Hyperbolic Distributed Operating System

Install HyperdOS (single-node setup from scratch)

curl -o install.bash https://raw.githubusercontent.com/HyperbolicLabs/Hyper-dOS/refs/heads/main/install.bash && bash install.bash

To access your cluster for maintenance

1. install k9s: https://k9scli.io/topics/install/

2. run the following command to save the kubeconfig (which should be kept secret)

# NOTE: you can skip this mkdir command if your .kube directory already exists
mkdir ~/.kube

# write the microk8s kubeconfig secret to the expected location
microk8s config > ~/.kube/config

# access your cluster using the kubeconfig
k9s

Installation Walkthrough Video

Headless Install

If you would like to install in a scripted environment, set these variables prior to running the install script:

export HEADLESS=true
export TOKEN=<your-hyperbolic-token>
export ALLOCATE_ROOT_DISK_GB=<integer-gb-of-root-disk-to-allocate-to-microceph>

Notes

• If you would like to run the install script yourself rather than curling from github, you are welcome to download and edit the install.bash file before running it on your node.

• We do not officially support operating systems other than Linux. That being said, if you would like to join the Hyperbolic Supply Network from a Windows or MacOS device, you are welcome to give it a shot:

  • https://microk8s.io/docs/install-alternatives

• We officially support single-node microk8s+microceph clusters only, HOWEVER - a custom multi-node cluster should work smoothly if configured properly. See below for customized installation guidelines:

(experimental) add more nodes to your cluster

Note: please reach out before doing this so we can support your cluster smoothly. As of right now, only single-node clusters will be shown to renters automatically.

add microk8s node

https://microk8s.io/docs/clustering

1. (on the new node) sudo snap install microk8s --classic --channel=1.32/stable
2. (on the original node) microk8s add-node
3. (on the new node) microk8s join <output-from-original-node>

add microceph node

https://canonical-microceph.readthedocs-hosted.com/en/quincy-stable/tutorial/multi-node/

1. (on the new node) sudo snap install microceph
2. (on the new node) sudo microceph init

Customized installation (existing kubernetes cluster)

Please get in touch if you are planning to install hyperdos on an existing multi-node cluster, we can help you get set up smoothly.

Prerequisites

• ArgoCD installed: https://argo-cd.readthedocs.io/en/stable/operator-manual/installation/
• NVIDIA Operator installed: https://docs.nvidia.com/datacenter/cloud-native/gpu-operator/latest/getting-started.html
• Namespaces hyperdos and instance
• You will need a StorageClass for rental instances to create PersistentVolumeClaims. We recommend microceph: https://github.com/canonical/microceph
• A ResourceQuota named hyperstore in the instance namespace. This will designate how much storage the network can use on your cluster.
• Please ensure at least 150GB of free disk space on each node before installing HyperdOS. Low disk space may lead to issues with your cluster, and failed rentals.
• Please disable auto-update for NVIDIA drivers - otherwise they may update at an inopportune time and cause your stake to be slashed.

configure helm repo and dry-run

sudo microk8s helm install --dry-run hyperdos hyperdos/hyperdos --version 0.1.3 --set token="DRY_RUN_NO_TOKEN"

install (without rolling updates)

# to disable automatic updates and pin to a specific git ref
sudo microk8s helm install hyperdos hyperdos/hyperdos --version 0.1.3 --set token="<YOUR_API_KEY>"

Uninstall

remove hyperdos from the cluster

# to uninstall
sudo microk8s helm uninstall hyperdos
sudo microk8s kubectl delete app hyperweb -n argocd

delete the cluster entirely

# run these commands on each node in your cluster
sudo snap remove --purge microk8s
sudo snap remove --purge microceph