terraform-state-recovery

Recover Terraform state by identifying orphaned resources using `terraform state list` and `terraform state show`, then remove stale entries with `terraform state rm` before re-applying. This clears references to resources that no longer exist in the recreated cluster, unblocking deployments.

Recreating a Kubernetes cluster leaves stale state entries for resources that no longer exist. Terraform apply then fails with resource-already-exists errors because it tries to manage orphaned Kubernetes namespaces and Helm releases still referenced in state but absent from the cluster.

Use `terraform state list` to enumerate all Kubernetes and Helm resources in state, then verify each exists in the cluster using kubectl. Resources present in state but missing from the cluster are orphaned and should be removed with `terraform state rm` before re-applying.

Yes, a provided Bash script automates orphaned resource detection and state removal by comparing state entries against cluster resources, eliminating manual `terraform state rm` commands and reducing error risk during recovery.

Verify state is clean by running `terraform plan` after removal to confirm no active infrastructure is affected. Use remote state locking to prevent concurrent modifications. Remove only confirmed orphaned entries; incorrect removals can break active resource management.

Use Terraform data sources for read-only cluster discovery, let GitOps tools like ArgoCD manage application resources independently, and enable remote state locking to enforce safe concurrent state operations across team members.