(Optional) Customize Cluster Installation¶
The Run:AI Admin UI cluster creation wizard requires the download of a Helm values file
runai-<cluster-name>.yaml. You may need to edit this file. Examples:
- Disable the installation of third-party dependencies which already exist on the Kubernetes cluster
- Security review and pre-installation of Kubernetes cluster-wide access roles
- Add an ingress point
- Add an outbound internet proxy
Disable the installation of third-party dependencies¶
The default Run:AI installation packages all of its third-party dependencies. If your Kubernetes environment already has some of these dependencies, you may disable their installation.
These dependencies are:
- Node feature discovery
- NVIDIA GPU feature discovery (also contains Node feature discovery)
- Prometheus. To disable Prometheus please contact Run:AI customer support
runai-<cluster-name>.yaml and set
false for the respective third-party
Review Access Roles¶
The default Run:AI installation assumes administrative rights to create its required access roles. You may want to want to provide your IT department with the ability to view and then pre-create these roles.
If you have not done so before, run:
helm repo add runai https://run-ai-charts.storage.googleapis.com helm repo update
helm pull runai/runai-cluster --untar cd runai-cluster/templates
clusterroles folder contains the following:
base.yaml- Mandatory Kubernetes Cluster Roles and Cluster Role Bindings
project-controller-ns-creation-and-user-auth.yaml- Automatic Project Creation and Maintenance. Provides Run:AI with the ability to create Kubernetes namespaces when the Run:AI administrator creates new Projects.
project-controller-cluster-wide-secrets.yaml- allow the propagation of Secrets. See Secrets in Jobs
priorityclasses folder contains a list of Priority Classes used by Run:AI
ocp folder contains OpenShift related access files.
Pre-install Access Roles¶
Your IT department can pre-apply the above YAML files after reviewing them.
In this case, in the values file, change
Add an Ingress point¶
Set aside an IP address for ingress access to containers (e.g. for Jupyter Notebooks, PyCharm, VisualStudio Code). See: Allow external access to Containers. Note that you can access containers via port forwarding without requiring an ingress point.
Add a Proxy¶
Allow outbound internet connectivity in a proxied network environment. See Installing Run:AI with an Internet Proxy Server.