HPA for data plane

Author: Nowjean

apis/v1alpha2/nginxproxy_types.go

@@ -388,6 +388,11 @@ type DeploymentSpec struct {
 	// +optional
 	Replicas *int32 `json:"replicas,omitempty"`
 
+	// Horizontal Pod Autoscaling.
+	//
+	// +optional
+	Autoscaling HPASpec `json:"autoscaling"`
+
 	// Pod defines Pod-specific fields.
 	//
 	// +optional
@@ -412,6 +417,41 @@ type DaemonSetSpec struct {
 	Container ContainerSpec `json:"container"`
 }
 
+type HPASpec struct {
+	// Enable or disable Horizontal Pod Autoscaler
+	Enabled bool `json:"enabled"`
+
+	// Annotation for Horizontal Pod Autoscaler
+	// Annotations is an unstructured key value map stored with a resource that may be
+	// set by external tools to store and retrieve arbitrary metadata. They are not
+	// queryable and should be preserved when modifying objects.
+	// More info: https://kubernetes.io/docs/concepts/overview/working-with-objects/annotations
+	// +optional
+	//HPAAnnotations map[string]string `json:"hpaAnnotations,omitempty" protobuf:"bytes,12,rep,name=hpaAnnotations"`
+
+	// behavior configures the scaling behavior of the target
+	// in both Up and Down directions (scaleUp and scaleDown fields respectively).
+	// If not set, the default HPAScalingRules for scale up and scale down are used.
+	// +optional
+	//Behavior *autoscalingv2.HorizontalPodAutoscalerBehavior `json:"behavior,omitempty" protobuf:"bytes,5,opt,name=behavior"`
+
+	// Minimun of replicas.
+	MinReplicas int32 `json:"minReplicas"`
+
+	// Minimun of replicas.
+	MaxReplicas int32 `json:"maxReplicas"`
+
+	// Target cpu utilization percentage of HPA.
+	//
+	// +optional
+	TargetCPUUtilizationPercentage *int32 `json:"targetCPUUtilizationPercentage,omitempty"`
+
+	// Target memory utilization percentage of HPA.
+	//
+	// +optional
+	TargetMemoryUtilizationPercentage *int32 `json:"targetMemoryUtilizationPercentage,omitempty"`
+}
+
 // PodSpec defines Pod-specific fields.
 type PodSpec struct {
 	// TerminationGracePeriodSeconds is the optional duration in seconds the pod needs to terminate gracefully.

apis/v1alpha2/zz_generated.deepcopy.go

@@ -264,9 +264,10 @@ The following table lists the configurable parameters of the NGINX Gateway Fabri
 | `certGenerator.ttlSecondsAfterFinished` | How long to wait after the cert generator job has finished before it is removed by the job controller. | int | `30` |
 | `clusterDomain` | The DNS cluster domain of your Kubernetes cluster. | string | `"cluster.local"` |
 | `gateways` | A list of Gateway objects. View https://gateway-api.sigs.k8s.io/reference/spec/#gateway for full Gateway reference. | list | `[]` |
-| `nginx` | The nginx section contains the configuration for all NGINX data plane deployments installed by the NGINX Gateway Fabric control plane. | object | `{"config":{},"container":{},"debug":false,"image":{"pullPolicy":"Always","repository":"ghcr.io/nginx/nginx-gateway-fabric/nginx","tag":"edge"},"imagePullSecret":"","imagePullSecrets":[],"kind":"deployment","plus":false,"pod":{},"replicas":1,"service":{"externalTrafficPolicy":"Local","loadBalancerClass":"","loadBalancerIP":"","loadBalancerSourceRanges":[],"nodePorts":[],"type":"LoadBalancer"},"usage":{"caSecretName":"","clientSSLSecretName":"","endpoint":"","resolver":"","secretName":"nplus-license","skipVerify":false}}` |
+| `nginx` | The nginx section contains the configuration for all NGINX data plane deployments installed by the NGINX Gateway Fabric control plane. | object | `{"autoscaling":{"enabled":false,"maxReplicas":11,"minReplicas":1,"targetCPUUtilizationPercentage":50,"targetMemoryUtilizationPercentage":50},"config":{},"container":{"resources":{}},"debug":false,"image":{"pullPolicy":"Always","repository":"ghcr.io/nginx/nginx-gateway-fabric/nginx","tag":"edge"},"imagePullSecret":"","imagePullSecrets":[],"kind":"deployment","plus":false,"pod":{},"replicas":1,"service":{"externalTrafficPolicy":"Local","loadBalancerClass":"","loadBalancerIP":"","loadBalancerSourceRanges":[],"nodePorts":[],"type":"LoadBalancer"},"usage":{"caSecretName":"","clientSSLSecretName":"","endpoint":"","resolver":"","secretName":"nplus-license","skipVerify":false}}` |
 | `nginx.config` | The configuration for the data plane that is contained in the NginxProxy resource. This is applied globally to all Gateways managed by this instance of NGINX Gateway Fabric. | object | `{}` |
-| `nginx.container` | The container configuration for the NGINX container. This is applied globally to all Gateways managed by this instance of NGINX Gateway Fabric. | object | `{}` |
+| `nginx.container` | The container configuration for the NGINX container. This is applied globally to all Gateways managed by this instance of NGINX Gateway Fabric. | object | `{"resources":{}}` |
+| `nginx.container.resources` | The resource requirements of the NGINX container. You should be set this value, If you want to use dataplane Autoscaling(HPA). | object | `{}` |
 | `nginx.debug` | Enable debugging for NGINX. Uses the nginx-debug binary. The NGINX error log level should be set to debug in the NginxProxy resource. | bool | `false` |
 | `nginx.image.repository` | The NGINX image to use. | string | `"ghcr.io/nginx/nginx-gateway-fabric/nginx"` |
 | `nginx.imagePullSecret` | The name of the secret containing docker registry credentials. Secret must exist in the same namespace as the helm release. The control plane will copy this secret into any namespace where NGINX is deployed. | string | `""` |

