Node Pool allows you to bring up additional pools of worker nodes each with a separate configuration including:
kube-aws creates a node pool in a single AZ by default. On top of that, you can add one or more node pool in an another AZ to achieve Multi-AZ.
Assuming you already have a subnet and a node pool in the subnet:
subnets: - name: managedPublicSubnetIn1a availabilityZone: us-west-1a instanceCIDR: 10.0.0.0/24 worker: nodePools: - name: pool1 subnets: - name: managedPublicSubnetIn1a
cluster.yaml file to add the second node pool:
subnets: - name: managedPublicSubnetIn1a availabilityZone: us-west-1a instanceCIDR: 10.0.0.0/24 - name: managedPublicSubnetIn1c availabilityZone: us-west-1c instanceCIDR: 10.0.1.0/24 worker: nodePools: - name: pool1 subnets: - name: managedPublicSubnetIn1a - name: pool2 subnets: - name: managedPublicSubnetIn1c
Launch the secondary node pool by running
$ kube-aws update \ --s3-uri s3://<my-bucket>/<optional-prefix>
Beware that you have to associate only 1 AZ to a node pool or cluster-autoscaler may end up failing to reliably add nodes on demand due to the fact that what cluster-autoscaler does is to increase/decrease the desired capacity hence it has no way to selectively add node(s) in a desired AZ.
Also note that deployment of cluster-autoscaler is currently out of scope of this documentation. Please read cluster-autoscaler's documentation for instructions on it.
If you've chosen to power your worker nodes in a node pool with an auto scaling group, you can customize
Please read the AWS documentation for more information on
MinInstancesInService for ASGs.
worker: nodePools: - name: pool1 autoScalingGroup: minSize: 1 maxSize: 3 rollingUpdateMinInstancesInService: 2
See the detailed comments in
cluster.yaml for further information.
Utilizing Spot Fleet gives us chances to dramatically reduce cost being spent on EC2 instances powering Kubernetes worker nodes while achieving reasonable availability. AWS says cost reduction is up to 90% but the cost would slightly vary among instance types and other users' bids.
Spot Fleet support may change in backward-incompatible ways as it is still an experimental feature. So, please use this feature at your own risk. However, we'd greatly appreciate your feedbacks because they do accelerate improvements in this area!
kube-aws updateto increase or decrease
targetCapacityof a spot fleet results in a complete replacement of the Spot Fleet hence some downtime. This is due to how CloudFormation works for updating a Spot Fleet
This feature assumes you already have the IAM role with ARN like "arn:aws:iam::youraccountid:role/aws-ec2-spot-fleet-role" in your own AWS account. It implies that you've arrived "Spot Requests" in EC2 Dashboard in the AWS console at least once. See the AWS documentation describing pre-requisites for Spot Fleet for details.
To add a node pool powered by Spot Fleet, edit node pool's
worker: nodePools: - name: pool1 spotFleet: targetCapacity: 3
To customize your launch specifications to diversify your pool among instance types other than the defaults, edit
worker: nodePools: - name: pool1 spotFleet: targetCapacity: 5 launchSpecifications: - weightedCapacity: 1 instanceType: t2.medium - weightedCapacity: 2 instanceType: m3.large - weightedCapacity: 2 instanceType: m4.large
This configuration would normally result in Spot Fleet to bring up 3 instances to meet your target capacity:
This is achieved by the
diversified strategy of Spot Fleet.
Please read the AWS documentation describing Spot Fleet Allocation Strategy for more details.
Please also see the detailed comments in
cluster.yaml and the GitHub issue summarizing the initial implementation of this feature for further information.