Monitoring a Rails app using Prometheus and Grafana — DevOps Project 08

This is me learning and building in public, I have complete 8 DevOps Projects by far, see them all here.

In this one, we will be using Prometheus and Grafana to monitor our rails application, Geeky.

First: key pair

2. security groups

Ensure that your security groups allow the necessary traffic between these instances. For example:

• Rails instance: Allow traffic on port 3000 (or whichever port your app is running on).
• Prometheus instance: Allow traffic on port 9090.
• Grafana instance: Allow traffic on port 3000.
• Allow internal communication between these instances on necessary ports.

We will create two security groups:

• app-sg: for rails
• monitoring-sg: for prometheus and grafana

Create instances

Application prereqs

On a temporary instance, installed ansible, now will run the playbook (this is me taking help from a previous project of mine, in which I wrote ansible playbook that deploys an app to production using just a few commands, here is the walkthrough)

Our app is up

Next we will be setting up prometheus and grafana

The prometheus scraping itself

First steps | Prometheus

Endpoint configured to show metrics

our target is up and running

Install Grafana on Debian or Ubuntu | Grafana documentation

Access it on port 3000 of your grafana instance

Add data source

Give the details

Now create a dashboard

And we can see our metrics on grafana

And on our application /metrics endpoint

/metrics endpoint

So we used monitoring tools — Prometheus and Grafana, to set up monitoring for our rails app.

With the stack of Prometheus and Grafana, we can do a lot more to monitor, analyze, and improve our application and infrastructure. Here are some advanced use cases and capabilities:

Advanced Monitoring and Alerting

1. Advanced Metrics Collection

• Custom Instrumentation: Besides basic metrics, you can instrument custom application-specific metrics to track key performance indicators (KPIs) specific to your business logic.
• Middleware Metrics: Collect metrics from middleware layers such as caching (Redis, Memcached), message queues (RabbitMQ, Kafka), and more.
• External Service Monitoring: Monitor the performance and availability of external services and third-party APIs your application depends on.

2. Alerting with Prometheus

• Alertmanager: Set up Prometheus Alertmanager to handle alerts. You can define alerting rules in Prometheus and configure Alertmanager to send notifications via email, Slack, PagerDuty, etc.
• Anomaly Detection: Use Prometheus to detect anomalies in your metrics, such as spikes in latency or error rates, and trigger alerts based on these anomalies.

3. Infrastructure Monitoring

• Node Exporter: Use the Prometheus Node Exporter to collect hardware and OS metrics from your servers. This includes CPU, memory, disk, and network metrics.

4. Security Monitoring

• Security Metrics: Collect and monitor security-related metrics such as authentication attempts, authorization failures, and other security events.

5. Webhooks and APIs

• Webhooks: Use Prometheus Alertmanager webhooks to integrate with other systems and automate responses to alerts.

Questions, suggestions always welcome.

Connect with me on X @ bhavyansh001, I am learning and building in public!