Pods in lower-weighted queue are not being evicted to make resources for Pods in greater weighted queue

[jchatter123]

What happened:

Pods in lower weighted queue continue running when Pods in higher-weighted queue needs resources

What you expected to happen:

Pods in lower priority queue would be evicted and resources would be returned back to higher-weighted queue to run Pods.

Extra Information:

• The total amount of resources available:

  • CPU: 1200m

• There are two queues

  • default and test

• The default queue has weight 1

  • This queue has a capacity greater than 1200m

• The test queue has weight 5

  • This queue has a capacity of 2 cpu (2000m)

• I’m scheduling a job (job-1) in the default queue that creates 5 pods, each taking 200m in cpu resources

  • Job-1 pods are taking 1000m in cpu resources altogether

• I’m scheduling a second job (job-2) in the default queue that creates 1 pod, which takes 200m in cpu resources.

• At this point, the default queue is using all cpu resources (1200m) and all pods of the two jobs are running successfully.

• I now schedule a third job (job-3) in the test queue which creates 5 pods, each taking 200m in cpu resources

• I expect the test queue (with weight 5) to take resources from the default queue to run the 5 pods of job-3

• However, this does not occur.

  • Instead, pods for job-3 remain pending until pods in the default queue finish and release resources

Output:

• The Pod Listing:

YAML :

• queue.yaml (YAML for test queue)

• vcjob.yaml (YAML for job-1)

• vcjob2.yaml (YAML for job-2)

• vcjob3.yaml (YAML for job-3)

Describing Resources

• Pod Group for job-1

• Pod Group for job-2

• Pod group for job-3

• Volcano Config Map

• The Default Queue

• The Test Queue

Environment:

• Volcano Version: 1.6.0
• Kubernetes version (use kubectl version):

• Cloud provider or hardware configuration:
  • Not using a cloud provider
  • Running minikube locally

[stale]

Hello 👋 Looks like there was no activity on this issue for last 90 days.
Do you mind updating us on the status? Is this still reproducible or needed? If yes, just comment on this PR or push a commit. Thanks! 🤗
If there will be no activity for 60 days, this issue will be closed (we can always reopen an issue if we need!).

[stale]

Closing for now as there was no activity for last 60 days after marked as stale, let us know if you need this to be reopened! 🤗