Install Grafana Agent on a Linux host using Ansible

This guide will show you how to install the Grafana Agent on a Linux host using Ansible and to use it to push logs to Grafana Cloud.

Before you begin

Before you begin, you should have the following available:

A Grafana Cloud account.
A Linux machine
Command line (terminal) access to that Linux machine with unzip binary installed
Account permissions sufficient to install and use the Grafana Agent on the Linux machine
Ansible installed on the Linux machine

Create a Grafana Agent configuration file

For this guide, we will configure the agent to collect and send logs from the Linux machine. The Agent configuration should look like this:

logs:
  configs:
    - clients:
      - basic_auth:
          password: <Your Grafana.com API Key>
          username: <User>
        url: https://logs-prod3.grafana.net/loki/api/v1/push
      name: default
      positions:
        filename: /tmp/positions.yaml
      scrape_configs:
        - job_name: integrations/node_exporter_direct_scrape
          static_configs:
            - targets:
                - localhost
              labels:
                instance: hostname
                __path__: /var/log/*.log
                job: integrations/node_exporter
      target_config:
        sync_period: 10s

Install Grafana Agent binary using Ansible

This Ansible playbook installs the Grafana Agent binary version v0.32.1 and also creates a systemd service to manage the Grafana agent. It creates a new user named grafana-agent on the Linux machine for running the Grafana Agent.

Create a file named grafana-agent.yml and add the following:

- name: Install Grafana Agent on a linux host
  connection: local
  hosts: localhost

  vars:
    agent_binary_location: <agent-binary-location>      # Example /usr/local/bin
    agent_config_location: <agent-config-location>      # Example /etc/grafana-cloud
    linux_architecture: <linux-architecture>            # Example linux-amd64
    agent_config_local_path: <agent-config-local-path>  # Example linux-agent-config.yml
    agent_version: 0.32.1

  tasks:
    - name: Download Grafana Agent binary
      get_url:
        url: "https://github.com/grafana/agent/releases/download/v{{ agent_version }}/grafana-agent-{{ linux_architecture }}.zip"
        dest: "/tmp/agent-linux.zip"
        mode: '0644'

    - name: Unarchive Grafana Agent binary
      unarchive:
        src: "/tmp/agent-linux.zip"
        dest: "{{ agent_binary_location }}"
        remote_src: yes
        mode: '0755'

    - name: Create directory for Grafana Agent
      file:
        path: "{{ agent_config_location }}"
        state: directory
        mode: '0755'

    - name: Create config file for Grafana Agent
      copy:
        src: "{{ agent_config_local_path }}"
        dest: "{{ agent_config_location }}/agent-config.yaml"

    - name: Add user 'grafana-agent'
      user:
        name: grafana-agent
        create_home: no
        shell: /bin/false

    - name: Create service file for Grafana Agent
      copy:
        dest: "/etc/systemd/system/grafana-agent.service"
        content: |
          [Unit]
          Description=Grafana Agent

          [Service]
          User=grafana-agent
          ExecStart={{ agent_binary_location }}/grafana-agent-{{ linux_architecture }} --config.file={{ agent_config_location }}/agent-config.yaml
          Restart=always

          [Install]
          WantedBy=multi-user.target

    - name: Start Grafana Agent service
      systemd:
        daemon_reload: yes
        name: grafana-agent
        enabled: yes
        state: restarted

Replace the following field values:
- <agent-binary-location> with the path where the Grafana Agent binary will be stored on the Linux machine.
- <agent-config-location> with the path where the Grafana Agent configuration will be stored on the Linux machine.
- <linux-architecture> with the architecture of your Linux machine.
- <agent-config-local-path> with the local path to the Grafana Agent configuration.

Run the Ansible playbook on the Linux machine

In the Linux machine’s terminal, run the following command from the directory where the Ansible playbook is located.

ansible-playbook grafana-agent.yml

Validate

The Grafana Agent service on the Linux machine should be active and running. You should see a similar output:

$ sudo systemctl status grafana-agent.service
  grafana-agent.service - Grafana Agent
    Loaded: loaded (/etc/systemd/system/grafana-agent.service; enabled; vendor preset: enabled)
    Active: active (running) since Wed 2022-07-20 09:56:15 UTC; 36s ago
  Main PID: 3176 (agent-linux-amd)
    Tasks: 8 (limit: 515)
    Memory: 92.5M
      CPU: 380ms
    CGroup: /system.slice/grafana-agent.service
      └─3176 /usr/local/bin/agent-linux-amd64 --config.file=/etc/grafana-cloud/agent-config.yaml

In a Grafana Cloud stack, click Explore in the left-side menu.
At the top of the page, use the dropdown menu to select your Loki logs data source. In the Log Browser, run the query {job="integrations/node_exporter"}

Summary

In this guide, you installed the Grafana Agent on a Linux node using Ansible and use it to pushed logs to Grafana Cloud.

To learn more about managing Grafana using Infrastructure as Code, see Provisioning Grafana Cloud with infrastructure as code.