Labelling instances

Attributing an instance to a project tells Carolina Cloud to count that instance’s compute cost toward the project’s totals. Any org member can label their own instances — you don’t need to be an admin.

Assigning an instance to a project

Every instance card has a project chip showing its current attribution — either the project name or No project.

1. Click the project chip on the instance.
2. In the picker, choose a project from your org’s list.
3. The chip updates immediately and the instance’s compute starts counting toward that project.

To remove attribution, open the picker and choose No project. The instance keeps running exactly as before — only its billing label changes.

Creating a project on the fly

If the project you want doesn’t exist yet, admins can create it without leaving the instance. In the project picker, type a name in the new project row and confirm — Carolina Cloud creates the project and assigns the instance to it in one step. (Names follow the same kebab-case rules.)

Re-assigning never rewrites history

This is the important part. When you move an instance from one project to another — or clear its project — Carolina Cloud does not retroactively re-label past spend.

Under the hood, the moment you re-assign:

1. The elapsed billing interval since the instance’s last bill is closed out and charged to the old project at the instance’s current size.
2. The billing clock resets.
3. The instance switches to the new project, and only usage from that point forward counts toward it.

So if a VM ran for three days under study-a and you move it to study-b, those three days stay attributed to study-a forever. study-b only ever sees what the VM costs after the switch. Your per-project history is always a faithful record of what actually ran where.

Note

Re-assigning a project is a billing-only operation. It never restarts, resizes, or interrupts the instance — the workload keeps running uninterrupted.

What gets attributed

Labelling an instance attributes its compute cost — vCPUs, memory, GPUs, and disk, billed hourly. Storage is attributed separately, through project storage buckets, not through the instances that happen to use them.

A project’s combined live and cumulative figures are on the Projects page.