[Bug] can not set submitter pod restart times

[xiaoyu1095]

Search before asking

• [X] I searched the issues and found no similar issues.

KubeRay Component

ray-operator

What happened + What you expected to happen

In my scenario, if the execution fails, it should be considered a failure and no retries are needed. However, I cannot disable retries entirely.

Reproduction script

submitterPodTemplate: spec: restartPolicy: Never containers:

• image: harbor.thupx.cn/docker.io/rayproject/ray:2.34.0 imagePullPolicy: IfNotPresent name: ray-job-submitter

Anything else

No response

Are you willing to submit a PR?

• [ ] Yes I am willing to submit a PR!

[xiaoyu1095]

apiVersion: ray.io/v1 kind: RayJob metadata: name: test spec: entrypoint: python main.py jobId: "123" submitterPodTemplate: spec: restartPolicy: Never containers:

• image: rayproject/ray:2.34.0 imagePullPolicy: IfNotPresent name: ray-job-submitter shutdownAfterJobFinishes: true rayClusterSpec: headGroupSpec: rayStartParams: dashboard-host: 0.0.0.0 template: spec: containers:
  • image: rayproject/ray:2.34.0 imagePullPolicy: IfNotPresent name: ray-master ports:
    • containerPort: 6379 name: gcs-server protocol: TCP
    • containerPort: 8265 name: dashboard protocol: TCP
    • containerPort: 10001 name: client protocol: TCP
    • containerPort: 8000 name: serve protocol: TCP volumeMounts:
    • mountPath: /tmp/ray/ name: log-volume rayVersion: 2.34.0 workerGroupSpecs:
      • groupName: small-group maxReplicas: 1 minReplicas: 1 replicas: 1 rayStartParams: {} template: spec: containers:
        • image: rayproject/ray:2.34.0 name: ray-worker resources: limits: cpu: 500m memory: 8Gi nvidia.com/gpu: 0 requests: cpu: 500m memory: 8Gi nvidia.com/gpu: 0

[andrewsykim]

Try setting submitterConfig.backoffLimit = 0