techjw / k8s-aws Goto Github PK

The focus of this code is to deploy a basic Kubernetes cluster into AWS for testing and training purposes. Upon successful completion, the following instances will have been provisioned:

• 1 k8s master node, with etcd co-located
• 2 k8s worker node
• 1 k8s ingress controller node

All instances are configured with public IPs, however firewall rules will only allow SSH and kubeapi access from the CIDR block (local_cidr) defined in terraform/terraform.tfvars, or the default value as set in variables.tf if no local_cidr is provided.

Additionally, the following support infrastructure will be provisioned:

• Custom VPC with a single subnet, with associated Internet gateway, and routing table
• Newly generated SSH keypair (local and setup as AWS key pair)
• Security groups for allowing SSH, kubeapi, and internal network traffic

• Amazon Web Service account (https://aws.amazon.com/free/)
• AWS CLI (https://aws.amazon.com/cli/)
• Terraform v0.11+ (https://www.terraform.io/downloads.html)

Create a customized terraform/terraform.tfvars file (tfvars file reference), key configurations are:

Other settings can be configured such as instances sizes and region, as listed in terraform/variables.tf.

1. Deploy the AWS infrastructure:

make prepare-instances
make create-instances

1. Once the deployment completes, setup and execute Kismatic Enterprise Toolkit:

make install-kismatic
make prepare-kubernetes
make install-kubernetes

• Note: you may change the deployment name in the Makefile, or by passing DEPLOY=myname to the prepare and install commands.

1. When finished with the cluster, you may destroy all the resources that were provisioned. To do so, run the following:

make destroy-instances
make remove-kismatic

• Note: If you changed the deployment name via command line, pass the name to the destroy command as well.