After GKE cluster update, kafka is not working.

[nevenara]

Hello, I installed strimzi using helm chart and kafka objects using deployment file. After cluster update pods main-kafka-0 and main-zookeeper-o dissapeer and therre is no chance to recreate it again on that cluster. After deleting ns and recreating it, PVC for zookeper and kafka doesn't want to be created (waiting for first consume). Do you have some idea how this can be solved. Thank you in advance.

apiVersion: kafka.strimzi.io/v1beta2
kind: Kafka
metadata:
namespace: kafka
name: main
spec:
kafka:
replicas: 1
listeners:
- name: plain
port: 9092
type: internal
tls: false
- name: tls
port: 9093
type: internal
tls: true
authentication:
type: tls
- name: external
port: 9094
type: loadbalancer
tls: true
authentication:
type: tls
configuration:
bootstrap:
annotations:
cloud.google.com/load-balancer-type: Internal
brokers:
- broker: 0
annotations:
cloud.google.com/load-balancer-type: Internal
storage:
type: jbod
volumes:
- id: 0
type: persistent-claim
size: 200G
deleteClaim: false
config:
offsets.topic.replication.factor: 1
transaction.state.log.replication.factor: 1
transaction.state.log.min.isr: 1
zookeeper:
replicas: 1
storage:
type: persistent-claim
size: 10G
deleteClaim: false
entityOperator:
topicOperator: {}

[scholzj]

You would have to provide the full logs from the operator, the configuration in a way where it is properly formatted and readable etc. Without that, nobody can tell you what is happening in your cluster.

[nevenara1234]

Hi, I attached logs.
kafka-logs.txt
kafka-pvc.txt
Tnx

[scholzj]

The log is not complete. So it is hard to say what the root issue might be from it. You have to share the full logs. If we stick with what you shared ... then it says quite clearly that:

Exceeded timeout of 300000ms while waiting for Pods resource main-zookeeper-0 in namespace kafka to be ready

So, the Pod is not ready ... what is is doing? What state is it in? Etc.