Install Microk8s

To install, please run the following command:

sudo snap install microk8s --classic

sudo snap alias microk8s.kubectl kubectl


microk8s.enable \
  dashboard \
  dns \
  gpu \
  ingress \
  metrics-server \
  prometheus \
  registry \

microk8s.kubectl config view --raw > ~/.kube/config

Polyaxon configuration

If you are here, we assume that you have a Microk8s cluster and Helm setup.

Please consider reading our configuration setup guides to have a deeper knowledge about how to configure and customize Polyaxon to fit your needs.

Create a config file config.yaml or polyaxon_config.yaml, and set up all information you want to override in the default config.


If you are not using Polyaxon with RBAC enabled you should disable it in your config.yaml file

  enabled: false

Use a NodePort service

serviceType: NodePort

Create a namespace for Polyaxon

Polyaxon installs and uses a namespace to run experiments independently of other applications running on your cluster, we recommend using polyaxon.

$ kubectl create namespace polyaxon

namespace "polyaxon" created

If you would like to use a different value, you must keep in mind to update the namespace value in your config.

Install Polyaxon

First of all, you need to add the Polyaxon helm repository to your helm, so you can install Polyaxon from it. This makes it easy to refer to Polyaxon chart without having to use a long URL each time.

helm repo add polyaxon
helm repo update


You can validate that your deployment config.yml file is compatible with the version you are trying to deploy:

polyaxon admin deploy -f config.yml --check


Now you can install Polyaxon with your config.yml file

You can use Polyaxon CLI to manage the deployment

polyaxon admin deploy -f config.yml

Or you can use Helm to do the same:

helm install polyaxon/polyaxon \
    --namespace=<NAMESPACE> \
    -f config.yml

<RELEASE_NAME> is an identifier used by helm to refer to this deployment. You need it when you are changing the configuration of this install or deleting it. We recommend using RELEASE_NAME = polyaxon or RELEASE_NAME = plx.

--namespace should be the same value of the namespace you created in the first step, we again recommend using polyaxon to make it always easy to remember.

Tip: We recommend using polyaxon for both the --name and --namespace to avoid too much confusion. The same command with polyaxon as a value:

helm install polyaxon/polyaxon \
--name=polyaxon \
--namespace=polyaxon \
-f config.yml

Note: “Release name already exists error” If you get a release named <RELEASE_NAME> already exists error, then you should delete the release by running helm delete --purge <RELEASE_NAME>.

You can see the pods being created by entering in a different terminal:

kubectl --namespace=<NAMESPACE> get pod

When helm is done deploying Polyaxon, it will output some instructions NOTES, these note will be different depending on your configuration (the service type used and/or ingress);

Env var for setting the CLI

set -xg POLYAXON_PORT (kubectl get --namespace polyaxon -o jsonpath="{.spec.ports[0].nodePort}" services polyaxon-polyaxon-api)
set -xg POLYAXON_IP (kubectl get nodes --namespace polyaxon -o jsonpath="{.items[0].status.addresses[1].address}")

env | grep POLYAXON


polyaxon config set --host=$POLYAXON_IP --port=$POLYAXON_PORT

Upgrade Polyaxon

To upgrade Polyaxon to a newer version, you can simply run the following command using Polyaxon CLI:

polyaxon deploy -f config.yml --upgrade

Or using Helm

helm upgrade polyaxon polyaxon/polyaxon -f config.yml

Applying configuration changes

The general method to modify your Kubernetes deployment is to:

  1. Make a change to the config.yml

  2. [Optional] run polyaxon deploy -f config.yml --check

  3. Run a polyaxon deploy -f config.yml --upgrade or helm upgrade:

    $ helm upgrade <RELEASE_NAME> polyaxon/polyaxon -f config.yml

    Where <RELEASE_NAME> is the parameter you passed to --name when installing Polyaxon with helm install.

    If you don’t remember it, you can probably find it by doing helm list.

    If you used the default values suggested in the docs, the <RELEASE_NAME> should be polyaxon

  4. Wait for the upgrade to finish, and make sure that when you do

kubectl --namespace=<NAMESPACE> get pod the pods are in Ready state.

Your configuration change has been applied!


DNS issues

If the dns isn’t reflected from the host to the pod. Please run:

sudo iptables -P FORWARD ACCEPT

This could solved the issue.

Build failure

If your builds using the native dockerizer are not succeeding, please use kaniko as build process.

Turn off Polyaxon

When you are done with Polyaxon, you can turn off the deployment, and depending on your persistence configuration you can keep all your data saved for future deployments.

You can also decide to completely turn off Polyaxon and remove the namespace and computational resources.

polyaxon admin teardown


helm del --purge polyaxon

Stop/Delete running experiments/jobs

Polyaxon will by default stop all running jobs/experiments before a teardown, unless you prefer not to trigger the pre-delete hooks, in that case you should clean them on your own.

Delete Helm release

Delete the helm release. This deletes all resources that were created by helm during the deployment.

$ helm delete <RELEASE_NAME> --purge

If you used the default values, the command should be,

$ helm delete polyaxon --purge

If for some reason, your deployment did not succeed, you might need to delete Polyaxon with this command instead, to avoid triggering pre-delete hooks

$ helm delete polyaxon --purge --no-hooks

Delete the namespace

Delete the namespace Polyaxon was installed in. This deletes any disks that may have been created to store user’s logs|database, and any IP addresses that may have been provisioned.

$ kubectl delete namespace <your-namespace>

If you used the default values, the command should be,

$ kubectl delete namespace polyaxon