I moved a small team project from Terraform to OpenTofu last quarter. The migration itself was trivial. That's the trap.
The problems started when we needed to integrate with our existing pipelines. Our CI/CD system had a Terraform-specific plugin with enterprise support. OpenTofu broke it. We spent a week building and maintaining a custom container image. That's week-one hidden cost.
Then there's the provider ecosystem. Yes, it's compatible today. But what happens when a major cloud provider prioritizes Terraform's new features? We're already seeing lag in niche provider updates. You're betting your team's velocity on a fork keeping perfect parity. That's a risk.
OpenTofu saves you the license headache but introduces operational fragility. For a solo dev, that's a fine trade. For a team with SLAs and compliance audits, you're swapping one vendor's lock-in for another kind of lock-in—to a community project with uncertain long-term backing. The TCO math doesn't work.
Oh wow, that's a really good point about the CI/CD plugins. I hadn't thought about that at all. I've only tried OpenTofu locally.
When you say "uncertain long-term backing," are you mainly worried about the pace of updates, or something else? Like, the project losing contributors? I'm trying to understand the risk better.
The contributor concern is real, but it's secondary. The primary financial risk is corporate backing. OpenTofu's stated goal is to remain a drop-in replacement, which means it inherently cannot pursue a divergent, innovative roadmap that would attract major engineering investment from large companies. It's positioned as a defensive fork.
The money for sustained, high-quality development in infrastructure tooling comes from vendors who need deep integrations. If AWS, Google, or Azure decide to build a feature that requires changes to the core, where does that development happen? It happens in Terraform. Over time, that's how true incompatibility emerges, not from a malicious split, but from a natural divergence in priorities. You'll be waiting for the OpenTofu community to reverse-engineer and reimplement.
You're not betting on the current contributors leaving. You're betting that a project with a derivative technical vision can out-pace or even match the resource allocation of the de facto standard, indefinitely. That's a poor bet for team infrastructure.