Cloning creates a new application from an existing one. Use it to stand up a staging copy of production, try a risky change against real data, or move an application to another resource.

Cloning from the dashboard

Open the application and choose Clone from its actions. Give the copy a label and a codename, pick the project and where it runs, and tick what to bring along under Settings, Addons and Cronjobs.

What can be copied

The copy always gets the source’s application settings, apart from health checks. Everything else is opt-in:

Copying data

An add-on can be cloned empty or with its data, for example a PostgreSQL database with its contents. Data is not copied in these cases:
  • MySQL and Valkey add-ons.
  • Shared add-ons. Shared storage can also only be cloned to the same resource and project.
  • A copy that runs in a different region from the source.

Where the copy runs

The copy runs on a resource you choose. When the resource is assigned to projects, the copy’s project has to be one of them. It can also run on Usage Billing instead of a resource. It then keeps the source’s CPU and RAM unless you give it another size, and the source’s region unless you pick another one.

What the copy deploys

A copy does not inherit the source’s deployment source. Either connect one to it - a GitHub repository, Git push or a container registry - or follow the source’s image:
  • Use Parent Image runs the copy on the source application’s image, while still letting you deploy other code to it by hand.
  • Enable Auto-Sync also redeploys the copy every time the source’s image is updated, which suits a staging copy that should always run what production runs. Manual deployments are turned off while it is on.
Both are available only when the copy runs in the same region as the source.

Cloning over the API

To run the copy on Usage Billing, send "billing": "pay_per_use" instead of resource_id, with an optional region_code and size. The full list of fields is in the API reference.