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

# Server

> Single-instance compute (VM)—Create, configuration panel, and API subtype instance.

## Summary

|                 |                                                                                                                                                                                                   |
| --------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Purpose**     | One **virtual machine** (or cloud **instance**) in the linked account—general-purpose compute for apps, bastions, workers, or anything that maps to a single VM.                                  |
| **Create menu** | “Deploy a web service.”                                                                                                                                                                           |
| **API subtype** | `instance` (cards may show **Server** or **instance**).                                                                                                                                           |
| **Panel**       | **`ConfigurationPanel`** — same family of controls as other compute: **region**, **OS / image (AMI or provider image)**, **instance type**, **boot volume** (type and size), and related options. |

## Configuration (typical)

After you choose **Server**, the panel loads provider data for the project’s cloud:

* **Region** — Where the instance runs; changing region reloads images and instance types for that scope.
* **CPU architecture** — For example **x86\_64** vs **arm64** where the cloud offers a choice; filters compatible images and sizes.
* **Operating system / image** — Horizontal selector of OS families (for example Amazon Linux, Ubuntu, Debian) with a **version/build** choice per family; images come from the platform’s image APIs for the project.
* **Instance type** — Family and size (vCPU, memory); pricing hints may appear when the API provides them.
* **Storage** — Boot volume **type** (for example `gp3` on AWS) and **size** (GiB), with guardrails per OS image where applicable.

Exact fields depend on **AWS vs GCP vs Azure** (the panel branches on `cloudProvider`).

## Canvas and connections

* Connect **servers** to **firewalls**, **runtimes** (install runtime on this host), **applications**, or other allowed pairs per your **resource connection** rules.
* **Terminal** resources may attach to eligible servers for shell access.

## Operations

* **Sync** from cloud (where implemented) refreshes status and metadata from the provider.
* **Delete** tears down the instance and related cloud objects per the unified resources API.

**API:** See [API documentation](/api/index) → **Servers & instances**, **Unified resources**, **Resource connections**.

[← Resource types hub](/gateways/resources)