k8saas-cli

A CLI for interacting with k8saas-api

Installing

Prebuilt binaries are available: https://github.com/confluentinc/cire-k8saas-cli/releases/.

Building

You'll need Go 1.15:

make build

Authentication

Calling the k8saas API requires you to be authenticated. The CLI will automatically pick up your Vault token from $HOME/.vault-token, so there is nothing for you to do aside from successfully completing cclogin against any environment.

Commands

The result of any subcommand is either an error message, or a cluster spec dumped as JSON so you can pipe it into other commands, like jq.

At the root the CLI takes one flag: api-env. This defaults to the production deployment of k8saas and should be where any clusters that used to be in cc-terraform get created. You can switch to stag or devel if you want to target a different k8saas environment.

create

This subcommand lets you create new clusters. It will tell you about any missing required flags. All flags are automatically generated and applied based on k8saas-structs.

The flag name matches the name= pair in the protobuf struct tag, with underscores replaced by dashes. Flags for values in AksConfig, EksConfig and GkeConfig are prefixed with aks., eks. or gke..

Required flags

If you want to know which flags are required, see the comments on the protobuf definition.

get

This subcommand lets you retrieve a single cluster, by its cluster ID.

delete

This subcommand lets you delete a single cluster, by its cluster ID.

Example

# Ensure you have a fresh vault token by simply running caaslogin
# This also verifies that you have the necessary permissions to perform the following statements
cclogin -e devel -c aws -r us-west-2 -b 0

./k8saas create \
    -name "cli-example" \
    -region eu-west-1 \
    -env devel \
    -cloud aws \
    -image-registry "037803949979.dkr.ecr.eu-west-1.amazonaws.com/confluentinc" \ # This varies by region. k8saas will grant access to the registry.
    -instance-type "r5.xlarge" \
    -eks.master-security-groups "sg-0fd1b86f58624d7bf" \
    -eks.subnet-ids "subnet-32c9c46a,subnet-5ddb1a3a,subnet-e16ea3a8" \
    -eks.vpc-id "vpc-949f5ef3" \
    -eks.worker-security-groups "sg-0cd9d39259c6c121f" 

The response will include an ID which can be used to retrieve the cluster:

k8saas get k8s-c6jh7

Note: The status of the cluster can be found in cluster.state. It's an integer which maps to a state constant in the proto definition.

After the cluster reached the state 4/active you can use it from the region specific bastion. Don't forget to run configure_kubecfg. The kubeconfig context will be: "k8saas-${ID}"

To delete your cluster:

k8saas delete k8s-c6jh7