Everything below happens in the panel. The same steps are available over the API; see the API guide for the requests.
1. Create a storage zone
A storage zone is a bucket. Give it a name, a region, and a quota if you want one — a quota of zero means no quota rather than no space. Tiers and replication regions can be changed later.
2. Upload an object
The object browser uploads in chunks through a resumable session, so a large file survives a dropped connection and a page reload. Folders are inferred from object paths; creating one writes a zero-length marker so an empty folder can exist.
3. Front it with a pull zone
A pull zone is the delivery side. Point it at the storage zone, set a cache TTL, and add a custom hostname if you have one. Until a pull zone exists, an object has no public URL and the panel says so rather than offering a link that would not work.
4. Check it
Copy the object's URL from its row menu and request it. The first request fills the edge cache; the second should be served from it. Delivery statistics record both.
Roles
New accounts start as Viewer, which is read-only. Creating and changing resources needs Operator; user administration needs Admin. The panel hides what your role cannot use, but the server is what enforces it.