How to Create and Manage Runners
There are two types of runners: self-hosted and GitHub-hosted. GitHub has standardized runners for you, but you can also create larger runners with more resources.
Ephemeral
GitHub runners are ephemeral meaning they are created on the fly and destroyed when the job is complete. This is the default behavior for GitHub-hosted runners.
Types and Sizes of Runners
| CPU | Memory (RAM) | Storage (SSD) | Architecture | Operating system (OS) |
|---|---|---|---|---|
| 6 | 14 GB | 14 GB | arm64 | macOS |
| 12 | 30 GB | 14 GB | x64 | macOS |
| 2 | 8 GB | 75 GB | x64, arm64 | Ubuntu |
| 4 | 16 GB | 150 GB | x64, arm64 | Ubuntu, Windows |
| 8 | 32 GB | 300 GB | x64, arm64 | Ubuntu, Windows |
| 16 | 64 GB | 600 GB | x64, arm64 | Ubuntu, Windows |
| 32 | 128 GB | 1200 GB | x64, arm64 | Ubuntu, Windows |
| 64 | 208 GB | 2040 GB | arm64 | Ubuntu, Windows |
| 64 | 256 GB | 2040 GB | x64 | Ubuntu, Windows |
The 4-vCPU Windows runner only works with the Windows 11 Desktop image.
Note: arm64 runners are currently in beta and subject to change.
GPU runners are also available.
| CPU | GPU | GPU card | Memory (RAM) | GPU memory (VRAM) | Storage (SSD) | Operating system (OS) |
|---|---|---|---|---|---|---|
| 4 | 1 | Tesla T4 | 28 GB | 16 GB | 176 GB | Ubuntu, Windows |
Auto-scaling and Scale Limits
If you hit scaling limits you can ask your AM or support to increase your concurrency limit.
ARC: Actions Runtime Configuration
You can automatically scale your self-hosted runners in response to webhook events.
Runner Groups
Runner groups simply allow you to manage which runners are available to which repositories. This is useful if you have a runner that is only available to a specific team.
GHR Networking
By default, GitHub-hosted runners have access to the public internet. However, you may also want these runners to access resources on your private network, such as a package registry, a secret manager, or other on-premise services.
API Gateways
You could run an API gateway on the edge of your private network that authenticates incoming requests with the OIDC token and then makes API requests on behalf of your workflow in your private network.
Wireguard
Static IPs
You have the option to create static IP addresses for your GitHub-hosted runners. This is useful if you need to whitelist the IP address of your runner in a firewall rule, for example.
VNET Injection (Azure Private Networking)
You can use VNET injection to connect your GitHub-hosted runners to your Azure virtual network.
Runner Labels
You label your runners to make it easier to target them in your workflows.
Example of using multiple runner labels
runs-on: [self-hosted, linux, x64, gpu]