Site Creation, Cloning, and Site Lifecycle Management
Creating a site
A site is created from the factory console or API by choosing an install profile/distribution, a site group, a domain, and an administrative account. The platform provisions a database, a files directory, and a stack, then runs the installer. During provisioning the site has a transient status (provisioning, active, failed). Failed provisioning should be diagnosed from task logs, not by manually editing the database.
Cloning
Clone creates a new sibling site in the same environment by copying the source site's database and files. The clone gets a new domain and its own stack; it does not share data with the source. Cloning is the standard way to create a template or a canary site. It is not a backup: it is a new site with its own lifecycle, and changes to the clone do not affect the source.
Lifecycle
Typical lifecycle: create/provision -> active -> (optionally) clone or back up -> delete. Deletion is usually a soft delete: the site enters a deleted/trash state and can be restored during a retention window; after that, garbage collection removes the stack permanently. Backups are separate artifacts and should be taken before risky operations. Environment refreshes (prod to test) are platform-level operations, not per-site clones.
Traps
Assuming clone is a backup; assuming deletion is immediate; assuming a clone keeps the source domain; expecting a deleted site to be recoverable forever.
Rule of thumb
Before any destructive action, take a backup, and verify the retention window and restore procedure. Treat the retention window as your only safety net.