add HorizontalPodAutoscaler to catalog

serving-catalog/core/deployment/base/hpa.yaml

@@ -0,0 +1,19 @@
+apiVersion: autoscaling/v2
+kind: HorizontalPodAutoscaler
+metadata:
+  name: singlehost-inference-hpa
+spec:
+  scaleTargetRef:
+    apiVersion: apps/v1
+    kind: Deployment
+    name: singlehost-inference-deployment-blueprint
+  minReplicas: 1
+  maxReplicas: 10
+  metrics:
+  - type: Pods
+    pods:
+      metric:
+        name: token-latency-ms
+      target:
+        type: AverageValue
+        averageValue: 50

serving-catalog/core/deployment/base/kustomization.yaml

@@ -4,3 +4,4 @@ kind: Kustomization
 
 resources:
 - deployment.yaml
+- hpa.yaml

serving-catalog/core/deployment/jetstream/gemma-7b-it/base/hpa.patch.yaml

@@ -0,0 +1,6 @@
+apiVersion: autoscaling/v2
+kind: HorizontalPodAutoscaler
+metadata:
+  labels:
+    app: gemma-7b-it-jetstream-inference-server
+  name: gemma-7b-it-jetstream-hpa

serving-catalog/core/deployment/jetstream/gemma-7b-it/base/kustomization.yaml

@@ -11,6 +11,11 @@ patches:
       kind: Deployment
     options:
       allowNameChange: true
+  - path: hpa.patch.yaml
+    target:
+      kind: HorizontalPodAutoscaler
+    options:
+      allowNameChange: true
   - path: job.patch.yaml
     target:
       kind: Job

serving-catalog/core/deployment/jetstream/llama3-8b/base/hpa.patch.yaml

@@ -0,0 +1,6 @@
+apiVersion: autoscaling/v2
+kind: HorizontalPodAutoscaler
+metadata:
+  labels:
+    app: llama3-8b-jetstream-inference-server
+  name: llama3-8b-jetstream-hpa

serving-catalog/core/deployment/jetstream/llama3-8b/base/kustomization.yaml

@@ -11,6 +11,11 @@ patches:
       kind: Deployment
     options:
       allowNameChange: true
+  - path: hpa.patch.yaml
+    target:
+      kind: HorizontalPodAutoscaler
+    options:
+      allowNameChange: true
   - path: job.patch.yaml
     target:
       kind: Job

serving-catalog/core/deployment/vllm/gemma-2b/base/hpa.patch.yaml

@@ -0,0 +1,6 @@
+apiVersion: autoscaling/v2
+kind: HorizontalPodAutoscaler
+metadata:
+  labels:
+    app: gemma-2b-vllm-inference-server
+  name: gemma-2b-vllm-hpa

serving-catalog/core/deployment/vllm/gemma-2b/base/kustomization.yaml

@@ -11,6 +11,11 @@ patches:
       kind: Deployment
     options:
       allowNameChange: true
+  - path: hpa.patch.yaml
+    target:
+      kind: HorizontalPodAutoscaler
+    options:
+      allowNameChange: true
   - path: service.patch.yaml
     target:
       kind: Service

serving-catalog/core/deployment/vllm/llama3-8b/base/hpa.patch.yaml

@@ -0,0 +1,6 @@
+apiVersion: autoscaling/v2
+kind: HorizontalPodAutoscaler
+metadata:
+  labels:
+    app: llama3-8b-vllm-inference-server
+  name: llama3-8b-vllm-hpa

serving-catalog/core/deployment/vllm/llama3-8b/base/kustomization.yaml

@@ -11,6 +11,11 @@ patches:
   path: deployment.patch.yaml
   target:
     kind: Deployment
+- options:
+    allowNameChange: true
+  path: hpa.patch.yaml
+  target:
+    kind: HorizontalPodAutoscaler
 - options:
     allowNameChange: true
   path: service.patch.yaml