Working
RKE2 DC Kafka is healthy. Direct TCP/TLS SNI to RKE2 node IPs and HAProxy IP on 443 succeeds from both spokes.
The next demo app should live in the workload repo, run on the selected OCP spoke, and consume RKE2 Kafka only after DNS, trust, and app-specific authentication are ready.
Architecture
| Layer | Repository or platform | Expected artifacts |
|---|---|---|
| Application source | New app repo or workload source tree | Java/JBoss code, Maven or Gradle build, Containerfile, unit tests. |
| Workload desired state | lab-workloads/ | Namespace, Deployment, Service, ConfigMap, runtime Secret reference, health probes, route or mesh resources. |
| Platform desired state | lab-gitops-full/ | Only shared platform prerequisites, policy, operators, or cluster-level resources. |
| Kafka platform | RKE2 Kafka | App-specific KafkaUser, ACLs, topic policy, cluster CA distribution, runtime SCRAM credentials. |
Kafka readiness
RKE2 DC Kafka is healthy. Direct TCP/TLS SNI to RKE2 node IPs and HAProxy IP on 443 succeeds from both spokes.
bootstrap.kafka.apps.sub.comptech-lab.com and broker-*.kafka.apps.sub.comptech-lab.com do not resolve yet. External 9094 is not reachable.
bootstrap.servers=bootstrap.kafka.apps.sub.comptech-lab.com:443
security.protocol=SASL_SSL
sasl.mechanism=SCRAM-SHA-512
Before deployment
*.kafka.apps.sub.comptech-lab.com to the HAProxy IP.mm2 or redis-applier.spoke-dc, then decide whether spoke-dr is active standby.Runtime shape