Prometheus and Grafana Tutorial for Monitoring Alluxio: Introduction, Environment Setup, and Manual Tuning

1. Introduction to Prometheus & Grafana

Prometheus is an open‑source monitoring solution that replaces traditional monitoring and alerting models with a centralized rule‑based system. Grafana, written in Go, is a popular open‑source visualization tool for time‑series data and supports many back‑end databases.

2. Environment Setup

To monitor Alluxio effectively, a robust monitoring stack is required. The steps are:

Pre‑condition: Alluxio must be installed and running together with Presto.

Configure Prometheus Server: download the tarball, extract it, and edit prometheus.yml (example snippet below).

Start Prometheus Server:

nohup <PATH_TO_PROMETHEUS_SERVER>/prometheus --web.enable-admin-api --config.file=<PATH_TO_PROMETHEUS_SERVER>/prometheus.yml ><PATH_TO_PROMETHEUS_SERVER>/prometheus.log 2>&1 &

Verify the UI at http://<PROMETHEUS_HOSTNAME>:9090/ and ensure all targets are up .

Configure Grafana: download, extract, and start it in the background.

nohup <PATH_TO_GRAFANA>/bin/grafana-server --homepath <PATH_TO_GRAFANA> web > <PATH_TO_GRAFANA>/grafana.log 2>&1 &

Access Grafana UI at http://<GRAFANA_HOSTNAME>:3000 (default admin/admin).

Add a Prometheus data source in Grafana (Configuration → Data sources → Add data source → Prometheus) and set the HTTP URL to the Prometheus server (default port 9090).

Configure Alluxio monitoring in prometheus.yml by adding the following scrape jobs (replace hostnames and ports as needed):

scrape_configs:
# Add Alluxio configurations
- job_name: "alluxio master"
metrics_path: '/metrics/prometheus/'
static_configs:
- targets: ['<LEADING_MASTER_HOSTNAME>:<MASTER_WEB_PORT>'] # default 19999
- job_name: "alluxio worker"
metrics_path: '/metrics/prometheus/'
static_configs:
- targets: ['<WORKER_HOSTNAME>:<WORKER_WEB_PORT>'] # default 30000
... (additional jobs for job master, job worker, fuse)

Import the Alluxio monitoring dashboard template into Grafana and adjust the data source if necessary.

3. Manual Tuning

Modify dashboards by adding panels or rows, selecting metrics (e.g., CPU), choosing instances or jobs, and using the Use query button. Adjust panel titles, descriptions, axis names, units, and visual styles (line, bar, point). Export data via Download CSV and save dashboards as JSON templates.

Typical steps include:

Click Add panel → configure query.

Customize visualization type and axis settings.

Inspect data and download CSV for further analysis.

4. Cleanup

Stop the monitoring services when finished:

pgrep -f prometheus | xargs kill

pgrep -f pushgateway | xargs kill

pgrep -f grafana | xargs kill

Conclusion

By integrating Prometheus and Grafana, a flexible and user‑friendly monitoring system for Alluxio is built, offering comprehensive metrics, customizable dashboards, and easy data export for different user needs.