Proposed runs for infrastructure destruction

Terraform and orchestration utilities like Atlantis allow you to preview a proposed run (plan) that will destroy infrastructure via the “-destroy” command line flag.

“tf plan -destroy” or “atlantis plan — -destroy” are very common workflows on our team to ensure we don’t leave orphaned Terraform resources as we manipulate our stacks.

It seems the only Spacelift supported workflow for destruction of infrastructure is through stack deletion. This is problematic for two reasons:

• The UI asymmetry between proposed runs and infrastructural deletion. There’s no way to confirm exactly what resources will be destroyed prior to hitting the “delete stack” button.

• Inability to monitor resource deletion and retry/adjust as needed. It’s important to know what’s happening as it’s happening. For example, if a problematic resource gets hung during deletion manual action may need to be taken during or after deletion, and deletion may need to be re-executed several times.