[Bug] Job CRD status not updated upon job completion

[z103cb]

Search before asking

• I searched the issues and found no similar issues.

KubeRay Component

ray-operator

What happened + What you expected to happen

When submitting a KubeRay job (ray job CRD) to be executed on a previously provisioned cluster, the ray job status is not set to v1alpha1.JobStatusSucceeded.

The underlying batchv1.Job status is completed.

Reproduction script

To reproduce the issue:

1. Create k8s cluster
2. Deploy operator (v1.0.0-rc.0+) in default namespace
3. Create ray-system namespace: kubectl create namespace ray-system
4. Create a KubeRay cluster: kubectl apply -f cluster.yaml.txt
5. Wait for cluster to be up an running
6. Create a KubeRay Job: kubectl apply -f job.yaml.txt
7. Watch the job which doesn't have its status updated.

The job batchv1.Job has a status of

status:
  completionTime: "2023-10-11T06:56:19Z"
  conditions:
  - lastProbeTime: "2023-10-11T06:56:19Z"
    lastTransitionTime: "2023-10-11T06:56:19Z"
    status: "True"
    type: Complete
  startTime: "2023-10-11T06:55:46Z"
  succeeded: 1

Also, deleting the ray job doesn't delete the associated batchv1.Job

kubectl delete rayjob -n ray-system sample-job
# rayjob.ray.io "sample-job" deleted
kubectl get job -n ray-system sample-job

# NAME         COMPLETIONS   DURATION   AGE
# sample-job   1/1           33s        28m

Anything else

Fairly consistent.

Are you willing to submit a PR?

• Yes I am willing to submit a PR!

[astefanutti]

@kevin85421 could you please assign it to me?