> For the complete documentation index, see [llms.txt](https://gata-1.gitbook.io/gata-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://gata-1.gitbook.io/gata-docs/use-guide/gata-decentralized-ai-cloud/instance-overview-and-management.md).
# Instance Overview and Management
On `My Deployments` page, you can manage all your instances.
#### 1. Instance overview
The instances are listed by status — **Running, Snapshotting, or Released** — along with their ID, Name, Cost, and available Actions.
* **Running** – The instance is active and accessible, billed per usage time.
* **Snapshotting** – The instance is creating a snapshot and is temporarily unavailable. No additional fees are charged during this process.
* **Released** – The instance has been terminated and its resources reclaimed. It cannot be restarted, and no further charges apply.
#### 2. Instance Details
Click the `arrow icon` on the right to expand the instance details and access management options.
#### Access
**SSH (Port 22)**
SSH connection details:
* **SSH Command**: `ssh -p root@`
* **Service Port**: Port assigned to this instance
* **User**: `root` (default)
* **Password**: Initial password (masked)
#### HTTP (Port 8888)
Runs **JupyterLab** by default. Click **Open** to access the instance directly.
* **Link** – View the HTTP address
* **Open** – Launch the link in a new browser tab
* **Copy** – Copy the URL
* **Edit** – Modify port settings *(if available)*
* **Delete** – Remove the configuration
**New Port Forwarding**
Click **New Port Forwarding** to add a rule:
* **Service Port** – Enter the port number
* **Protocol** – Select the protocol (default **TCP**). For web apps (e.g., JupyterLab, Dashboards), enable **HTTPS** for secure browser access.
#### Billing
All billing information and cost details associated with the instance.
#### Config
Detailed configuration parameters of the instance.
#### Hardware
Detailed hardware specifications of the instance, including GPU and CPU models, GPU count, GPU memory, system RAM, and storage capacity.
#### 3. Instance Management (Running State)
#### Take a snapshot of the current instance
You can create a snapshot of the current instance to preserve its full system state and file changes (not those in Data Hub). The snapshot includes installed packages, configurations, and all directory-level modifications. You can later use this snapshot as a **template** to quickly launch new instances without repeating the setup process.
❕During snapshot creation, the instance is temporarily inaccessible, so it’s recommended to create a snapshot before releasing it.
Click `Take Snapshot` and enter the required details
* **Name** – Specify a name for the snapshot.
* **Auto-release after saving** – If enabled, the instance will be automatically released once the snapshot is complete; otherwise, it will return to the *Running* state.
#### Release the current instance
Releasing an instance frees its resources and changes its status to `Released`. Use this option when the instance is no longer needed, as it stops billing and reclaims allocated resources.
Click `Release`.
Click `Stop & Release` to confirm your action.
The status transitions from **Releasing** to **Released**. Instances in the Released state cannot be restarted. To create a new instance from a saved environment, select a **Snapshot** during instance creation.
