OpenShift Integration Operator

What this operator adds

OpenShift Integration Operator is the lifecycle and UX layer for Apache Camel on OpenShift. It does not replace Camel Quarkus. You design flows visually, try them ephemerally, then promote them to GitOps.

The operator is listed in the OperatorHub community catalog (community-operators, package openshift-integration-operator, channel candidate-v0). Current CSV on OperatorHub is 0.8.2. It is community software (Apache 2.0), not a Red Hat supported product.

This full lab installs it with the Validated Patterns clusterGroup.subscriptions entry in values-hub.yaml. OpenShift GitOps reconciles a Subscription in namespace openshift-integration (AllNamespaces). After the CSV is Succeeded, the OpenShift Console shows Integration Platform.

This module complements Kaoto in DevSpaces. DevSpaces edits YAML routes in a workspace. The console plugin edits IntegrationFlow custom resources on the cluster.

Operator reconciliation: IntegrationFlow CR to worker Deployment

How this lab installs it

The Pattern creates namespace openshift-integration with a cluster-scoped OperatorGroup and this Subscription:

apiVersion: operators.coreos.com/v1alpha1
kind: Subscription
metadata:
  name: openshift-integration-operator
  namespace: openshift-integration
spec:
  channel: candidate-v0
  name: openshift-integration-operator
  source: community-operators
  sourceNamespace: openshift-marketplace
  installPlanApproval: Automatic

Lite (values-lite.yaml) leaves this Subscription enabled so you still get Kaoto in the console when DevSpaces is off.

Open the console plugin

  1. Log in to the OpenShift Console.

  2. Open Integration PlatformIntegration Flows.

  3. Search, filter by type or phase, and follow links to Pods and Tekton Pipelines.

Integration Flows list — search

Console tour

Screenshots below come from the operator documentation site.

Inspect the visual diagram

The SVG diagram shows saga branching (choice, multicast). Click a node to highlight the matching YAML.

Visual saga flow diagram with branching logic

Edit in Kaoto

The plugin embeds the Kaoto canvas so you drag Camel components in the console — the same designer you use in DevSpaces, bound to spec.kaotoDesign.

Kaoto designer embedded in the OpenShift Console

Sync YAML to Git

In GitOps mode the YAML editor keeps kaotoDesign in sync with the configured repository.

YAML editor for kaotoDesign with Git sync

Configure spec and status

The spec panel exposes deployment mode, engine, targeting, resilience, and circuit-breaker state.

IntegrationFlow spec and status — resilience

Check platform health

The status dashboard reports Operator API, Kaoto, Gitea, Argo CD, Tekton, and OpenTelemetry Collector.

Platform Status — Operator

Monitor KPIs

Overview cards show total flows, running count, error rate, and distribution by type and mode.

Flow Overview — KPIs

Stream worker logs

The Logs tab tails the worker pod (ephemeral or GitOps), with container selector and follow mode.

Flow logs — live worker pod streaming

Pause, resume, stop

Lifecycle controls confirm Pause, Resume, and Stop against the current phase.

Lifecycle controls — Pause

Quick Try (ephemeral)

Quick Try deploys a precompiled Camel Quarkus worker from kaotoDesign without Git, Tekton, or Argo CD. Set deploymentMode: EPHEMERAL and a TTL. Deleting the CR removes owned Deployments.

Apply this sample in openshift-integration (or any namespace the operator watches):

apiVersion: platform.io/v1alpha1
kind: IntegrationFlow
metadata:
  name: ephemeral-camel-demo
  namespace: openshift-integration
spec:
  engine: CAMEL
  integrationType: CAMEL_ROUTE
  deploymentMode: EPHEMERAL
  kaotoDesign: |
    - route:
        id: quick-try
        from:
          uri: "timer:tick"
          parameters:
            period: 10000
        steps:
          - log:
              message: "Ephemeral quick-try flow"
  ephemeral:
    ttlSeconds: 3600
oc apply -f ephemeral-camel-demo.yaml
oc get integrationflow ephemeral-camel-demo -n openshift-integration -w

The CDC processor in kafka-cdc stays a Camel Quarkus Deployment from the cdc-pipeline chart. Use IntegrationFlow for new routes you want to try in the console, then promote to GitOps when the route is ready.

Do not point a long-running workshop flow at production credentials from Quick Try. TTL expires the worker; extend it from the console if you need more time.

Workers, MCP, and telemetry

The operator selects a precompiled worker image from the route (or you pin spec.ephemeral.workerImage). Domain images cover core, messaging, HTTP, data, cloud, and AI extensions.

HPA v2 scales each IntegrationFlow worker independently

Camel routes can expose MCP tools, and flows can call models. That complements Lightspeed MCP and AI Playground + Kafka MCP already in this lab.

MCP bridge — AI tools discover and invoke Camel routes

Workers emit OpenTelemetry. The console colors canvas nodes from an SSE telemetry stream.

OpenTelemetry collector to console SSE overlay

Verify on the cluster

oc get csv -n openshift-integration
oc get subscription openshift-integration-operator -n openshift-integration
oc get consoleplugin
oc get crd integrationflows.platform.io

Expect CSV openshift-integration-operator.v0.8.2 (or newer on channel candidate-v0) in phase Succeeded, and ConsolePlugin integration-console-plugin. Then open Integration Platform in the console.

Official documentation