Skip to content

Build to Manage - Node.js application monitoring and logging lab

During this lab we will instrument a sample Node.js application for logging to use with log analytics tools like Elastic stack as well as for monitoring with Prometheus and Grafana.

Instrumentation of the application code for monitoring and logging is part of the general concept we call Build to Manage. It specifies the practice of activities developers can do in order to provide manageability aspects as part of an application release.

Lab outline

  • Fork and clone the Github repository and review the source code
  • Configure logging library and add log messages
  • Configure and run Elastic stack in Docker Compose
  • Integrate application logs with Elastic stack
  • What to instrument? - the RED method
  • Instrument application code with Node.js client library for Prometheus
    • Enable a default set of metrics offered by prom-client library
    • Define custom metrics (counter and histogram)
  • Configure and run Prometheus and Grafana stack in Docker Compose
  • Run example PromQL queries
  • Configure sample alert in Prometheus
  • Configure Grafana
    • Import sample dashboard
    • Define a custom dashboard panel
  • Deploy to IBM Cloud Private cluster and configure monitoring using ICP Prometheus and Grafana
  • Configure kubernetes liveness probe using application health check


Install the following software on your workstation or use the provided VM.

  • Node.js
  • npm
  • Docker
  • Docker Compose
  • kubectl

Review the application code and run it locally

Logon to GitHub using your user and password. Access the following repository and click Fork.

From now on you will work on your own fork of the application project<username>/b2m-nodejs

where <username> is your GitHub username.

Clone the b2m-nodejs lab repository to your home directory using:

git clone<username>/b2m-nodejs

Most of the commands in this lab should be executed from the b2m-nodejs/src directory:

cd b2m-nodejs/src

Review the application code in server.js. Application simulates a transaction with random response time on the following URL:


About 20% of requests should return error and 500 HTTP response code.

Run the Node Package Manager to install all required node modules:

npm install 

and then

npm start server.js

to start the application. Node.js server should start and listen on port 3001.

Use internet browser to access http://localhost:3001/checkout

{"status":"RSAP0001I: Transaction OK","transactionTime":"22ms"}


{"error":"RSAP0010E: Severe problem detected"}

Refresh the page a couple of times and you should see random transaction response times.