clearml-helm-charts/charts/clearml-serving/templates/clearml-serving-inference-hpa.yaml

{{- if .Values.clearml_serving_inference.autoscaling.enabled }}
apiVersion: {{ include "common.capabilities.hpa.apiVersion" ( dict "context" $ ) }}
kind: HorizontalPodAutoscaler
metadata:
  name: {{ include "clearmlServing.fullname" . }}-inference-hpa
  namespace: {{ .Release.Namespace | quote }}
  annotations: {}
  labels:
    clearml.serving.service: {{ include "clearmlServing.fullname" . }}-inference
spec:
  scaleTargetRef:
    apiVersion: "apps/v1"
    kind: Deployment
    name: {{ include "clearmlServing.fullname" . }}-inference
  minReplicas: {{ .Values.clearml_serving_inference.autoscaling.minReplicas }}
  maxReplicas: {{ .Values.clearml_serving_inference.autoscaling.maxReplicas }}
  metrics:
    {{- if .Values.clearml_serving_inference.autoscaling.targetCPU }}
    - type: Resource
      resource:
        name: cpu
        {{- if semverCompare "<1.23-0" (include "common.capabilities.kubeVersion" .) }}
        targetAverageUtilization: {{ .Values.clearml_serving_inference.autoscaling.targetCPU }}
        {{- else }}
        target:
          type: Utilization
          averageUtilization: {{ .Values.clearml_serving_inference.autoscaling.targetCPU }}
        {{- end }}
    {{- end }}
    {{- if .Values.clearml_serving_inference.autoscaling.targetMemory }}
    - type: Resource
      resource:
        name: memory
        {{- if semverCompare "<1.23-0" (include "common.capabilities.kubeVersion" .) }}
        targetAverageUtilization: {{ .Values.clearml_serving_inference.autoscaling.targetMemory }}
        {{- else }}
        target:
          type: Utilization
          averageUtilization: {{ .Values.clearml_serving_inference.autoscaling.targetMemory }}
        {{- end }}
    {{- end }}
{{- end }}