Install a Kloudfuse POC

A Kloudfuse POC (proof of concept, a trial) runs on a small Kubernetes cluster in your cloud account: AWS, Google Cloud, or Azure. One script, install.sh, installs it in one of two ways:

Install What you provide What install.sh does

New cluster (Terraform)

A cloud account you can sign in to

Terraform creates the network, the cluster and its nodes, a storage bucket, and a static IP, then Kloudfuse is installed on it.

Existing cluster (Helm)

A cluster with a node for Kloudfuse, a storage bucket, and a static IP

Installs Kloudfuse on it with Helm. No Terraform.

For cloud requirements, settings, and upgrades, see POC Overview. For errors, see POC Troubleshooting.

Prerequisites

  • license.json, the license file Kloudfuse gave you.

  • A DNS name for Kloudfuse, such as kfuse.example.com, and access to its DNS. The install asks you to add an A record for it.

  • For an existing cluster: a kubeconfig context with cluster-admin, and a node labelled and tainted kf_role=kfuse with at least 16 vCPU and 128 GiB. Existing cluster requirements lists everything.

  • The required tools. install.sh checks for them first.

Signed in to your cloud

Cloud Sign in with

AWS

aws sso login (or aws configure for access keys)

Google Cloud

gcloud auth login, then gcloud auth application-default login

Azure

az login, then az account set --subscription <subscription ID>

For an existing cluster, you can get its kubeconfig context by running…​

  • aws eks update-kubeconfig --name <cluster>,

  • gcloud container clusters get-credentials <cluster>, or

  • az aks get-credentials --name <cluster> --resource-group <group>.

Installation

Create an empty folder and copy the license.json file to it. Open a terminal there and run:

curl -fsSLO https://install.kloudfuse.io/kfuse-poc/install.sh && bash install.sh

Then:

  1. Choose install type: 1 for a new cluster, 2 for a cluster you already have.

  2. New cluster: choose the cloud (aws, gcp, or azure). Existing cluster: confirm your kubeconfig context with Enter, or type another one; the cloud is read from the cluster.

  3. install.sh downloads the matching bundle into kfuse-poc-<version>/kfuse-poc-<bundle>/ and carries on from there.

  4. Answer the questions. Press Enter to take the value in [brackets]. You choose the password of the admin login here: remember it, as the cluster keeps only its hash.

  5. New cluster: review Terraform’s plans and type yes.

  6. Add the DNS record it prints (<domain> A <static IP>), then press Enter.

  7. It installs Kloudfuse and waits until it answers. Then open https://<domain>; and log in as admin.

On a new cluster, the whole install, including creating the cloud resources, takes about 30 minutes. On an existing cluster, it takes about 10–15 minutes.

Post-install

Keep everything in the bundle folder. Run follow-up actions on the install from there:

Action Run, in the bundle folder

Carry on after a stop, or reinstall

./install.sh (every step is safe to repeat)

Change a setting, or the admin password

./install.sh, and answer No to Reinstall with these settings?

Add Helm values the install doesn’t set

Put them in extended-values.yaml, then ./install.sh

Remove Kloudfuse

./uninstall.sh

./uninstall.sh removes only the items that the install created.

The folder you started in also gets a copy of uninstall.sh, so ./uninstall.sh works from there too: it finds the install in the bundle folder and removes it.

Required Tools

install.sh checks them first and shows a table of what it found.

Tool Version

kubectl

Matching the cluster

Helm

3.8 or later, or 4.x

yq (brew install yq)

4.x

curl

any

unzip

any

htpasswd

any

dig or nslookup

any

openssl (only with your own TLS certificate)

any

Terraform (new cluster only)

1.5 or later (1.x)

AWS CLI (new cluster on AWS)

2.x

gcloud and gke-gcloud-auth-plugin (new cluster on Google Cloud)

current

Azure CLI (new cluster on Azure)

current

On an existing cluster, the cloud’s CLI is needed only if your kubeconfig signs in with it.