> ## 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.

# Scalable server

> Auto scaling groups and scaled compute—API subtype scalable_server.

## Summary

|                 |                                                                                                                                                                           |
| --------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Purpose**     | **Horizontally scaled** compute: auto scaling groups (AWS), managed instance groups (GCP), VM scale sets (Azure), or equivalent—depending on the linked cloud.            |
| **Create menu** | “Create an Auto Scaling Group” (AWS-oriented wording; other providers use their native scaling resource).                                                                 |
| **API subtype** | `scalable_server`.                                                                                                                                                        |
| **Panel**       | **`ScalableServerConfigurationPanel`** — extends the same ideas as **Server** (region, OS/image, instance type, volume defaults) with **scaling** and **network** fields. |

## Configuration (typical)

* **Region**, **CPU architecture**, **OS / AMI**, **instance type**, **boot volume** — Same pattern as [Server](/gateways/resources/server), with provider-specific defaults (for example default volume types: GCP `pd-balanced`, Azure `Premium_LRS`, AWS `gp3`).
* **Capacity** — **`minSize`**, **`maxSize`**, **`desiredCapacity`** — bounds and current desired count for the group.
* **Instance network access (AWS)** — **`public`** (instances get public IPv4 and public subnets) vs **`private`** (no public IP; outbound often via NAT). Other clouds apply analogous private/public patterns where supported.

## Canvas and connections

* Often placed **behind** an HTTP/S front end when your deployment provides one, and protected by [firewalls](/gateways/resources/firewall).
* **Runtime** resources can connect to group members where the platform supports installing runtimes on scaled instances.

## Operations

* **Capacity changes** may use dedicated API paths (for example PATCH capacity) in addition to general resource updates.
* **Delete** removes the scaling group and underlying instances according to cloud behavior.

**API:** [API documentation](/api/index) → **Scalable servers**, **Resources**, **Resource connections**.

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