Monitoring OMRS

Setting up system monitoring for openmrs

What is System Monitoring?

System monitoring involves tracking and analyzing various aspects of a computer system's performance and health to ensure it operates efficiently and securely. This includes:Resource Utilization,Performance Metrics,Health Checks,Logs and Alerts,Security Monitoring.

OpenMRS System monitoring using  Prometheus and grafana

Prerequisites

1. Docker

2. Docker swarm 

3. Grafana for visualization

4. Prometheus for collecting and storing metric data

Note:All the above  should be installed on the monitoring server

5. An exporter for exposing metrics.This should be installed on a server having openmrs.

Setup Instructions

1. Prepare a monitoring server,this should be separate from the server running openmrs.

2. Setup grafana and Prometheus on your monitoring server.Grafana is used for  visualization and prometheus is used for collecting and storing metrics.

3. For a quick installation of grafana and prometheus,you can clone  https://github.com/I-TECH-UW/digi-monitoring to your monitoring server that already has the grafana and prometheus setup.

1. Startup the services  by running mk.sh on your terminal

2. This starts,the reverse proxy,prometheus and grafana

• Reverse proxy

Acts as an intermediary between clients and backend servers, a reverse proxy can hide the internal network structure from the outside world.

• Prometheus

For  storing time-series data.

• Grafana

Creates dashboards and visualizations for various types of data sources, including time-series data.

3.  Monitor the Target Server ie the server that is running  openmrs.

• To monitor that target server,make sure you have an exporter installed on the target server .

These exporters collect particular metrics from the target systems and deliver them in a Prometheus-readable way.

Examples of exporters: i) Node exporter which enables you to measure various machine resources such as memory, disk and CPU utilization,

ii) cAdvisor for container-based metrics.

For our guide,we shall concentrate on node exporter so as to capture some metrics around the memory,disk and cpu utilization for the server running openmrs.

• Install node exporter on the openmrs server,instructions can be found here ie using docker https://last9.hashnode.dev/how-to-download-and-run-node-exporter-using-docker

Or for a linux machine that is not running docker https://docs.techdox.nz/node-exporter/  

• Expose the url for the node exporter as seen in the node exporter setup instructions

4.  Configure the node exporter with prometheus

• Configure the url of the node exporter exposed in the above step within your prometheus.yml file under the repository you cloned on step c.

For example as seen here:     https://github.com/I-TECH-UW/digi-monitoring/blob/63ce1c38afc43cc2d582a119cf75968697af8c0d/monitoring/prometheus/prometheus.yml#L27

      f)  Restart the prometheus service using docker restart <serviceName>

5. Visualize and Analyze Metrics

• Access your grafana instance and create the monitoring dashboard

• From this point you can see the metrics from your dashboard for the openmrs server you are monitoring

Expected Behavior After Installation

• You should be able to access the grafana ui after installation

• You should be able to view metrics of the target instance on your dashboard after signing in to grafana