Add keda example

Author: pdabelf5

examples/shared-examples/autoscaling-keda/README.md

@@ -0,0 +1,28 @@
+# Enable Keda scale to 0 in off hours
+
+This example is based on the official Keda [example](https://keda.sh/docs/2.18/scalers/cron/#scale-to-0-during-off-hours).
+
+## 1. Deploy the Ingress Controller
+
+1. Follow the [installation](https://docs.nginx.com/nginx-ingress-controller/installation/installation-with-manifests/)
+   instructions to deploy the Ingress Controller.
+
+## 2. Deploy Keda
+
+1. Follow the [installation](https://keda.sh/docs/2.18/deploy/) instructions to suit your deployment preference.
+
+## 3. Apply the Keda ScaledObject
+
+Adjust the setting to suit your requirements.  See the [ScaledObject](https://keda.sh/docs/2.18/reference/scaledobject-spec/) reference for more details
+
+1. Apply `scaled-object.yaml`
+
+```console
+kubectl apply -f scaled-object.yaml
+```
+
+## 4. Configure deployment replicas
+
+1. Remove the `spec.replicas` from your deployment to allow Keda to autoscale your pods.
+
+2. Once the end time has expired, 5 minutes from then the NGINX Ingress Controller pods should scale to 0.

examples/shared-examples/autoscaling-keda/scaled-object.yaml

@@ -0,0 +1,22 @@
+apiVersion: keda.sh/v1alpha1
+kind: ScaledObject
+metadata:
+  name: cron-scaledobject
+  namespace: nginx-ingress
+spec:
+  scaleTargetRef:
+    name: nginx-ingress
+  minReplicaCount: 0
+  maxReplicaCount: 4
+  cooldownPeriod: 300
+  triggers:
+    - type: cron
+      metadata:
+        timezone: Europe/London
+        start: 0 09 * * *
+        end: 30 17 * * *
+        desiredReplicas: "1"
+    - type: cpu
+      metricType: Utilization
+      metadata:
+        value: "80"