Re-assessing ops: Ansible, Poetry, Terrakube, Gitbook and more

The benefit of dedicated technical study time on Saturdays is seeing the bigger picture. As I come to the end of my Site Reliability Engineer course on Udacity, I’m spending more and more time following the additional resources provided. Which leads to yet other sites and articles, and to a fundamental reassessments of what I had thought was “done” for ops building blocks. Thoughts from my research today:

    • Conda vs Poetry – for dependencies when developing in Python, while I use Conda fairly extensively could Poetry be a more elegant solution? Or as suggested here, should be used to complement each other?

    • Sphinx/Doxygen vs Gitbook – I understand and have seen deployments of Sphinx for automated code generation. Is Gitbook are more modern successor, or just different? I’m looking for something that is as close as possible to realtime code and its state with minimal management effort from engineering.

    • Terraform Cloud vs Spacelift vs Terrakube – what collaboration platforms are useful if a team scales and every engineer is fluent in the basics of IaC, k8s and public cloud deployment? Not hearing good things about Terraform Cloud pricing on Reddit, and Spacelift is focused on policy/governance which seems to be for more sophisticated/mature setups. Terrakube could be a good middle ground offering access policies, org mgmt and visualization. There’s also Atlantis but seems to be limited in scope.

    • Terraform vs Ansible – similar to the first one, my view was originally either or. But according to this article or this video it’s more of a team play. in my view would be Ansible as app config support to core Terraform infra mgmt workflow. I haven’t used it so far with Terraform. But as ops scale, could configuration management get increasingly tricky without Ansible?

So as you can see no aha moments or earth shattering breakthroughs. Rather an increased awareness of technology tradeoffs and discussion starters for deeper ops discussions. PS. the image above was AI-generated by Adobe Firefly. Didn’t work out too bad, even with my basic prompting skills. And if you really want to know, it was “hip AI engineer working on software overlooking Tokyo Bay” 🙂

Leave a Reply