aws-cluster-autoscaler

This chart has been deprecated as of version 0.3.1 and will not be updated. Please use the cluster-autoscaler chart instead.

The cluster autoscaler on AWS scales worker nodes within an AWS autoscaling group.

TL;DR:

console

$ helm install stable/aws-cluster-autoscaler -f values.yaml

Where values.yaml contains:

autoscalingGroups:
  - name: your-asg-name
    maxSize: 10
    minSize: 1

Introduction

This chart bootstraps an aws-cluster-autoscaler deployment on a Kubernetes cluster using the Helm package manager.

Prerequisites

Kubernetes 1.3+ with Beta APIs enabled

Installing the Chart

In order for the chart to configure the aws-cluster-autoscaler properly during the installation process, you must provide some minimal configuration which can't rely on defaults. This includes at least one element in the autoscalingGroups array and its three values: name, minSize and maxSize. These parameters cannot be passed to helm using the --set parameter at this time, so you must supply these using a values.yaml file such as:

autoscalingGroups:
  - name: your-asg-name
    maxSize: 10
    minSize: 1

To install the chart with the release name my-release:

console

$ helm install stable/aws-cluster-autoscaler --name my-release -f values.yaml

The command deploys aws-cluster-autoscaler on the Kubernetes cluster using the supplied configuration. The configuration section lists the parameters that can be configured during installation.

Tip: List all releases using helm list

Verifying Installation

The chart will succeed even if the three required parameters are not supplied. To verify the aws-cluster-autoscaler is configured properly find a pod that the deployment created and describe it. It must have a --nodes argument supplied to the ./cluster-autoscaler app under Command. For example (all other values are omitted for brevity):

Containers:
  aws-cluster-autoscaler:
    Command:
      ./cluster-autoscaler
      --cloud-provider=aws
      --nodes=1:10:your-asg-name
      --scale-down-delay=10m
      --skip-nodes-with-local-storage=false
      --skip-nodes-with-system-pods=true
      --v=4

Uninstalling the Chart

To uninstall/delete the my-release deployment:

console

$ helm delete my-release

The command removes all the Kubernetes components associated with the chart and deletes the release.

Configuration

The following table lists the configurable parameters of the aws-cluster-autoscaler chart and their default values.

Parameter	Description	Default
`autoscalingGroups[].name`	autoscaling group name	None. You must supply at least one.
`autoscalingGroups[].maxSize`	maximum autoscaling group size	None. You must supply at least one.
`autoscalingGroups[].minSize`	minimum autoscaling group size	None. You must supply at least one.
`awsRegion`	AWS region	`us-east-1`
`image.repository`	Image	`k8s.gcr.io/cluster-autoscaler`
`image.tag`	Image tag	`v0.5.4`
`image.pullPolicy`	Image pull policy	`IfNotPresent`
`extraArgs`	additional container arguments	`{}`
`nodeSelector`	node labels for pod assignment	`{}`
`podAnnotations`	annotations to add to each pod	`{}`
`replicaCount`	desired number of pods	`1`
`resources`	pod resource requests & limits	`{}`
`scaleDownDelay`	time to wait between scaling operations	`10m` (10 minutes)
`service.annotations`	annotations to add to service	none
`service.clusterIP`	IP address to assign to service	`""`
`service.externalIPs`	service external IP addresses	`[]`
`service.loadBalancerIP`	IP address to assign to load balancer (if supported)	`""`
`service.loadBalancerSourceRanges`	list of IP CIDRs allowed access to load balancer (if supported)	`[]`
`service.servicePort`	service port to expose	`8085`
`service.type`	type of service to create	`ClusterIP`
`skipNodes.withLocalStorage`	don't terminate nodes running pods that use local storage	`false`
`skipNodes.withSystemPods`	don't terminate nodes running pods in the `kube-system` namespace	`true`

Specify each parameter you'd like to override using a YAML file as described above in the [installation](#Installing the Chart) section.

You can also specify any non-array parameter using the --set key=value[,key=value] argument to helm install. For example,

console

$ helm install stable/aws-cluster-autoscaler --name my-release \
    --set awsRegion=us-west-1

IAM Permissions

The worker running the cluster autoscaler will need access to certain resources and actions:

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": [
                "autoscaling:DescribeAutoScalingGroups",
                "autoscaling:DescribeAutoScalingInstances",
                "autoscaling:SetDesiredCapacity",
                "autoscaling:TerminateInstanceInAutoScalingGroup"
            ],
            "Resource": "*"
        }
    ]
}

Unfortunately AWS does not support ARNs for autoscaling groups yet so you must use "*" as the resource. More information here.