User Guide

Welcome to the Flux Operator user guide! If you come here, we are assuming you have a cluster with the Flux Operator installed and are interested to submit your own custom resource to create a MiniCluster, or that someone has already done it for you. If you are a developer wanting to work on new functionality or features for the Flux Operator, see our Developer Guides instead.

Containers Available

All containers are provided under The latest tag is the current main branch, a “bleeding edge” version, and we also provide releases, each of which has YAML for x86 or ARM associated with a release container. For dates before June 30, 2023, we provide the other pinned containers in case you want a previous version:

These were primarily experimental versions run for experiments like Kubecon!


Quick Install

We generally recommend that you install a release, e.g.,


# For x86
kubectl apply -f${VERSION}/flux-operator.yaml

# For ARM
kubectl apply -f${VERSION}/flux-operator-arm.yaml

You can also install from the current main branch “bleeding edge” latest:

kubectl apply -f
kubectl apply -f

Note that from the repository, these configs are generated with:

$ make build-config
$ make build-config-arm

and then saved to the main branch or release where you retrieve it from.

Helm Install

We optionally provide an install with helm, which you can do either from the charts in the repository:

$ git clone 
$ cd flux-operator
$ helm install ./chart

Or directly from GitHub packages (an OCI registry):

# helm prior to v3.8.0
$ helm pull oci://

And install!

$ helm install chart-0.1.0.tgz 
NAME: flux-operator
LAST DEPLOYED: Fri Mar 24 18:36:18 2023
NAMESPACE: default
STATUS: deployed

Development Install

For developer instructions, please see our developer documentation.

Tool Install

We maintain a tool Flux Cloud that is able to bring up clusters, install the operator, and optionally run experiments and bring them down. We currently support a handful of clouds (AWS and Google) and if you find yourself wanting a way to easily generate and save results for experiments, this might be the way to go. If you have a cloud or environment you want to deploy to that isn’t supported, please let us know.

Next Steps

For next steps, you might do the following:

1. Verify Install

Regardless of what you chose above, from this point on (after the operator is installed) there are some checks that you can do to see that everything worked. First, ensure the operator-system namespace was created:

$ kubectl get namespace
NAME              STATUS   AGE
default           Active   12m
kube-node-lease   Active   12m
kube-public       Active   12m
kube-system       Active   12m
operator-system   Active   11s
$ kubectl describe namespace operator-system
Name:         operator-system
Labels:       control-plane=controller-manager
Annotations:  <none>
Status:       Active

No resource quota.

No LimitRange resource.

You can find the name of the operator pod as follows:

$ kubectl get pod --all-namespaces -o wide
operator-system   operator-controller-manager-6c699b7b94-bbp5q   2/2     Running   0             80s    ip-192-168-28-166.ec2.internal   <none>           <none>

2. Create Namespace

You’ll likely Make your namespace for the flux-operator:

$ kubectl create namespace flux-operator

3. Validate your container (optional)

Your main container (with flux installed) has a basic set of requirements and we provide a simple tool to sanity check the most simple of these requirements, the Flux Operator Validator. You are encouragd to run this script, although it’s not required- you can just as easily go through the list and verify the points on your own.

4. Apply your custom resource definition

Ensure that your custom resource definition matches the namespace you just created. Then apply your CRD. You can use an example Here is using a default we provide:

$ kubectl apply -f

Please let us know if you would like an example type added - we have plans for many more but are prioritizing them as we see them needed. And now you can get logs for the manager:

$ kubectl logs -n operator-system operator-controller-manager-6c699b7b94-bbp5q

And then watch your jobs!

$ kubectl get -n flux-operator pods

And don’t forget to clean up! Leaving on resources by accident is expensive! This command will vary depending on the cloud you are using. Either way, it’s good to check the web console too to ensure you didn’t miss anything. Next, you might be interested in ways to submit jobs or how to build images in our Developer Guides.

Last update: Sep 10, 2023