FinPay API Monitoring

Production observability for a fintech payment platform — built with Grafana Cloud, Prometheus, and Grafana Alloy.

Live Production Dashboard

◉ View Live Dashboard →

Opens the live FinPay production dashboard in Grafana Cloud — showing real-time CPU, memory, HTTP rates, response times and event loop lag.

What We Monitor

FinPay is a production-style payment API built with Node.js, PostgreSQL, and Redis. This monitoring setup tracks every critical component of the system in real time.

Layer	Technology	What We Monitor
API	Node.js + Express	Request rate, response time, error rate
Runtime	Node.js	CPU, memory, event loop lag
Cache	Redis (Upstash)	Hit/miss ratio, memory
Database	PostgreSQL (Railway)	Query performance
CI/CD	GitHub Actions	Pipeline success/failure
Infrastructure	Railway	Uptime, CPU, memory

Architecture

finpay-api (Railway)
    └── /metrics endpoint (prom-client)
            ↑
        Grafana Alloy       ← scrapes every 15s
            ↓
    Grafana Cloud           ← stores metrics
            ↓
    Grafana Dashboard       ← visualises data

Stack

• prom-client — Prometheus client for Node.js
• Grafana Alloy — lightweight metrics agent
• Grafana Cloud — hosted Prometheus + dashboards
• PromQL — query language for metrics

Key Insight

Before adding any monitoring, understand what your app does. Each architectural decision — JWT auth, Redis rate limiting, atomic transactions — should have a corresponding metric that tells you when that component is misbehaving.

This is the difference between generic CPU dashboards and meaningful production observability.

Monitoring Sections

Section	What You'll Learn
App Instrumentation	Adding prom-client, exposing /metrics
Grafana Alloy Setup	Installing and configuring the scrape agent
Grafana Dashboard	Building panels with PromQL
Security	Keeping secrets out of git
Production Deployment	Monitoring the live Railway app

✦ Test Your Knowledge

1.What is the primary purpose of adding monitoring to a fintech application?

ATo make the app look more professional

BTo detect and respond to incidents before users are impacted

CTo slow down the application intentionally

DTo replace logging entirely

2.Which tool scrapes the /metrics endpoint and pushes data to Grafana Cloud?

APrometheus Server

BGrafana Agent

CGrafana Alloy

DDatadog

3.Why must you understand the application architecture before setting up monitoring?

ATo write better code

BBecause each architectural decision needs a corresponding metric

CTo reduce cloud costs

DMonitoring works the same regardless of the app