Monitoring Linux Servers with Prometheus and Grafana


Monitoring the health and performance of Linux servers is essential for maintaining a stable and reliable IT infrastructure. In this article, we will explore how to set up a powerful monitoring solution using Prometheus and Grafana. Prometheus is an open-source monitoring and alerting toolkit, while Grafana provides feature-rich visualization and dashboarding capabilities.

Section 1: Prometheus Server

  1. Prometheus Overview: Learn about the Prometheus server, its role in monitoring, and the services it collects data from, such as prometheus.service, alertmanager.service, and blackbox_exporter.service.

  2. Prometheus Configuration: Understand the configuration files and directories for Prometheus on the server, located in /etc/prometheus, and how to start the service using systemd.

  3. Accessing Prometheus Console: Access the Prometheus console through HTTP basic authentication using web.yml. Discover how to view all clients and their metrics from the console. (Console URL: http://app108:9090/graph?g0.expr=&

  4. Alerting with Alertmanager: Explore the close relationship between the Prometheus and Alertmanager services. Find out where the Alertmanager config files are located in /etc/alertmanager, and how to manage and browse alerts. (Alertmanager URL: http://app108:9093/#/alerts)

  5. Using Blackbox Exporter: Learn about blackbox_exporter.service, which monitors specific ports or checks alerts. Understand how to configure it through the main Prometheus config file. (Alerts URL: http://app108:9090/alerts)

Section 2: Grafana Server

  1. Grafana Overview: Discover how Grafana complements Prometheus by providing user-friendly dashboards for visualizing Prometheus alerts and server performance statistics. Understand the services running on the Grafana server, including grafana-server.service.

  2. Grafana Configuration: Explore the configuration file grafana.ini located at /etc/grafana, and find out where plugins are deployed in /var/lib/grafana.

  3. Main Dashboard: Explore the primary Grafana dashboard that displays Prometheus alerts and provides links to other relevant pages. (Dashboard URL:

  4. Server Stats Dashboard: Understand the server stats dashboard that showcases data collected by the node exporter services. (Dashboard URL:

  5. Gluster Stats Dashboard: Learn about the Gluster stats dashboard displaying data collected by gluster-exporter.service. (Dashboard URL:

  6. Ping Dashboard: Explore the ping dashboard showing data collected by blackbox_exporter.service. (Dashboard URL:

Section 3: Clients Configuration

  1. Node Exporter Service: Understand how clients use node_exporter.service to monitor the operating system. Access the data exposed on port 9100. (Example URL: http://ap103.vmbs.local:9100/metrics)

  2. Gluster Exporter Service: Discover the configuration and port (9713) of the gluster-exporter.service, used to monitor Gluster server nodes. (Example URL: http://ap103.vmbs.local:9713/metrics)


Setting up Prometheus, Grafana, and Gluster Monitoring together can greatly enhance your server monitoring capabilities. Here’s a step-by-step guide to help you get started:

  1. Install and Configure Prometheus: Begin by installing Prometheus on your designated Prometheus server. You can find detailed installation instructions in the official Prometheus documentation here. Customize the prometheus.yml configuration file to scrape metrics from your desired targets, including Gluster servers and node exporters.

  2. Prometheus Alerting Configuration: Set up alerting rules in the prometheus.yml file to trigger alerts based on specific metrics thresholds. Check the official documentation on alerting rules for details.

  3. Configure Alertmanager: Install and configure Alertmanager to manage alerts and send notifications. The official Alertmanager documentation can guide you through the process of installation and configuration.

  4. Install and Configure Grafana: Install Grafana on your Grafana server and configure it to connect to Prometheus as a data source. You can follow the official Grafana documentation on data sources and installation to get started.

  5. Import Grafana Dashboards: Import pre-built Grafana dashboards to visualize Prometheus metrics effectively. You can find various community-contributed dashboards on Grafana Dashboard Library.

  6. Setup Gluster Monitoring: To monitor Gluster-specific metrics, install and configure the Gluster exporter on Gluster server nodes. You can find the official Gluster exporter documentation on GitHub.

  7. Node Exporter Configuration: Ensure the node_exporter is running on all target servers you want to monitor. Refer to the official Prometheus node_exporter documentation for setup instructions.

  8. Test and Refine: After setting up the monitoring components, perform tests to verify metrics collection and alerting functionality. Adjust alerting rules and dashboards as needed to suit your specific monitoring requirements.

Remember that this is a basic guide to get you started. For more detailed configurations and advanced setups, refer to the official documentation for Prometheus, Grafana, and Gluster Monitoring. The Prometheus official documentation and Grafana documentation provide comprehensive resources to fine-tune your monitoring stack and optimize server performance. Additionally, the Gluster exporter GitHub repository offers valuable insights for monitoring Gluster filesystems.

By following these steps and utilizing the available documentation, you can build a robust monitoring system to keep your Linux servers and the environment in check.


By following this comprehensive guide, you will have successfully set up a powerful monitoring system for your Linux servers using Prometheus and Grafana. With real-time insights into your infrastructure’s health and performance, you can proactively address issues, ensure high availability, and optimize the overall efficiency of your IT environment. Happy monitoring!

Note: The URLs and ports mentioned in this article are for example purposes only and should be replaced with your actual configurations.