CFK Cluster link

Ops Cluster Replication
1

Hi

I have created this ClusterLink, but kafka pods reports this errors en destination cluster:

[ERROR] 2025-02-16 16:54:05,790 [kafka-admin-client-thread | cluster-link-client-admin-clusterlink-tfn-global-destination-2] org.apache.kafka.clients.NetworkClient initiateReverseConnect - [AdminClient clientId=cluster-link-client-admin-clusterlink-tfn-global-destination-2] Error initiating reverse connection to node localhost:0 (id: -1 rack: null)

kind: ClusterLink
apiVersion: platform.confluent.io/v1beta1
metadata:
name: clusterlink-tfn-global
namespace: kafkades
spec:
configs:
connections.max.idle.ms: "350000"
cluster.link.retry.timeout.ms: "900000"
sourceInitiatedLink:
linkMode: Destination
destinationKafkaCluster:
kafkaRestClassRef:
name: kafkarestclass
namespace: kafkades
sourceKafkaCluster:
authentication:
type: mtls
tls:
enabled: true
secretRef: kafka-tls-secret-kafkaadmin
bootstrapEndpoint: kafkapro.sistemasti.int:443
clusterID: pm9MXZPnTqqhRc7PjV7U8w
kafkaRestClassRef:
name: kafkarestclass
namespace: kafkapro
mirrorTopics:
- name: test
mirrorTopicOptions:
autoCreateTopics:
enabled: true
2

Hi, There might be a configuration issue, such as a missing or incorrect bootstrap server definition in the cluster link configuration. An error with localhost:0 suggests that the bootstrap server is not correctly set. You can check the kafka properties file and ensure that the correct bootstrap server is specified.

3

Hi

bootstrap server definition in the cluster link configuration is correct. The cluster source is expoxed by openshift route and the bootstrapEndpoint is this route of OpenShift; perhaps the issue is because the reverse connection doesn´t work fine witk openshift route / haproxy balancer.