexible, easy-to-maintain setup that prioritizes updates based on risk levels ⚠️-keeping your patching process efficient and adaptable 🚀.

In this blog, we'll take a simple Ansible server update script and turn it into a Risk-Based Update System. Here, servers with the lowest risk get patched first, giving us a chance to test thoroughly before moving on to higher-priority systems.

• Ansible Automation:
  • Dynamic Risk-Based Updates Using Python and Excel"
  • Host File
  • Dynamic Host List
  • Why Not Use a Hosts File?

The secret sauce? Setting up well-defined groups to make this flow seamlessly. But the real question is: can we pull this off without major changes to our Ansible script from last time? Let's find out!

Host File

The host file is at the heart of this change. In the last post, we used a static file grouped by server types. Now, we're adding a second layer of grouping by risk level-which does add some complexity to the host file.

But here's the twist: what if our host file could be dynamically generated from a more generic source? That would keep things flexible and save us from endless file editing!

Dynamic Host List

Ansible can work with dynamically created host files, which gives us a more flexible way to keep track of servers. In this example, we'll use an Excel file to organize our hosts.

Example hosts_data.xlsx Structure:

Host Name	Server Environment	Ansible User	Server Type	DNS	Notes
mint	dev	richard	desktop	desktop.sebostech.LOCAL	Mint desk top
ansible_node	dev	ansible_admin	Ansible	ansible_node.sebostech.local	Development server; Only updates monthly
clone_master	dev	ansible_admin	clone	clone.dev.sebostech.local	Development server; Only updates monthly
mele	staging	richard	nas	nas.stage.sebostech.local	Testing server; Used for application testing
pbs	production	root	backup server	pbs.prod.sebostech.local	Testing server; Used for application testing
pve	production	root	hypervisor	api.stage.sebostech.local	Testing server; Used for application testing
samba	production	richard	nas	nas.prod.sebostech.local	Critical server; Requires daily backup
firewall	production	richard	firewall	firewall.sebostech.local	Critical server; Requires daily backup

Most IT departments already have a list of servers stashed in an Excel file, so why not put it to good use? This approach makes it easy to keep our Ansible hosts organized and up-to-date without constant manual updates.

But how does Ansible use the Excel file? Let's dive into how we can transform this data into a usable dynamic inventory!

## This will run agains all host
ansible-playbook -i dynamic_inventory.py playbook.yml

You can also use environment variables option to target specific groups, based on Server Environment, Server Type, or even a combination of both:

## Just production
SERVER_ENVIRONMENT="production" 
ansible-playbook -i dynamic_inventory.py playbook.yml --limit "high:web"

## Just nas
SERVER_TYPE="nas" 
ansible-playbook -i dynamic_inventory.py playbook.yml --limit "high:web"

## production nas
SERVER_ENVIRONMENT="production" 
SERVER_TYPE="nas" 
ansible-playbook -i dynamic_inventory.py playbook.yml --limit "high:web"

Need new groups? Just update the Excel file and adjust the Python script accordingly-easy as that!

For a look at the Python code, see here.

Why Not Use a Hosts File?

When I first started using Ansible, the hosts file was my go-to. But as I added more servers, especially ones with dual roles, that file got more and more complex.

Could you use a traditional hosts file to achieve this? Sure-but there are a few drawbacks.

With a hosts file, you'd likely end up with duplicate entries or additional variables to capture all the structure you need. An Excel file, on the other hand, provides a clean, easy-to-maintain structure that keeps things organized.

In a corporate environment, there's a good chance there's already at least one Excel file with a server list, so why not take advantage of it?

If you'd like me to dive deeper into the Python code, just let me know!