Kueue deletes workloads for jobs with limits, but no requests

[mimowo]

What happened:

Kueue creates and deletes in a loop (the cycle is <1s) a workload for a job with limits, but without requests.

What you expected to happen:

Kueue does not delete the workloads and allows the jobs to complete.

How to reproduce it (as minimally and precisely as possible):

0. Setup Kueue with the main local queue
1. Create a job from the yaml

apiVersion: batch/v1
kind: Job
metadata:
  name: job-longrun-only-limits
  annotations:
    kueue.x-k8s.io/queue-name: main
spec:
  suspend: true
  template:
    spec:
      restartPolicy: Never
      containers:
      - name: job-longrun
        image: centos:7
        resources:
          limits:
            cpu: 100m
            memory: "200Mi"
        command: ["bash"]
        args: ["-c", 'sleep 120 && echo "Hello world"']
        imagePullPolicy: IfNotPresent
  backoffLimit: 0

Anything else we need to know?:

Environment:

• Kubernetes version (use kubectl version): master (before 1.27)
• Kueue version (use git describe --tags --dirty --always): master (before 0.3.0)
• Cloud provider or hardware configuration:
• OS (e.g: cat /etc/os-release):
• Kernel (e.g. uname -a):
• Install tools:
• Others:

[alculquicondor]

This was introduced in #317.

We need to fix by changing the logic that matches a Workload and a Job. This logic should be included in the job integration library, maybe at the podtemplate level. cc @kerthcet

We should also have an integration test.

[kerthcet]

/assign
Let me fix this at first.

[kerthcet]

I'd like to default the container again then process deepEqual. See #597