Skip to content


Step 1: ~/.kubefirst

If you just recently ran install from your localhost, you'll already have the file on your localhost at ~/.kubefirst that's needed to destroy. If you don't have this file locally, you'll need to download it from your s3 bucket that was created during provisioning and add it to your home directory.

Step 2: Destroy

With your ~/.kubefirst file in place, run:

kubefirst cluster destroy

Step 3: Clean (optional)

Cleaning files generated by the installer.

As AWS selection will created S3 buckets, if you don't want to persist this buckets, just execute:

kubefirst clean  --destroy-buckets --destroy-confirm

OR to just remove the files generated locally and preserve the S3 buckets for future re-use, execute:

kubefirst clean  

This step it is not mandatory, but before any new execution of the kubefirstCLI installation for local or cloud selections you will be requested to execute this command, so installer can start from a fresh point.


Avoiding tools re-download

The kubefirstCLI download some tools used during cluster provisioning, for example, Terraform, Helm, and Kubectl, in versions compatible with Kubefirst and stores them in the K1 folder. If you are using Kufibefirst to demo in conferences or using poor connections (mobile, hotels) you should consider using this additional flag --preserve-tools for each cycle of create/destroy. This will preserve tools downloaded and will save time and network bandwidth during cluster provisioning.