Your submission was sent successfully! Close

You have successfully unsubscribed! Close

Thank you for signing up for our newsletter!
In these regular emails you will find the latest updates about Ubuntu and upcoming events where you can meet our team.Close

CNI with Kube-OVN

Kube-OVN is a CNI implementation based on OVN that provides a rich set of networking features for advanced enterprise applications. For a list of features, see the Kube-OVN GitHub repository.

Deploying Charmed Kubernetes with Kube-OVN

To deploy a cluster with Kube-OVN, deploy the charmed-kubernetes bundle with the Kube-OVN overlay:

juju deploy charmed-kubernetes --overlay kube-ovn-overlay.yaml

You can apply any additional customisation overlays that would apply to charmed-kubernetes to this deployment as well.

Kube-OVN configuration options

A full list of Kube-OVN configuration options and their descriptions can be found in the Kube-OVN charm page.

Checking the current configuration

To check the current configuration settings for Kube-OVN, run the command:

juju config kube-ovn

Setting a config option

To set an option, simply run the config command with an additional <key>=<value> argument. For example, to configure the Kube-OVN pinger:

juju config kube-ovn pinger-external-address= pinger-external-dns=example.internal

Config settings which require additional explanation are described below.

Configuring the default subnet

When Kube-OVN first installs, it creates a default subnet named ovn-default that is used to assign IPs to pods. By default, this subnet has CIDR

To configure the default pod subnet with a different CIDR at deploy time, create a file named default-subnet-overlay.yaml that contains the following:


Then include it as an overlay when you deploy Charmed Kubernetes. Note that the order of the overlays is important:

juju deploy charmed-kubernetes --overlay kube-ovn-overlay.yaml --overlay default-subnet-overlay.yaml

Changing the default subnet after deployment

Note that during the process of changing the default subnet, workload pods belonging to that subnet will experience temporary downtime.

To change the default subnet after deployment, first edit the ovn-default subnet to have the desired config, then rebuild all pods under the ovn-default subnet to pick up the new config.

Next, update the Kube-OVN charm configuration to match the new config:

juju config kube-ovn default-cidr= default-gateway=

Configuring the join subnet

Kube-OVN creates a subnet named join that is used to assign IP addresses to the ovn0 interface of Kubernetes nodes, enabling them to participate in the broader Kube-OVN network and communicate directly with pods. By default, the join subnet has CIDR

To configure the join subnet with a different CIDR at deploy time, create a file named join-overlay.yaml that contains the following:


Then include it as an overlay when you deploy Charmed Kubernetes:

juju deploy charmed-kubernetes --overlay kube-ovn-overlay.yaml --overlay join-overlay.yaml

Changing the join subnet after deployment

Note that during the process of changing the join subnet, Kubernetes nodes may be temporarily unable to reach pods.

To change the join subnet after deployment, do the following:

  1. Delete the join subnet
  2. Cleanup allocated config
  3. Update the kube-ovn charm configuration:
    juju config kube-ovn node-switch-cidr= node-switch-gateway=
  4. Wait a minute, then verify that the join subnet has been recreated with the expected configuration:
    kubectl get subnet
  5. Reconfigure the ovn0 NIC address

Configuring kube-ovn-pinger

The kube-ovn-pinger service is a DaemonSet that collects OVS status and a variety of metrics about network connectivity from each Kubernetes node. By default, it is configured to check external network connectivity by pinging and

To change the external addresses used by kube-ovn-pinger, update the pinger-external-address and pinger-external-dns config options:

juju config kube-ovn pinger-external-address= pinger-external-dns=example.internal

Observability configuration

The Kube-OVN charm integrates seamlessly with Prometheus and Grafana from the Canonical Observability Stack. See the COS observability stack documentation for instructions on how to deploy the required observability components: Deploy the COS Lite observability stack on MicroK8s


Kube-OVN exposes metrics for all its components and network quality. See Kube-OVN upstream documentation for more information about the metrics available: Kube-OVN Monitor Metrics

Prometheus and Kube-OVN charm should be related through a cross-model relation; for more information about how to create cross-model relations for the COS observability stack, see the overlays section in the COS documentation: Deploy the COS Lite observability stack on MicroK8s


Kube-OVN charm uses the dashboard from upstream. These use the metrics scraped by Prometheus from the various Kube-OVN components. You can find the available dashboards in the upstream documentation: Kube-OVN Grafana Dashboards

To fetch the dashboards, the Kube-OVN charm must be related (via a cross-model relation) to Grafana from the COS observability stack.

Using a private Docker registry

For a general introduction to using a private Docker registry with Charmed Kubernetes, please refer to the Private Docker Registry page.

In addition to the steps documented there, you will need to upload the following image to the registry:

The Kube-OVN charm will automatically use the image registry that kubernetes-control-plane is configured to use. If needed, you can override the Kube-OVN image registry by setting the image-registry config:

export IP=`juju exec --unit docker-registry/0 'network-get website --ingress-address'`
export PORT=`juju config docker-registry registry-port`
juju config kube-ovn image-registry=$REGISTRY


If there is an issue with connectivity, it can be useful to inspect the Juju logs. To see a complete set of logs for Kube-OVN:

juju debug-log --replay --include=kube-ovn

For additional troubleshooting pointers, please see the dedicated troubleshooting page.

We appreciate your feedback on the documentation. You can edit this page or file a bug here.

See the guide to contributing or discuss these docs in our public Mattermost channel.