[Bug]: About PVC

[Json0215]

Description

I am using the Slurm Operator to deploy a Slurm cluster on Kubernetes for Machine Learning workloads. I need to mount a shared NFS PVC (ReadWriteMany) to both the slurm-controller and slurm-worker pods at a specific path (e.g., /data) to share code and datasets. However, after configuring volumes and volumeMounts in the Helm values (under controller and nodesets), the resulting Pods do not contain these mount points. It seems the Operator or the Helm chart is ignoring these custom volume definitions in the CRD spec.How should i config this?

Steps to Reproduce

The content of values.yaml is as follows:

controller:
  # 挂载共享 PVC 到 Controller
  volumeMounts:
    - name: shared-home
      mountPath: /home  # 目录挂载
  volumes:
    - name: shared-home
      persistentVolumeClaim:
        claimName: slurm-common-pvc

configFiles:
  gres.conf: |
    AutoDetect=nvidia

# 3. 计算节点配置 (NodeSet)
nodesets:
  slinky:
    replicas: 1
    nodeSelector:
      kubernetes.io/hostname: "k8s-n3"
    extraConf: "Gres=gpu:1"
    
    volumeMounts:
      - name: shared-home
        mountPath: /home
    volumes:
      - name: shared-home
        persistentVolumeClaim:
          claimName: slurm-common-pvc

    slurmd:
      resources:
        limits:
          nvidia.com/gpu: "1"
        requests:
          nvidia.com/gpu: "1"
    
    partition:
      enabled: true

where slurm-common-pvc is the name of pvc ，which is defined as follow:

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: slurm-common-pvc
  namespace: slurm
spec:
  accessModes:
    - ReadWriteMany
  storageClassName: nfs-client  
  resources:
    requests:
      storage: 15Gi

Expected Behavior

Additional Context