Observing Node.js: Using Metrics to Improve Your Application Performance

This is a test application used to genearte data for the talk at the OpenJS World 2021.

Disclaimer: It can be used as a guideline for setting up metrics and dev envirnoment with Prometheus and Grafana, but this service does not have many of the characteristics you want in a production-ready service. Please DO NOT use it in production.

Usage

This setup was made to work with MacOS and only tested on MacOS Catalina.

Setting up containers

1. make sure you have docker desktop installed and on the latest version
2. run the ./setup-docker.sh script (WARN this script will remove existing containers)

Running the server

1. cd app
2. npm install
3. npm start

Server commands

Within the app folder, the following npm scripts are available:

• npm start: starts the server
• npm restart: restarts the server
• npm stop: stops the server
• npm baseline-load: uses autocannon to send a baseline load to the /echo endpoint
• npm load: short for autocannon (you should use autocannon parameters)

Visualizing the metrics

Please notice:

• The metrics might take a coupe minutes to show once the server is started.
• Some metrics will require the baseline-load to run for a couple minutes before they appear.

Steps:

1. Go to Grafana (http://localhost:3001)
2. login with credentials admin:admin
3. select the OpenJS World Dashboard from the Dashboards panel

Troubleshooting

1. npm install fails:
   • maybe you need XCode command line tools installed?
2. ./setup-docker.sh fails:
   • open the file and try to execute each command independently on your shell.