> ## Documentation Index
> Fetch the complete documentation index at: https://docs.rafftechnologies.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Lifecycle states

> How a VM moves through provisioning, running, stopped, and terminal states

<sub>Updated May 8, 2026</sub>

Every VM has a `status` field that reflects where it is in its lifecycle. Most actions you take in the dashboard or via the API are state transitions.

## State diagram

```
Creation: initiating → provisioning → booting → active
Stop:     active     → finalizing   → passive
Start:    passive    → booting      → active
Reboot:   active     → booting      → active
Delete:   any        → terminated
Failure:  any        → failure      (non-billable)
```

## State reference

| Status         | Description                               | Billable |
| -------------- | ----------------------------------------- | -------- |
| `initiating`   | Queued, initial setup before provisioning | Yes      |
| `provisioning` | VM being created in the hypervisor        | Yes      |
| `booting`      | Starting up                               | Yes      |
| `active`       | Running and accessible                    | Yes      |
| `passive`      | Stopped, resources reserved               | Yes      |
| `finalizing`   | Shutting down                             | Yes      |
| `failure`      | Creation or operation failed              | No       |
| `unknown`      | Unhandled state                           | No       |

## Why `passive` is billable

A stopped VM still holds CPU, RAM, disk, and IP allocations on the hypervisor. Those reservations are billed at the same per-hour plan rate even when the OS isn't running. To stop billing, **delete** the VM (and its volumes, snapshots, and backups, if no longer needed). See [Billing model](/products/build/virtual-machines/concepts/billing-model) for how subscription credits work when you delete mid-term.

## `provisioning` during snapshots and backups

While a snapshot or scheduled/instant backup is running on a VM, the dashboard temporarily shows the VM in **`provisioning`** state — even though the VM was already up and serving traffic.

This is **normal and not a problem**:

* The VM continues running. Network traffic, SSH/RDP sessions, and applications keep working.
* The state flips back to `active` automatically when the capture completes — snapshots are typically 1–2 minutes, backups around 3 minutes on average; both scale with the size of the disk being captured.
* It's the same `provisioning` label used during creation; the dashboard reuses the state because the hypervisor is doing capture work in the background.

If you see `provisioning` and you didn't trigger anything, check the **Snapshots** or **Backups** tab on the VM detail page — there's almost certainly a scheduled job running. No action needed; just wait for it to finish.

| What you see                          | What it actually means                                                                   |
| ------------------------------------- | ---------------------------------------------------------------------------------------- |
| `provisioning` right after **Create** | Real provisioning — VM being built                                                       |
| `provisioning` on an existing VM      | Snapshot or backup in progress; VM is still serving                                      |
| `provisioning` stuck for 10+ minutes  | Possible issue — see [Troubleshooting](/products/build/virtual-machines/troubleshooting) |

## Related

<CardGroup cols={2}>
  <Card title="Manage power" icon="power-off" href="/products/build/virtual-machines/quickstart-guides/manage-power">
    Trigger state transitions.
  </Card>

  <Card title="Pricing" icon="circle-info" href="/products/build/virtual-machines/details/pricing">
    What each state costs.
  </Card>
</CardGroup>