AWS Copilot deprecated: FAQ¶

Is AWS Copilot CLI deprecated?¶

Yes. AWS ended support on 2026-06-12 and archived the repository on 2026-06-22. See end of support.

Will my Copilot services stop running?¶

No. They are CloudFormation stacks and keep running. Two dates matter for Copilot's custom-resource Lambdas, which run on nodejs20.x: creating them is blocked from 2027-07-29, and updating them from 2027-08-31 (AWS Lambda runtime deprecation table, checked 2026-09-30; AWS can change these dates).

Can I just delete the Copilot CloudFormation stacks after importing into Terraform?¶

Not safely without preparation. A normal stack delete removes every resource without DeletionPolicy: Retain, and Copilot's custom resources delete certificates, DNS records and bucket contents. Deleting the last service stack also makes the env-controller remove the shared load balancer, NAT gateways and EFS. ecsodus retain-patches every stack first. On AWS this stopped every Delete handler from running (evidence).

Why doesn't `delete-stack --retain-resources` work?¶

It only applies to stacks already in DELETE_FAILED. On a healthy stack the first delete removes everything.

Does ecsodus change anything in my AWS account?¶

No. ecsodus only reads: its AWS clients refuse every call that isn't Describe/List/Get/Lookup. You run the generated runbook yourself, and each mutating step in it is gated by an ecsodus check.

Will the migration cause downtime?¶

Adopt-in-place moves no traffic. The same ECS service, load balancer and VPC keep serving, and only the owner changes from CloudFormation to Terraform. In the AWS end-to-end run the service answered HTTP 200 throughout.

How is this different from cf2tf or `terraform plan -generate-config-out`?¶

Those convert templates or generate starting-point configuration. ecsodus produces HCL with the exact deployed values, so the first plan is a pure import. It also handles the parts that make Copilot dangerous to tear down: retain patches, custom-resource handlers, nested addon stacks, the app StackSet, and teardown order.

Should I use ECS Express Mode instead?¶

Express Mode is AWS's recommended destination for new deployments, and it can be managed by Terraform. It is a rebuild, though: traffic moves, and you migrate stateful resources yourself. ecsodus keeps what you already run, and its report tells you whether each service would fit Express Mode later.

No. This is the AWS Copilot CLI, Amazon's tool for deploying containers to ECS and App Runner.

What about AWS App Runner and Proton?¶

App Runner closed to new customers on 2026-04-30. Support for it is planned for v0.2. Proton reaches end of support on 2026-10-07.