Skip to content
CodeFloe

Hardware

CodeFloe provides the following servers to process CI/CD jobs:

NameTypeArchitectureCPUMemoryJob CapacityCPU/JobMem/Job
artemisHetzner AX42 (dedicated)amd641664 GB847 GiB
ci-1devxy PVE cluster (zeus)amd64822 GB947 GiB
ci-2devxy PVE cluster (hera)amd64816 GB627 GiB
ci-3devxy PVE cluster (poseidon)amd641632 GB1347 GiB
ci-arm1Hetzner CAX41 (cloud)arm641632 GB647 GiB
gaiaApple Mac Mini M4arm641032 GB547 GiB

The ci-1, ci-2 and ci-3 runners are virtual machines on a self-managed Proxmox VE cluster, which itself runs on rented Hetzner bare metal. They sit on desktop-class CPUs (an AMD Ryzen 9 7950X3D for ci-1, an Intel Core i9-13900 for ci-2 and ci-3), which buys a noticeably higher single-core clock than the cloud instances. Each VM carries a share of its host’s rental, listed on the infrastructure hardware page. Because they have no public address, they reach the Crow server and Forgejo through a Headscale tailnet.

The total job capacity of all runners is split 50/50 between Crow CI and Forgejo Actions:

NameCrow CIForgejo Actions
artemis53
ci-154
ci-233
ci-358
ci-arm133
Sum2121

gaia is not part of this split and is scheduled independently. Its listed capacity and resource limits refer to its Crow Docker agent.

The hosted runners listed above configure a memory limit of 7 GiB and a CPU quota of 4 cores per container. The only CPU exception is ci-2, which limits containers to 2 cores because it has the smallest CPU allocation.

Limits apply to Crow step containers and Forgejo Actions job containers, rather than an entire pipeline. Parallel containers each have their own limits, while sharing the host’s available resources. These limits are configured in the IaC repository.