Skip to search

GPUQuota

inference.llmkube.dev / v1alpha1

apiVersion: inference.llmkube.dev/v1alpha1 kind: GPUQuota metadata: name: example
View raw schema
apiVersion string
APIVersion defines the versioned schema of this representation of an object. Servers should convert recognized schemas to the latest internal value, and may reject unrecognized values. More info: https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#resources
kind string
Kind is a string value representing the REST resource this object represents. Servers may infer this from the endpoint the client submits requests to. Cannot be updated. In CamelCase. More info: https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#types-kinds
metadata object
spec object required
GPUQuotaSpec defines the desired state of GPUQuota. Exactly one of Selector or NamespaceRef must be set; they are mutually exclusive. Selector targets pods by label across one or more namespaces; NamespaceRef pins the quota to a single namespace.
costBudgetRef string
CostBudgetRef is the name of a CostBudget resource (future CRD) that caps the dollar cost of GPU usage under this quota.
gpuCount integer required
GPUCount is the maximum number of GPUs this quota allows.
format: int32
minimum: 0
minPriority string
MinPriority is the minimum scheduling priority a pod must declare to be admitted under this quota. Pods with a lower priority are denied regardless of available headroom.
enum: critical, high, normal, low, batch
namespaceRef string
NamespaceRef pins this quota to a single namespace. Mutually exclusive with Selector.
selector object
Selector matches pods by label across one or more namespaces. Mutually exclusive with NamespaceRef.
matchExpressions []object
matchExpressions is a list of label selector requirements. The requirements are ANDed.
key string required
key is the label key that the selector applies to.
operator string required
operator represents a key's relationship to a set of values. Valid operators are In, NotIn, Exists and DoesNotExist.
values []string
values is an array of string values. If the operator is In or NotIn, the values array must be non-empty. If the operator is Exists or DoesNotExist, the values array must be empty. This array is replaced during a strategic merge patch.
matchLabels object
matchLabels is a map of {key,value} pairs. A single {key,value} in the matchLabels map is equivalent to an element of matchExpressions, whose key field is "key", the operator is "In", and the values array contains only "value". The requirements are ANDed.
vramBytes integer
VRAMBytes is the maximum total VRAM (bytes) this quota allows. Zero means no VRAM cap.
format: int64
minimum: 0
status object
GPUQuotaStatus defines the observed state of GPUQuota.
admissionDenials integer
AdmissionDenials is the cumulative count of pods denied by this quota.
format: int64
lastDenial string
LastDenial is the timestamp of the most recent admission denial.
format: date-time
usedGPUCount integer
UsedGPUCount is the number of GPUs currently allocated under this quota.
format: int32
usedVRAMBytes integer
UsedVRAMBytes is the total VRAM (bytes) currently allocated under this quota.
format: int64

No matches. Try .spec.costBudgetRef for an exact path