Developer Deployment Instructions
Prerequisites
- RHEL 8.x jumpbox that is registred with Red Hat
- Review and run the openshift-ai-workload.sh
- If using demo.redhat.com use
RHEL 8 Base
instance to configure a jumpbox and run the fork and run rhel8-bootstraphttps://github.com/tosin2013/rhel8-bootstrap/actions/workflows/configure-rhel8-bastion.yml
to run GitHub Actions to configure the jumpbox. Igf you do not want to run this on your laptop.
SSH into RHEL jumpbox
Deploying GPUs for AI Workloads on OpenShift on AWS
Configure SSL for OpenShift Console
Configure jumpbox for deployment
Deploy Gitea
Configure Gitea
Click on route in OpenShift Console (1) Click on Register (2) Register a new account (3) Click on New Migration click on Git (4) Click on Migrate From Git
- https://github.com/tosin2013/redhat-edge-ai-industrial-demo-infra.git (5) Edit the following folders in the repo
- Replace with your gita URL
File path:clusters/overlays/aws/patch-applicationset-repo-revision.yaml
File path 2:clusters/overlays/aws/patch-application-repo-revision.yaml
Apply configure changes to argocd
Ensure the ArgoCD application is in sync
Configure Enviornment
- Run tekton pipeline - The Tekton pipeline is used to build the container image for the application.
- Configure rhods - The Red Hat OpenShift Data Science environment is used to train the AI model.
- Run rhods Notebooks - Run notebooks to train the AI model.