MongoDB Monitoring with Prometheus and Grafana

Monitoring MongoDB performance and health is crucial for maintaining database reliability and efficiency. One of the best ways to achieve this is by using mongodb_exporter, which allows Prometheus to collect MongoDB metrics and visualize them with Grafana.

This guide will cover setting up mongodb_exporter using Docker and Systemd, integrating it with Prometheus, and visualizing data in Grafana.

Prerequisites

Before getting started, ensure you have:

• A running MongoDB instance
• Docker installed (if using Docker setup)
• Prometheus and Grafana installed

User Setup for MongoDB Exporter

To ensure MongoDB Exporter has the required permissions, create a dedicated user (mongodb_exporter) with the following roles:

{
   "role": "clusterMonitor",
   "db": "admin"
},
{
   "role": "read",
   "db": "local"
}

These roles allow the exporter to:

• Monitor cluster-wide metrics (clusterMonitor role)
• Read local database statistics (read role on local database)

Creating the MongoDB Exporter User

Connect to MongoDB and run the following command:

use admin
db.createUser({
   user: "mongodb_exporter",
   pwd: "your-secure-password",
   roles: [
      { role: "clusterMonitor", db: "admin" },
      { role: "read", db: "local" }
   ]
})

Important: Replace your-secure-password with a strong, unique password.

Setting Up mongodb_exporter

Option 1: Using Docker

To run mongodb_exporter using Docker, execute the following command:

docker run -d -p 9216:9216 percona/mongodb_exporter:0.42 \
  --mongodb.uri=mongodb://<username>:<password>@127.0.0.1:27017/admin?authSource=admin

This command runs the exporter and exposes it on port 9216, allowing Prometheus to scrape MongoDB metrics.

Option 2: Using Systemd

Step 1: Download and Install mongodb_exporter

# Download the exporter
wget https://github.com/percona/mongodb_exporter/releases/download/v0.42.0/mongodb_exporter-0.42.0.linux-amd64.tar.gz

# Extract the archive
tar -xvf mongodb_exporter-0.42.0.linux-amd64.tar.gz

# Move to the extracted directory
cd mongodb_exporter-0.42.0.linux-amd64

# Copy binary to system path
sudo cp mongodb_exporter /usr/local/bin/

Step 2: Create a Systemd Service File

Create the service file:

sudo nano /etc/systemd/system/mongodb_exporter.service

Add the following configuration:

[Unit]
Description=MongoDB Exporter
Wants=network-online.target
After=network-online.target

[Service]
User=<username>
Group=<username>
Type=simple
ExecStart=/usr/local/bin/mongodb_exporter \
    --mongodb.uri=mongodb://<username>:<password>@<host>:27017/admin?authSource=admin \
    --web.listen-address=":9216" \
    --mongodb.direct-connect \
    --mongodb.global-conn-pool \
    --collect-all \
    --collector.replicasetstatus \
    --collector.dbstats \
    --collector.topmetrics \
    --mongodb.connect-timeout-ms=180
Restart=always
RestartSec=10

[Install]
WantedBy=multi-user.target

Note: Replace <username>, <password>, and <host> with your actual values.

Step 3: Start and Enable the Service

# Reload systemd daemon
sudo systemctl daemon-reload

# Start the service
sudo systemctl start mongodb_exporter

# Enable on boot
sudo systemctl enable mongodb_exporter

# Check status
sudo systemctl status mongodb_exporter

Configuring Prometheus

Step 1: Edit Prometheus Configuration

Add the MongoDB exporter target to your prometheus.yml:

scrape_configs:
  - job_name: 'mongodb'
    static_configs:
      - targets: ['localhost:9216', 'host:9216']

Step 2: Restart Prometheus

sudo systemctl restart prometheus

Visualizing Metrics in Grafana

Import a MongoDB Dashboard

1. Navigate to Dashboards > Import
2. Use an existing dashboard ID from Grafana Labs
   • Recommended: Search for “MongoDB” dashboards
   • Popular options include dashboard IDs like 2583 or 12079
3. Select your Prometheus data source
4. Click Import

Once configured, you can monitor MongoDB performance with real-time metrics and alerts.

Key Metrics to Monitor

Metric	Description
mongodb_connections	Current number of connections
mongodb_opcounters_total	Database operations count
mongodb_ss_mem_resident	Resident memory usage
mongodb_rs_members_health	Replica set member health
mongodb_dbstats_dataSize	Database data size

Exporter Flags Reference

Flag	Description
--mongodb.uri	MongoDB connection URI
--web.listen-address	Address to listen on for metrics
--mongodb.direct-connect	Direct connection mode
--collect-all	Enable all collectors
--collector.replicasetstatus	Collect replica set status
--collector.dbstats	Collect database statistics
--collector.topmetrics	Collect top metrics

Troubleshooting

Common Issues

1. Connection refused: Ensure MongoDB is accessible from the exporter
2. Authentication failed: Verify the user credentials and roles
3. Metrics not appearing: Check if port 9216 is open and accessible

Verify Exporter is Running

# Check metrics endpoint
curl http://localhost:9216/metrics

Conclusion

This guide provides a step-by-step approach to setting up mongodb_exporter with Prometheus and Grafana for MongoDB monitoring. With this setup, you can track database performance, resource utilization, and health metrics effectively.

Next Steps

• Set up alerting rules in Prometheus for critical metrics
• Create custom Grafana dashboards for your specific use case
• Configure log aggregation for complete observability

---

For more information, refer to the official mongodb_exporter repository.