Deploying to an existing VPC

kube-aws tries its best to not modify your existing AWS resources. It's our users' responsibility to ensure existing AWS resources provided to kube-aws are properly configured.

Please note that you don't need to care about modifications if you've instructed kube-aws to create all the AWS resources for you i.e. you've omitted vpcId and routeTableId from cluster.yaml.

What kube-aws does modify

• Adding a record set for Kubernetes API Endpoint to an existing hosted zone you've provided via the hostedZoneId configuration key in cluster.yaml
• Adding one or more subnet(s) to an existing VPC specified by the vpcId
• Associating one or more subnet(s) to an existing route table specified by the routeTableId

See cluster.yaml for more details.

All the other configurations for existing AWS resources must be done properly by users before kube-aws is run.

For example, if you're deploying a cluster to an existing VPC:

• An internet gateway or a NAT gateway needs to be added to VPC before cluster can be created
  • Or all the nodes will fail to launch because they can't pull docker images or ACIs required to run essential processes like fleet, hyperkube, etcd, awscli, cfn-signal, cfn-init.
• Existing route tables must have a route to Internet in some form. For example, a default route to an internet gateway or to a NAT gateway via 0.0.0.0/0 would be needed or your cluster won't come up. See a relevant issue about it.
• Existing route tables and/or subnets to be reused by kube-aws must be tagged with the key kubernetes.io/cluster/$CLUSTER_NAME and "shared" as a value.
  • Or Kubernetes will fail to create ELBs correspond to Kubernetes services with type=LoadBalancer
• "DNS Hostnames" must be turned on before cluster can be created
  • Or etcd nodes are unable to communicate each other thus the cluster doesn't work at all