Docker is a powerful tool. Containers are extremely useful.
But especially today, when resources like RAM and SSD storage are becoming expensive again, consciously avoiding containers — with proper planning — can benefit both your budget and the environment.
Less overhead.
Less complexity.
Fewer layers that need to be maintained and patched.
For many classic workloads like:
- web servers
- mail servers
- simple APIs
- internal tools
bare metal or a traditional VM setup is often more than sufficient — and more efficient.
Docker is a tool.
And like with any tool: not every problem is a nail. 🔨
As an Ansible lover, I often find that services are actually faster to set up without containers.
Whether I use Docker or not is, in many cases, just a simple variable in my playbooks.
That said: for quick testing, Docker is perfect.
And when you need something running fast and reliably, containers are often the best choice. No question about that.
For me, it’s not about Docker vs. no Docker.
It’s about choosing the right level of abstraction for the actual problem you’re solving.
Where do you consciously use containers — and where do you deliberately avoid them?
