Overview

System log monitoring involves collecting, aggregating, and analyzing the logs produced by your running services. Unlike system metrics monitoring, which tracks CPU or memory usage, log monitoring focuses purely on application and container logs.

This guide describes how to enable log-only monitoring for an OpenMRS server running Docker Compose, using:

• Grafana (visualization & querying)

• Loki (log storage)

• Promtail (log exporter for Docker logs)

No node-level or system-level metrics exporters (e.g., Node Exporter) are required.

Prerequisites On the Monitoring Server

Use a dedicated machine separate from the OpenMRS application server. Ensure the following are installed:

• Docker

• Docker Swarm (optional depending on your stack)

• Loki (for log storage)

• Grafana (for dashboards and log viewing)

On the OpenMRS Docker Server

Install a log exporter that can read Docker logs and send them to Loki:

• Promtail – recommended for Docker log collection

Promtail reads OpenMRS container logs and forwards them to the monitoring server.

Setup Instructions

1. Prepare the Monitoring Server

Set up a dedicated monitoring server that will run:

• Loki (log index + storage)

• Grafana (log visualization)

• A reverse proxy (optional)

2. Install Grafana & Loki

Grafana provides log-querying and visualization, while Loki stores all incoming logs.

For a quick setup, clone the DIGI monitoring repository:

git clone GitHub - DIGI-UW/digi-monitoring

The repository includes configuration for:

• Reverse proxy

• Grafana

• Loki

3. Start the Monitoring Stack

Inside the repository directory, run:

./mk.sh

This script deploys:

• Reverse proxy

• Grafana

• Loki

About the Reverse Proxy

It provides:

• Secure access to Grafana and Loki endpoints

• Hidden internal ports and services

• Central authentication if configured

4. Install a Log Exporter on the OpenMRS Server

To collect logs from the OpenMRS Docker Compose environment, install Promtail.

Recommended: Promtail (for Docker logs)

On the OpenMRS server, create a promtail-config.yamll file:

server:

  http_listen_port: 9080

  grpc_listen_port: 0

positions:

  filename: /var/log/promtail-positions.yaml

clients:

• url: http://<MONITORING_SERVER_IP>:3100/loki/api/v1/push

scrape_configs:

• job_name: "docker-logs"

    static_configs:

• targets:

• localhost

        labels:

          host: "<OPENMRS_SERVER_NAME>"

          __path__: /var/lib/docker/containers/*/*.log

Add Promtail to your OpenMRS Docker Compose deployment:

 promtail:

• image: grafana/promtail:latest

•  container_name: promtail

•  restart: unless-stopped

  volumes:

• /var/lib/docker/containers:/var/lib/docker/containers:ro

• /var/log:/var/log:ro

• ./promtail-config.yaml:/etc/promtail/config.yml

    command: -config.file=/etc/promtail/config.yml

Start Promtail using the command:

Promtail will now forward all container logs (OpenMRS, MySQL, etc.) to the monitoring server.

5. Configure Loki as a Data Source in Grafana

On the monitoring server, access Grafana:

http://<MONITORING_SERVER_IP>:3000

Default login: admin / admin

Steps:

1. Go to Configuration → Data Sources

2. Click Add data source

3. Select Loki

4. Set the URL: http://loki:3100

5. Save & test.

Grafana is now connected to your log store.

6. Validate the Setup

A. Check Promtail Logs

On the OpenMRS server: docker logs promtail

Look for messages indicating successful pushes to Loki.

B. Check Loki Status

Visit:http://<MONITORING_SERVER_IP>:3100/ready

Should respond with ready.

C. Query Logs in Grafana

iOpen: http://<MONITORING_SERVER_IP>:3000/explore

Choose Loki as the data source.

Run a basic query:

{job="docker-logs"}

You should now see logs from all OpenMRS-related containers.

Conclusion

After completing the setup, your monitoring environment will:

• Collect Docker logs from the OpenMRS server

• Store logs efficiently using Loki

• Enable real-time log viewing and search in Grafana

• Support log filtering by service, container, or log level

• Provide historical logs for debugging and analysis

This log-only monitoring setup ensures that you have clear visibility into container behavior without collecting system metrics.