Use Kueue as default

[Bobbins228]

Issue link

RHOAIENG-1058

What changes have been made

KubeRay and Kueue are now the primary resources responsible for creating Ray Clusters through the SDK.
mcad=False by default meaning Ray Clusters are created via Kube Ray and Routes/Ingresses are created seperately.

The local queue label is applied to the Ray Cluster when mcad=False.
The SDK will try to gather the default local queue's name by reading the annotation "kueue.x-k8s.io/default-queue":"true" from the local queue. If the SDK is unsuccessful it will raise an error asking for the LQ name to be specified in the ClusterConfiguration.

Verification steps

There are two ways to test this.

1. SDK gathers the default local queue name
2. User provides the local queue name

1: SDK gathers the default local queue name

• Create a Local Queue with the new annotation (or add the annotation to your own LQ).

apiVersion: kueue.x-k8s.io/v1beta1
kind: LocalQueue
metadata:
    namespace: ns
    name: local_queue_name
    annotations:
      "kueue.x-k8s.io/default-queue": "true"
spec:
  clusterQueue: cluster_queue_name

• Run through a default notebook.
• A yaml file will be generated with the kueue.x-k8s.io/queue-name: LQ_NAME label
• Create the Ray Cluster
• Check if a Workload was created for the Ray Cluster

2: User provides the Local Queue name

Note: This method is taken into account over the SDK gathering the Local Queue name

• Run through a demo notebook
• Add the parameter local_queue="LOCAL_QUEUE_NAME" to the ClusterConfiguration
• Check the generated yaml file for the appropriate label
• Create the Ray Cluster
• Check if a Workload was created for the Ray Cluster

Other Changes

The status for RayClusters includes suspended now.
Altered list_all_queued and list_all_clusters to reflect this.
Added the functions above to the simple import scheme.

Note: This PR should not be merged until RHOAIENG-1056 and RHOAIENG-3270 are merged.

Checks

• I've made sure the tests are passing.
• Testing Strategy
  • Unit tests
  • Manual tests
  • Testing is not required for this change

[KPostOffice]

Nice! A few comments

[dimakis]

@Bobbins228 This PR needs to be carefully timed as to when it will be merged. just making sure this is thought about before someone merges it

[astefanutti]

/lgtm

[openshift-ci]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please ask for approval from astefanutti. For more information see the Kubernetes Code Review Process.

The full list of commands accepted by this bot can be found here.

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[dgrove-oss]

/lgtm

I'll rebase #484 ontop of this as soon as this is merged. There will be some minor merge conflicts, but should be manageable.

[Bobbins228]

@dgrove-oss
This PR now just depends on the two following issues and it can be merged 👍
https://issues.redhat.com/browse/RHOAIENG-1056
https://issues.redhat.com/browse/RHOAIENG-1031

[dimakis]

@dgrove-oss This PR now just depends on the two following issues and it can be merged 👍 https://issues.redhat.com/browse/RHOAIENG-1056 https://issues.redhat.com/browse/RHOAIENG-1031

@Bobbins228 https://issues.redhat.com/browse/RHOAIENG-1031 has been merged by the looks of it so we're now just waiting for #495

[astefanutti]

/lgtm

[openshift-ci]

New changes are detected. LGTM label has been removed.