OpenShift AI step-by-step

⏱ ~10 min

OpenShift AI on the hub hosts the dashboard, model connections, and (in later modules) Native MaaS. This module shows where LLM credentials and edge inference (YOLO) live before you expose them through gateway routes.

Open the OpenShift AI dashboard

https://rhods-dashboard-redhat-ods-applications.apps.cluster.example.com

Log in with your workshop credentials.

Explore maas-workshop

AI Secret Flow 
hub-login guest (register first)
oc get namespace maas-workshop

In the dashboard: Data Science Projects → maas-workshop → browse model connections (Llama Scout 17B, Granite 3.2, etc.).

List MaaS model connections

oc get secret -n maas-workshop -l opendatahub.io/dashboard=true \
  -o custom-columns=NAME:.metadata.name,TYPE:.metadata.annotations.opendatahub\.io/connection-type
Expected output
NAME                             TYPE
openshift-ai-maas-credentials    generic
Secret openshift-ai-maas-credentials is synced from Vault path secret/hub/maas-credentials via External Secrets Operator (see Module 06 — Vault).

Verify DataScienceCluster on hub

hub-login guest (register first)
oc get datasciencecluster -o custom-columns=NAME:.metadata.name,PHASE:.status.phase
oc get dscinitialization -o custom-columns=NAME:.metadata.name,PHASE:.status.phase
Expected output
NAME          PHASE
default-dsc   Ready

NAME           PHASE
default-dsci   Ready

GitOps configuration

# charts/all/openshift-ai-hub/templates/all.yaml (excerpt)
apiVersion: dscinitialization.opendatahub.io/v2
kind: DSCInitialization
metadata:
  name: default-dsci
---
apiVersion: datasciencecluster.opendatahub.io/v2
kind: DataScienceCluster
metadata:
  name: default-dsc
spec:
  components:
    dashboard:
      managementState: Managed
    kserve:
      managementState: Managed
      defaultDeploymentMode: RawDeployment
# charts/all/openshift-ai-hub/values.yaml (excerpt)
maas:
  endpoint: "https://maas-rhdp.apps.maas.redhatworkshops.io"
  credentialsSecretName: openshift-ai-maas-credentials
  models:
    - id: llama-scout-17b
      displayName: Llama Scout 17B (workshop default)

Verify KServe on spokes

east-login guest (register first)
oc get inferenceservice -n neuroface-cv

Check InferenceService status

oc get inferenceservice -n neuroface-cv -o custom-columns=NAME:.metadata.name,READY:.status.conditions[0].status
Expected output
NAME                    READY
yolo-ppe-serving        True

Test MaaS from hub

hub-login guest (register first)
curl -sk https://rhods-dashboard-redhat-ods-applications.apps.cluster.example.com/ -o /dev/null -w "OpenShift AI dashboard: HTTP %{http_code}\n"
Expected output
OpenShift AI dashboard: HTTP 200

What you learned

  • OpenShift AI exposes the dashboard and maas-workshop project for exploration

  • NeuroFace chat and Lightspeed reach LLMs through AI Gateway (https://ai-gateway.apps.cluster.example.com/v1)

  • DSCInitialization and DataScienceCluster CRs are GitOps-managed

  • Platform API keys (ai-gateway-platform-keys) are auto-generated in Vault at install time

Next

Continue with Module 04 — Discover RHCL Gateway objects.