# Administrator CLI

The NVIDIA Run:ai Administrator (`runai-adm`) is a lightweight tool designed to support infrastructure administrators by simplifying two key tasks:

* [Collecting logs](/self-hosted/2.20/infrastructure-setup/procedures/logs-collection.md) for troubleshooting and sharing with NVIDIA Run:ai support.
* Configuring [node roles](/self-hosted/2.20/infrastructure-setup/advanced-setup/node-roles.md) in the cluster for optimal performance and reliability.

This section outlines the installation and usage of the NVIDIA Run:ai Administrator CLI to help you get started quickly.

## Prerequisites

Before installing the CLI, review the following:

* **Operating system**: The CLI is supported on Mac and Linux.
* **Kubectl**: The Kubernetes command-line interface must be installed and configured to access your cluster. Follow the [official guide](https://kubernetes.io/docs/tasks/tools/install-kubectl/).
* **Cluster administrative permissions**: The CLI requires a Kubernetes profile with administrative privileges.

## Installation

To install the NVIDIA Run:ai Administrator CLI, ensure that the CLI version matches the version of your NVIDIA Run:ai cluster. You can either install the latest version or a specific version from the [list](https://runai.jfrog.io/ui/native/cli/runai-admin-cli/).

### Installing the Latest Version

Use the following commands to download and install the latest version of the CLI:

<details>

<summary><strong>Mac</strong></summary>

```bash
wget --content-disposition https://app.run.ai/v1/k8s/admin-cli/darwin  
chmod +x runai-adm  
sudo mv runai-adm /usr/local/bin/runai-adm
```

</details>

<details>

<summary><strong>Linux</strong></summary>

```bash
wget --content-disposition https://app.run.ai/v1/k8s/admin-cli/linux  
chmod +x runai-adm  
sudo mv runai-adm /usr/local/bin/runai-adm
```

</details>

### Installing a Specific Version

To install a specific version of the Administrator CLI that matches your NVIDIA Run:ai cluster version, append the version number to the download URL. Refer to the list of available versions linked above for the correct version number.

<details>

<summary><strong>Mac</strong></summary>

```bash
wget --content-disposition https://app.run.ai/v1/k8s/admin-cli/<version>/darwin # Replace <version> with the desired version in the format vX.X.X (e.g., v2.19.5) 
chmod +x runai-adm  
sudo mv runai-adm /usr/local/bin/runai-adm
```

</details>

<details>

<summary><strong>Linux</strong></summary>

```bash
wget --content-disposition https://app.run.ai/v1/k8s/admin-cli/<version>/linux # Replace <version> with the desired version in the format vX.X.X (e.g., v2.19.5)
chmod +x runai-adm  
sudo mv runai-adm /usr/local/bin/runai-adm
```

</details>

### Verifying Installation

Verify your installation completed successfully by running the following command:

```sh
runai-adm version  
```

## Reference

### Node Roles

To set or remove node rules using the `runai-adm` tool, run the following:

```sh
runai-adm set node-role [--runai-system-worker | --gpu-worker | --cpu-worker] <node-name>
```

```sh
runai-adm remove node-role [--runai-system-worker | --gpu-worker | --cpu-worker] <node-name>
```

{% hint style="info" %}
**Note**

Use the `--all` flag to set or remove a role to all nodes.
{% endhint %}

### Collect Logs

To collect logs using the `runai-adm` tool:

1. Run the following command:

   ```sh
   runai-adm collect-logs
   ```
2. Locate the generated compressed log file.


---

# 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://run-ai-docs.nvidia.com/self-hosted/2.20/reference/cli/administrator-cli.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.