Skip to content

openshift-deployment.md

Latest commit

 

History

History
147 lines (98 loc) · 4.01 KB

openshift-deployment.md

File metadata and controls

147 lines (98 loc) · 4.01 KB

Deploying Agents to Red Hat OpenShift

This guide covers deploying any agent from this repository to an OpenShift cluster using Helm.

Prerequisites

  • oc CLI installed
  • Helm 3 installed
  • Podman installed (recommended) or Docker — only needed for Option A below
  • Access to a container registry (e.g., Quay.io) — only needed for Option A below
  • An OpenShift cluster with permissions to create Deployments, Services, and Routes

Steps

1. Login to OpenShift and Container Registry

# Token-based login (recommended — avoids credentials in shell history)
oc login --token=<token> --server=https://<cluster-api-url>

If using Option A (local build + push), also log in to your container registry:

# Quay.io
podman login quay.io

# Docker Hub
podman login docker.io

# OpenShift internal registry
podman login $(oc get route default-route -n openshift-image-registry -o jsonpath='{.spec.host}')

2. Choose an Agent

Navigate to the agent you want to deploy:

cd agents/langgraph/react_agent   # or any other agent

3. Configure Environment

make init        # creates .env from .env.example

Edit .env:

API_KEY=your-api-key-here
BASE_URL=https://<model-endpoint>/v1
MODEL_ID=llama3.2:3b

4. Build the Container Image

Option A: Build locally and push to a registry

Requires Podman (or Docker) and a registry account (e.g., Quay.io).

make build    # builds the image locally
make push     # pushes to the registry specified in CONTAINER_IMAGE

The Makefile auto-detects Podman or Docker (preferring Podman).

Option B: Build in-cluster via OpenShift BuildConfig

No Podman, Docker, or registry account needed — just the oc CLI.

make build-openshift

This creates a BuildConfig (if it doesn't exist) and uploads your local source to OpenShift, which builds the image in-cluster using its internal registry.

After the build completes, set CONTAINER_IMAGE in your .env to the internal registry URL:

CONTAINER_IMAGE=image-registry.openshift-image-registry.svc:5000/<namespace>/<agent-name>:latest

Replace <namespace> with your OpenShift project name (run oc project -q to check).

5. Preview Rendered Manifests (optional)

Before deploying, you can inspect exactly what Kubernetes resources will be created:

make dry-run

Secrets are redacted in the output.

6. Deploy with Helm

make deploy

Under the hood, make deploy runs helm upgrade --install with your .env values, passing secrets via a temporary file that is cleaned up after deployment. If any required environment variables are missing, it will fail with a clear error listing which variables need to be set.

7. Verify

oc get pods -l app=<agent-name>
oc get route <agent-name>

The route URL is your agent's public endpoint.

8. Remove

make undeploy

Customizing Deployment

Each agent has a values.yaml that overrides the shared chart defaults at charts/agent/values.yaml. You can:

  • Change resources: edit resources.requests / resources.limits in the agent's values.yaml

  • Disable OpenShift Route and use K8s Ingress instead:

    helm upgrade --install <agent-name> ../../charts/agent \
  -f values.yaml \
  --set openshift.route.enabled=false \
  --set ingress.enabled=true \
  --set ingress.host=my-agent.example.com

  • Add environment variables: add entries to env: in the agent's values.yaml

  • Add volumes: add entries to volumes: and volumeMounts: (see the agentic_rag agent for an example)

Shared Helm Chart

All agents share a single Helm chart at charts/agent/. The override chain is:

charts/agent/values.yaml        <-- global defaults
  agents/.../values.yaml        <-- agent-specific overrides
    --set flags                 <-- CLI overrides at deploy time