Terraform vs Pulumi for managing cloud infra in Nairobi

I've been on Pulumi with Python for about six months, after years on Terraform. Drift detection is decent. Pulumi has a `pulumi refresh` like Terraform's, and catching drift during previews feels more transparent since you can write custom logic to handle it in Python. The learning curve for a Terraform team is mostly thinking in code instead of config. If they know basic Python, give them a week or two. On pricing, the free tier covers single users, but a small team will likely hit the $50/user/month Professional tier once you need state locking and a shared backend.

Hey I made the switch a few weeks ago and Pulumi's preview shows drift clearly because you can write the checks in Python, so it feels easier to debug than Terraform's static config. The learning curve is just picking up basic Python if you already know Terraform and the free tier works for a small team but you may need the $50/user professional plan if you want state locking.

Good point from the last reply about self-hosting the backend on S3, that's key for keeping costs down. Another thing: since you're in Nairobi, maybe trial Pulumi with a small non-critical project first, like a dev environment, to see how it handles slower internet or occasional dropouts when pushing state. Saves you committing fully before you know how it feels day-to-day.

Hey, thanks for laying out the Nairobi team context so clearly. One thing worth checking early: Pulumi's state backend can live in an S3 bucket, same as Terraform. So if cost is the main blocker, you can self-host the backend and stay on the free tier longer than most teams assume. Happy to point you at the exact docs if your colleague is still on the managed service.

Solid thread already, and I like the S3 backend point. For a small team, that really changes the math. One thing I'd add: since you're cost-sensitive, also think about the time cost of debugging. Terraform loops and conditionals are doable but clunky, and if your team already knows Python, Pulumi might save you hours on those tricky bits. Nairobi internet aside, the day-to-day pain is the same everywhere. Start with that dev project, test how state uploads feel on a slow day, and see if the Python way clicks for you. If it does, the $50 per user is easier to justify.

Solid advice already. One thing I'd add from a Nairobi angle: we've had days where the connection to AWS drops for a bit, and Pulumi's Python code can retry or fail more gracefully than HCL if you build that in. Try it on a small dev project first, like the last person said. That way you learn the feel without risking your main setup or paying anything extra.