# Connect cluster

Follow these steps to connect Kubernetes cluster with Supervisely to run Apps on it.

{% hint style="warning" %}
**BETA** Keep in mind that this is a beta version and some features may not work as expected.
{% endhint %}

### Add a Kubernetes cluster

Go to the `Team Cluster` page, click `Add` button and select `Kubernetes` in the popup window.

![](/files/Pp4JyRs6FUAjzbCzITdT)

You will see the following window:

![](/files/CiUNpAFgs2gAJMGFw2jl)

### Deploy required resources

You can either download the configuration and deploy it manually or use the command provided in the window above to deploy the resources automatically.

### Configure the options

`Server address` - this is the address of your Supervisely server that is accessible from Kubernetes cluster.

`Service account token` - long-term token that will be used to authenticate Kubernetes cluster with Supervisely. You can find the token value using `kubectl -n supervisely get secret sly-task-manager-token -o jsonpath='{.data.token}' | base64 -d`

`API endpoint` - this is the address of your Kubernetes cluster API server that is accessible from Supervisely server. You can find the API address using `kubectl cluster-info`

`Certificate Authority` - this is the CA.crt certificate that is used to verify the identity of the Kubernetes cluster API server. You can find the CA value using `kubectl -n supervisely get secret sly-task-manager-token -o jsonpath='{.data.ca\.crt}'`

`Namespace` - a namespace in which all the resources required to deploy Supervisely Apps will be created.

`Registry secret` - a secret from the `namespace` above that will be used to pull container images from the registry. This is included in the configuration file that you will receive from Supervisely.

`Ingress address` - this is the address used to proxy requests to the Supervisely Apps, essentially this is the address used to access your Ingress.

`Ingress path prefix` - this is the path prefix used for the Ingress rules to avoid conflicts with other services running on the same Kubernetes cluster.

`GPU capability of the cluster` - this setting notifies Supervisely that it can use this cluster for GPU deployments. Make sure that you have installed [NVIDIA Device plugin](https://github.com/NVIDIA/k8s-device-plugin) on your cluster first and confirm the installation by checking nvidia node labels via `kubectl get nodes --show-labels`.

`Ingress manifest for Apps routing` - this is the manifest that will be used to route traffic to the Supervisely Apps. You can use the default NGINX manifest or provide your own. This is a work in progress and we will be adding more manifests in the future.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.supervisely.com/enterprise-edition/kubernetes/agent.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.