How often are you using Ansible in your daily work? Do you mainly rely on Ansible for configuration management and server deployment, Terraform for provisioning infrastructure, or a combination of both? I'm also curious whether AI assistants or agents are changing how you build and run that automation.
5 Answers
I manage a small game company with a few dozen virtual machines, build systems, and automated QA machines, and nearly all of it is managed through Ansible. AI has made the process easier because I can ask it to help create or modify a playbook, review the changes, and then run the automation myself instead of letting an agent make uncontrolled changes directly.
The two tools usually solve different problems, so using both is common. Terraform defines and provisions the desired infrastructure, while Ansible—or another configuration-management tool—configures systems and corrects drift inside that infrastructure. In some environments, tools like Puppet, Rudder, or Salt fill the configuration-management role instead. Ansible can still be useful for inventory, mass access, deployments, and server configuration even when it isn’t the primary management tool.
For us, Ansible runs daily for configuration management and several times a week for virtual-machine builds. We also use it for monthly Linux patching and smaller tasks such as managing relay allowlists. AI is limited to autocomplete or reviewing YAML. Even when using AI, the changes should go through infrastructure-as-code and normal testing, because direct agent changes can leave systems in an unpredictable state.
Ansible is especially valuable for repetitive operational work. One organization used it to notify customers, disable monitoring, patch Linux and Unix servers, and restore monitoring afterward. They could patch around 30 machines with one action and finish most of the work in a few minutes of a maintenance window. The initial automation effort was substantial, but the long-term operational savings were significant.
I use Ansible constantly, especially because it gives me a clear, repeatable record of what was done and why. I often reproduce customer issues, so being able to document the exact steps in a playbook and keep them in version control is extremely useful. It also becomes living documentation for the team.
That makes sense. Is it mainly that your everyday tasks don’t need Ansible, or could you technically get by without it but prefer the repeatability?
We use it the same way. Our environment is small and varied, so the time savings aren’t always huge, but readable playbooks and Git history mean we don’t have to maintain separate documentation. AI can help draft tasks, but only when the prompts are specific and the results are reviewed.
The main use cases I had in mind were inventory management and deploying or configuring multiple servers.