I am running an EKS cluster and I have a HorizontalPodAutoscaler created for autoscaling number of pods based on average CPU utilisation.
How to do the same for Average memory utilization?
Suppose all of the pods running in an EKS clusters, have used average of 70% of memory they are allocated (using resources), then the deployment should be autoscaled.
How to do this? Is creating a custom metric in CloudWatch the only way?
Even if cloudWatch is the only way, how to do that? Is there a specific documentation or tutorial or blog that does this?
Please try the below HPA configuration object.
apiVersion: autoscaling/v2beta1
kind: HorizontalPodAutoscaler
metadata:
name: nginx-hpa
namespace: default
spec:
scaleTargetRef:
apiVersion: extensions/v1beta1
kind: Deployment
name: nginx
minReplicas: 2
maxReplicas: 10
metrics:
- type: Resource
resource:
name: memory
targetAverageUtilization: 70and apply the object using kubectl apply