---
title: "Installing RealTheory in a Cluster"
slug: "installing-realtheory-in-a-cluster"
updated: 2026-10-30T11:32:32Z
published: 2026-10-30T11:32:32Z
canonical: "resources.realtheory.io/installing-realtheory-in-a-cluster"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://resources.realtheory.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Installing RealTheory in a Cluster

In just a few minutes, you can customize, generate, and apply the RealTheory Collector deployment YAML file, which installs RealTheory into your cluster.

## Prerequisite

Before you install RealTheory into your cluster, consider the following:

- Review the [minimum requirements for the RealTheory Collector installation](https://resources.realtheory.io/docs/realtheory-collector-minimum-requirements)
- You must have a user account inside the organizational account.
- Determine if your environment uses custom or third-party resources, as you must add their API groups to the deployment manifest.
- Establish a labeling protocol if your environment is comprised of multiple clusters.  

See [How to Leverage RealTheory Labels For Scalability and Automation](https://resources.realtheory.io/docs/how-to-leverage-realtheory-labels-for-scalability-and-automation).

## Installing RealTheory

1. Navigate to **Settings** > **Agent** > **Deployment**.
2. Ensure that the minimum requirements for a successful deployment are met:
  - Is your cluster running a supported version of Kubernetes? [I need help with this](https://resources.realtheory.io/docs/how-to-verify-the-cluster-is-running-a-supported-version-of-kubernetes)
  - Does your cluster have the metrics server installed? [I need help with this](https://resources.realtheory.io/docs/how-to-verify-the-metrics-server-is-installed-in-the-cluster)
  - Do you have the appropriate RBAC permissions necessary to install RealTheory in your cluster? [I need help with this](https://resources.realtheory.io/docs/how-to-verify-you-have-the-appropriate-rbac-permissions-to-install-the-realtheory-collector)
  - If your cluster uses a proxy server for outbound traffic, do you have the server's proxy settings? [I need help with this](https://resources.realtheory.io/docs/how-to-verify-you-have-configured-the-proxy-settings-correctly)
3. (Optional) Provide a descriptive **Cluster name** that has meaning based on the cluster's function, purpose, location, etc.
4. (Optional) Provide a **Cluster description**; a comment that describes the cluster.
5. (Optional) To let automation change workloads in this cluster, turn on **Make this cluster automation-ready**. See [Making a Cluster Automation-Ready](/v1/docs/making-a-cluster-automation-ready).
6. (Optional) Provide any of the following labels:

**Note:** The labels you provide will only be associated with objects within the RealTheory application; no labels in the cluster will be altered.
  - **Cloud account identifier** - an identifier for the cloud provider associated with the cluster such as an Account ID, Account Name, Subscription ID, or Project ID
  - **Account owner** - the account owner or contact person for this cluster
  - **Cluster labels** - key/value pairs that specify meaningful and relevant identifying attributes of the cluster  

**Important:** Labels can be leveraged to configure and manage certain RealTheory features such as rate card policies, access control policies, and rule management automatically and at scale. You should review these features to help you identify labels that can simplify feature management.
7. If you want to associate the labels specified in Step 6 with all the Kubernetes objects in your cluster, leave **Associate these labels with all Kubernetes objects within RealTheory** turned on.
8. (Optional) If the cluster is a GKE Autopilot cluster, expand **More options** and turn on **This is a GKE Autopilot cluster**, so that RealTheory calculates costs accurately.
9. If the RealTheory Collector must be a specific architecture (e.g., for security auditing or registry compatibility), expand **More options** and select **arm64** or **amd64**.
10. To ensure that the appropriate CPU and memory requests are set for the RealTheory Collector, expand **More options** and specify the **Approximate cluster vCPU count** for your cluster.

**Note:** The performance of the RealTheory Collector is impacted by the size of the cluster where the Collector is installed. On larger clusters, the CPU and memory requests configured for the `realtheorycollector` pod must be higher to ensure that the Collector has the resources it needs to function efficiently.
11. If a proxy server is required to connect to the Internet, expand **More options**, select **Manual proxy configuration** in the **Outbound proxy** section, and specify the proxy server details.
12. If you need RealTheory to monitor resources outside the standard defaults—such as Custom Resources (CRDs) or additional Networking—expand **More options** and add any API Group names in **Additional API Groups**.
13. Verify the settings you provided in Steps 3 - 12.
14. Click **SHOW YAML**.

The RealTheory Collector deployment YAML will be generated and displayed in the right-hand panel.
15. Review the YAML file and then do one of the following:
  - Download the YAML file to the default download location on your local system.
  - Copy the YAML file to your clipboard and paste it into an application of your choosing.
16. Run the following command to ensure that kubectl is configured to point to the cluster that you want to install the RealTheory collector on:

`kubectl config use-context <cluster-context-name>`

where `<cluster-context-name>` is the name of the cluster context that you want to use.
17. Run the following command on the cluster you want to manage:

`kubectl apply -f <filename.yaml>`

where `<filename.yaml>` is the path to the YAML file you created in Step 15.

Your cluster will be visible in **Visualizations** within a few minutes.

**Note:** If you collect GPU metrics, add the GPU environment variables to the YAML file before you apply it. See [Enabling GPU Metric Collection](https://resources.realtheory.io/docs/enabling-gpu-metric-collection).

**See Also** [Enabling GPU Metric Collection](/v1/docs/enabling-gpu-metric-collection)