charts/nginx-gateway-fabric/README.md

@@ -8,13 +8,15 @@ rules:
 - apiGroups:
   - ""
   - apps
+  - autoscaling
   resources:
   - secrets
   - configmaps
   - serviceaccounts
   - services
   - deployments
   - daemonsets
+  - horizontalpodautoscalers
   verbs:
   - create
   - update

charts/nginx-gateway-fabric/templates/clusterrole.yaml

@@ -12,7 +12,15 @@ spec:
   kubernetes:
     {{- if eq .Values.nginx.kind "deployment" }}
     deployment:
+      {{- if .Values.nginx.replicas }}
       replicas: {{ .Values.nginx.replicas }}
+      {{- end }}
+      autoscaling:
+        enabled: {{ .Values.nginx.autoscaling.enabled }}
+        minReplicas: {{ .Values.nginx.autoscaling.minReplicas }}
+        maxReplicas: {{ .Values.nginx.autoscaling.maxReplicas }}
+        targetCPUUtilizationPercentage: {{ .Values.nginx.autoscaling.targetCPUUtilizationPercentage }}
+        targetMemoryUtilizationPercentage: {{ .Values.nginx.autoscaling.targetMemoryUtilizationPercentage }}
       {{- if .Values.nginx.pod }}
       pod:
         {{- toYaml .Values.nginx.pod | nindent 8 }}

charts/nginx-gateway-fabric/templates/nginxproxy.yaml

@@ -98,6 +98,45 @@
     "nginx": {
       "description": "The nginx section contains the configuration for all NGINX data plane deployments\ninstalled by the NGINX Gateway Fabric control plane.",
       "properties": {
+        "autoscaling": {
+          "properties": {
+            "enabled": {
+              "default": false,
+              "description": "Enable or disable Horizontal Pod Autoscaler",
+              "required": [],
+              "title": "enabled",
+              "type": "boolean"
+            },
+            "maxReplicas": {
+              "default": 11,
+              "required": [],
+              "title": "maxReplicas",
+              "type": "integer"
+            },
+            "minReplicas": {
+              "default": 1,
+              "description": "hpaAnnotations: {}",
+              "required": [],
+              "title": "minReplicas",
+              "type": "integer"
+            },
+            "targetCPUUtilizationPercentage": {
+              "default": 50,
+              "required": [],
+              "title": "targetCPUUtilizationPercentage",
+              "type": "integer"
+            },
+            "targetMemoryUtilizationPercentage": {
+              "default": 50,
+              "required": [],
+              "title": "targetMemoryUtilizationPercentage",
+              "type": "integer"
+            }
+          },
+          "required": [],
+          "title": "autoscaling",
+          "type": "object"
+        },
         "config": {
           "description": "The configuration for the data plane that is contained in the NginxProxy resource. This is applied globally to all Gateways\nmanaged by this instance of NGINX Gateway Fabric.",
           "properties": {
@@ -313,6 +352,14 @@
         },
         "container": {
           "description": "The container configuration for the NGINX container. This is applied globally to all Gateways managed by this\ninstance of NGINX Gateway Fabric.",
+          "properties": {
+            "resources": {
+              "description": "The resource requirements of the NGINX container. You should be set this value, If you want to use dataplane Autoscaling(HPA).",
+              "required": [],
+              "title": "resources",
+              "type": "object"
+            }
+          },
           "required": [],
           "title": "container",
           "type": "object"

charts/nginx-gateway-fabric/values.schema.json

@@ -228,6 +228,28 @@ nginx:
   # -- The number of replicas of the NGINX Deployment.
   replicas: 1
 
+  autoscaling:
+    # Enable or disable Horizontal Pod Autoscaler
+    enabled: false
+    #hpaAnnotations: {}
+    minReplicas: 1
+    maxReplicas: 11
+    targetCPUUtilizationPercentage: 50
+    targetMemoryUtilizationPercentage: 50
+    # behavior:
+    #   scaleDown:
+    #     stabilizationWindowSeconds: 300
+    #     policies:
+    #     - type: Pods
+    #       value: 1
+    #       periodSeconds: 180
+    #   scaleUp:
+    #     stabilizationWindowSeconds: 300
+    #     policies:
+    #     - type: Pods
+    #       value: 2
+    #       periodSeconds: 60
+
   image:
     # -- The NGINX image to use.
     repository: ghcr.io/nginx/nginx-gateway-fabric/nginx
@@ -431,9 +453,12 @@ nginx:
 
   # -- The container configuration for the NGINX container. This is applied globally to all Gateways managed by this
   # instance of NGINX Gateway Fabric.
-  container: {}
-    # -- The resource requirements of the NGINX container.
-    # resources: {}
+  container:
+    # -- The resource requirements of the NGINX container. You should be set this value, If you want to use dataplane Autoscaling(HPA).
+    resources: {} 
+      # requests:
+      #   cpu: 1000m
+      #   memory: 1Gi
 
     # -- The lifecycle of the NGINX container.
     # lifecycle: {}