Instance States and Lifecycle
Each instance managed by Thinfinity® Cloud Manager has two state values that describe it from independent angles: an Operational Status (what the VM is doing right now) and a Record State (whether the instance is enabled, errored, or in an internal transition).
This article describes both layers, the transitions between operational states, and how to interpret each value when troubleshooting.
For day-to-day operations on instances, see Manage Deployed Instances.
This article covers VM-based instances. For Kubernetes models, pod lifecycle is governed by the Kubernetes cluster — the Power Cycle, Enable / Disable, and Detach transitions described below behave differently or do not apply.
Operational Status
The Operational Status is the high-level state of the underlying VM. It drives the colored status icon shown in the leftmost column of the Instances table.
| Status | Meaning | Status icon |
|---|---|---|
| Active | The VM is running and accepting connections. | Green circle (On) |
| Inactive | The VM is powered off but still provisioned and ready to start. | Red circle (Off) |
| Provisioning | A Terraform Apply is in progress. The VM is being created. | Yellow circle (transient) |
| Activating | The VM is in the process of starting up. | Yellow circle (transient) |
| Inactivating | The VM is in the process of shutting down. | Yellow circle (transient) |
| Deleting | A Terraform Destroy is in progress. The VM and its record are being removed. | Gray circle |
| Creating Image | An image capture is in progress against this instance — the Create Image action was invoked (POST /images), and the instance is being captured into a Bucket (the Cloud Manager container for custom images). The instance is unavailable for user sessions while the capture runs. |
Yellow circle (transient) |
When a Detach completes successfully, the instance disappears from the table. There is no “Destroyed” terminal status that lingers.
Record State
The Record State is the availability layer for the instance. It tracks whether the instance is currently enabled for user assignment, has been disabled, or has hit an error. It is independent of the Operational Status — a VM can be Inactive but Enabled, or Active but Disabled.
| Record State | Meaning |
|---|---|
| Reset | Internal transitional state used while Cloud Manager synchronizes after a configuration change. Not normally visible for long. |
| Enabled | The instance is available for user assignment. This is the default for newly provisioned and healthy operational instances. |
| Disabled | The instance has been administratively disabled and is hidden from user assignment, even if its VM is still running. See Manage Deployed Instances for the Disable action. |
| Error | A lifecycle operation has failed. The instance requires administrator attention. The status icon is replaced by an alert overlay. |
| Deleting | A Terraform Destroy is in progress on the instance’s deployment. The instance record is being removed. |
| Deleted | Internal terminal state used briefly between the Detach completion and the record disappearing from the Instances table. Not normally visible to administrators. |
Status Icon Reference
The leftmost column of the Instances table renders a status icon that combines the Operational Status with overlays from the Record State.
| Visual | Meaning |
|---|---|
| Green circle | Operational Status is Active |
| Red circle | Operational Status is Inactive |
| Yellow circle (transient) | Operational Status is Provisioning, Activating, Inactivating, or Creating Image |
| Gray circle | Operational Status is Deleting |
| Forbidden overlay on the colored circle | Record State is Disabled (overlays the operational status) |
| Alert icon replacing the colored circle | Record State is Error (replaces the operational status) |
Click any status icon to open the View Log dialog for that instance.
Lifecycle Transitions
Provisioning
Triggered by: a Deploy action (manual via the Deployments tab, or automated by the model’s deployment schedule).
The instance record is created in Cloud Manager. A Terraform Apply runs against the cloud provider:
- Operational Status begins at Provisioning.
- On success, Operational Status moves to Inactive or Active depending on the model’s Power and Scheduling Settings. Record State becomes Enabled.
- On failure, Record State moves to Error.
See Provisioning Failures for failure-mode diagnosis.
Power Cycle
Triggered by: a Start or Stop action, an idle timeout, or a scheduled power event.
The cloud resource remains provisioned. Only the VM’s run state changes:
- Start: Inactive → Activating → Active.
- Stop: Active → Inactivating → Inactive.
Record State is unchanged by power transitions.
Enable / Disable
Triggered by: an Enable or Disable action (per row, or via the bulk header kebab).
This toggles only the Record State (between Enabled and Disabled) — the VM’s power state is unaffected. A Disabled instance is hidden from user assignment but its VM continues running unless also Stopped.
Detach
Triggered by: a manual Detach action on an instance whose Record State is Error (per row, or via the bulk header kebab). On an instance in any other state the action fails and nothing changes.
No Terraform Destroy is executed. The Cloud Manager record is marked deleted and the instance disappears from the table. Anything still present in the cloud provider is untouched.
To destroy the VM and its cloud resources, use Destroy on the corresponding deployment — see Deploy an Instance.
Detach is irreversible. See the warning in Manage Deployed Instances.
Error
A Record State transition to Error can happen at any point when:
- A Terraform Apply or Destroy operation fails.
- A power operation times out without confirmation from the provider.
- The provider returns an unexpected response that Cloud Manager cannot reconcile.
In the Error state, scheduled power events are paused for the affected instance until the error is resolved.
Diagnosing the Error State
When an instance is in the Error state:
In the Instances table, locate the instance — its status icon will be the red alert.
Click the alert icon (or the View Log action in the row’s Actions column) to open the lifecycle log for the instance.
Identify the failed operation and the error message returned by the cloud provider or by Terraform.
Cross-reference with the relevant troubleshooting page:
- Provisioning errors: Provisioning Failures
- Connectivity errors: Connectivity Issues
- Terraform-specific errors: Terraform Errors
Resolve the underlying issue (credentials, quotas, network, template). The next operation that reaches the cloud provider — for example, a Start — will exercise the corrected configuration.
If the instance cannot be recovered, detach it and provision a fresh one from the model. Detaching only removes the Cloud Manager record, so check the provider console for any resource the failed deployment left behind.
See Also