This guide explains how to extend the intelligence of OpenShift Lightspeed by integrating specialized knowledge for Kiali and Istio. By configuring a custom Retrieval-Augmented Generation (RAG) database, you ensure the service provides precise, context-aware assistance for your service mesh queries.

The OpenShift Lightspeed service leverages Large Language Models (LLMs) to provide intelligent, context-aware responses. To ensure the model has deep expertise in your specific environment, you can use the Bring Your Own Knowledge (BYOK) tool to integrate a RAG database.

By connecting this database, you bridge the gap between general AI knowledge and specific technical documentation, ensuring more accurate troubleshooting and configuration advice for service mesh workloads.

We have curated and optimized specialized knowledge sets for the following components:

Because Kiali and Istio are tightly related, both knowledge bases are merged into a single index at build time for richer, unified context. The image is published at:

quay.io/kiali/kiali-byok:v0.0.1

• You are logged in to the OpenShift Container Platform web console as a user with the cluster-admin role.
• You have an LLM provider available for use with the OpenShift Lightspeed Service.
• You have installed and configured the OpenShift Lightspeed Operator.

Modify the OLSConfig CR to deploy the pre-packaged RAG database alongside the existing one:

1. In the OpenShift Container Platform web console, click Operators → Installed Operators.
2. Select All Projects in the Project dropdown.
3. Click OpenShift Lightspeed Operator.
4. Click OLSConfig, then click the cluster configuration instance in the list.
5. Click the YAML tab.
6. Insert the spec.ols.rag section as shown below and click Save.

apiVersion: ols.openshift.io/v1alpha1
kind: OLSConfig
metadata:
  name: cluster
spec:
  ols:
    rag:
      - image: quay.io/kiali/kiali-byok:v0.0.1
        indexPath: /rag/vector_db/kiali
        indexID: kiali-istio-docs

Note: Kiali and Istio documentation are merged into a single index (kiali-istio-docs) at build time, so only one entry is needed in the CR.

After saving, the operator will restart the Lightspeed pod with the new indexes mounted. To confirm the indexes loaded correctly, check the pod logs:

oc logs -n openshift-lightspeed deployment/lightspeed-app-server | grep -E "vector index|All indexes"

Expected output:

INFO: Loading vector index #0...
INFO: Vector index #0 is loaded.
INFO: Loading vector index #1...
INFO: Vector index #1 is loaded.
INFO: All indexes are loaded.

Once loaded, Lightspeed will automatically retrieve relevant Kiali and Istio documentation when answering service mesh questions, and cite the source pages in its responses.

See the releases page for the full list.