Chore: Deprecate desiredOptimizedAlloc.accelerator

[shuynh2017]

Requirements: #809
Proposal: #809 (comment)

• We will first mark as deprecated (soft deprecation) with message what to use instead of this deprecated field.

Deprecating a required field in a Kubernetes CRD with Kubebuilder involves a multi-step process for backward compatibility, primarily by relaxing the requirement in the current version and marking it as deprecated, or by introducing a new API version

• Once this PR is approved, we open an issue to remove it (breaking change) for the next API version bump.

Notes:

• Kubebuilder has this annotation // +kubebuilder:deprecatedversion:warning=, but this is for the whole api, not for just a field.
• The annotation Deprecated used in this PR is according to recommended practice.
• The only visible change to users is:

$ kubectl explain VariantAutoscaling.status.desiredOptimizedAlloc
GROUP:      llmd.ai
KIND:       VariantAutoscaling
VERSION:    v1alpha1

FIELD: desiredOptimizedAlloc <Object>


DESCRIPTION:
    DesiredOptimizedAlloc indicates the target optimized allocation based on
    autoscaling logic.

FIELDS:
  accelerator   <string>
    Accelerator is the type of accelerator for the optimized allocation.
    Deprecated: This field is deprecated and will be removed in a future                 <=====
    version. Use node selector or node affinity from scale target instead.

  lastRunTime   <string>
    LastRunTime is the timestamp of the last optimization run.

  numReplicas   <integer>
    NumReplicas is the number of replicas for the optimized allocation.
    nil means no optimization decision has been made yet.

[shuynh2017]

@ev-shindin. Pls help to review. Thanks.

[ev-shindin]

Look at my comment. Also ensure that both CRD YAML files are generated by make manifests from the Go types + kubebuilder markers.

[shuynh2017]

/ok-to-test

[github-actions]

🚀 Kind E2E (full) triggered by /ok-to-test

View the Kind E2E workflow run

[github-actions]

🚀 OpenShift E2E — approve and run (/ok-to-test)

View the OpenShift E2E workflow run

[github-actions]

GPU Pre-flight Check ✅

GPUs are available for e2e-openshift tests. Proceeding with deployment.

Resource	Total	Allocated	Available
GPUs	50	31	19

Cluster	Value
Nodes	16 (7 with GPUs)
Total CPU	993 cores
Total Memory	10383 Gi
GPUs required	4 (min) / 6 (recommended)

[shuynh2017]

/ok-to-test

[github-actions]

🚀 Kind E2E (full) triggered by /ok-to-test

View the Kind E2E workflow run

[github-actions]

🚀 OpenShift E2E — approve and run (/ok-to-test)

View the OpenShift E2E workflow run

[github-actions]

GPU Pre-flight Check ✅

GPUs are available for e2e-openshift tests. Proceeding with deployment.

Resource	Total	Allocated	Available
GPUs	50	30	20

Cluster	Value
Nodes	16 (7 with GPUs)
Total CPU	993 cores
Total Memory	10383 Gi
GPUs required	4 (min) / 6 (recommended)

[ev-shindin]

Great work! Thanks!

[lionelvillard]

/lgtm