Skip to content

README.md

Latest commit

 

History

History
31 lines (29 loc) · 2.18 KB

README.md

File metadata and controls

31 lines (29 loc) · 2.18 KB
Raw

Cluster Autoscaler

Cluster Autoscaler enables dynamic scaling of Kubernetes cluster nodes via AWS API integration with the ASG service

Notes

  • The auto-discovery mode is advised as it is based on the pre-populated tags on the worker ASG (k8s.io/cluster-autoscaler/enabled and k8s.io/cluster-autoscaler/<cluster-name>) where cluster-name is the environment variable set on the eksd-terraform project
  • If you opt to use the manual mode - remember to define the specific workers ASG/s name/s and their lower/upper bounds on the autoscalingGroups values
  • The below chart depends on the cloud-config ConfigMap which is pre-deployed in the kube-system namespace - if you will use another namespace for Cluster Autoscaler you will need to replicate it
  • Always make sure to set min=max=desired capacity for the masters ASG as Cluster Autoscaler may try to scale it down (which will potencially brick the EKS-D cluster)

Installation example

  • Ensure Helm is installed and the kubectl context is correctly pointing to the relevant Kubernetes cluster
  • Assuming you opt for auto-discovery, the below clusterName value will be populated according to the kubectl context (you may change it as needed)
  • Run the below code snippet to create the values file and run the Helm installation: 
    helm repo add autoscaler https://kubernetes.github.io/autoscaler
helm repo update
cat <<EOF | tee values.yaml
awsRegion: us-east-1
autoDiscovery:
  clusterName: $(kubectl config current-context | cut -d '@' -f2)
cloudConfigPath: config/cloud.conf
extraVolumes:
  - name: cloud-config
    configMap:
      name: cloud-config
extraVolumeMounts:
  - name: cloud-config
    mountPath: config
EOF
helm install cluster-autoscaler autoscaler/cluster-autoscaler --namespace=kube-system --values values.yaml