fix(helm): wire controller manager config so leader election from ConfigMap is actually used

[boussoufiane]

What happened?
Problem
When deploying with replicaCount > 1, multiple operator pods run concurrently without leader election being effectively enabled, because the mounted controller_manager_config.yaml (which sets leaderElection.leaderElect: true) is never passed to the manager process.

Root cause
In fact , The Deployment starts /manager without --config=…, while the ConfigMap is mounted at /controller_manager_config.yaml. As a result, the binary falls back to CLI defaults (--leader-elect defaults to false), so no Lease objects are created and HA replicas act as concurrent controllers instead of leader / standby.

Did you expect to see something different?
having one leader instead of 2 concurent Pods

How to reproduce it (as minimally and precisely as possible):
Replica = 2 for operator and cass-operator
Environment

• K8ssandra Operator version:

  docker.io/k8ssandra/k8ssandra-operator@sha256:3fb182ff0317687682dacdfb272c5ad03440d12dd889965964951816b5de3acb

• Kubernetes version information:

NAME                STATUS   ROLES           AGE    VERSION
oi-test-master-a1   Ready    control-plane   324d   v1.31.4
oi-test-master-b1   Ready    control-plane   324d   v1.31.4
oi-test-master-c1   Ready    control-plane   324d   v1.31.4
oi-test-worker-a1   Ready    <none>          324d   v1.31.4
oi-test-worker-a2   Ready    <none>          324d   v1.31.4
oi-test-worker-a3   Ready    <none>          240d   v1.31.4
oi-test-worker-a4   Ready    <none>          175d   v1.31.4
oi-test-worker-b1   Ready    <none>          324d   v1.31.4
oi-test-worker-b2   Ready    <none>          324d   v1.31.4
oi-test-worker-b3   Ready    <none>          240d   v1.31.4
oi-test-worker-b4   Ready    <none>          175d   v1.31.4
oi-test-worker-c1   Ready    <none>          324d   v1.31.4
oi-test-worker-c2   Ready    <none>          324d   v1.31.4
oi-test-worker-c3   Ready    <none>          240d   v1.31.4
oi-test-worker-c4   Ready    <none>          175d   v1.31.4

• Kubernetes cluster kind:

  insert how you created your cluster: Kubespray

• Manifests:

https://github.com/k8ssandra/k8ssandra-operator/blob/v1.31.0/charts/k8ssandra-operator/templates/deployment.yaml#L37

apiVersion: apps/v1
kind: Deployment
metadata:
  name: {{ include "k8ssandra-common.fullname" . }}
  labels: {{ include "k8ssandra-common.labels" . | indent 4 }}
    control-plane: k8ssandra-operator
  {{- with include "k8ssandra-common.annotations" . }}
  annotations:
    {{- . | nindent 4 }}
  {{- end }}
spec:
  replicas: {{ .Values.replicaCount }}
  selector:
    matchLabels:
      {{- include "k8ssandra-common.selectorLabels" . | nindent 6 }}
      control-plane: k8ssandra-operator
  template:
    metadata:
      {{- with include "k8ssandra-common.annotations" (dict "context" . "annotations" .Values.podAnnotations) }}
      annotations:
        {{- . | nindent 8 }}
      {{- end }}
      labels:
        {{- include "k8ssandra-common.labels" . | indent 8 }}
        control-plane: k8ssandra-operator
    spec:
      {{- with .Values.imagePullSecrets }}
      imagePullSecrets:
        {{- toYaml . | nindent 8 }}
      {{- end }}
      serviceAccountName: {{ include "k8ssandra-common.serviceAccountName" . }}
      securityContext:
        {{- toYaml .Values.podSecurityContext | nindent 8 }}
      containers:
      - command:
        - /manager
        # args:
        # - --config=controller_manager_config.yaml

BTW , i do not know why it is commented ? it seems to be the cause of the issue

• K8ssandra Operator Logs:

insert K8ssandra Operator logs relevant to the issue here

Anything else we need to know?:

kubectl get lease -A | grep k8ssandra gives nothing and in the logs i see duplicate events .
so this confirm that there is no election

:Proposed change
Uncomment / add container args so the manager loads the mounted config file, e.g. --config=/controller_manager_config.yaml (path aligned with the volumeMount).
Optionally expose extraArgs / manager.extraArgs in values.yaml for flexibility without forking the chart.