Hands On - OpenTelemetry

Abstract

You can start by reading the abstract (in French) of this hands on.

Prerequisites

• OS session with administrator permissions
• IDE like IntelliJ or Visual Studio Code, supporting Java
• PowerShell 7 if you're using Windows: get it from the documentation using WinGet
• Bash (or zsh of fish) if you're using MacOS or Linux
• Docker
• Git and a GitHub account
• mise: don't forget to properly add it to your shell with mise activate <shell> after the installation
• Headlamp, unless you master the command line around kubectl

Repository content

• microservices: 4 Spring Boot applications
• observability: configuration of the observability stack
• presentation-slides: the presentation in format Slidev
• signoz: configuration of SigNoz
• traffic-simulation: script to simulate the traffic on the microservices

The presentation slides are available on GitHub Pages.

Setup

1. Fork this repository (you will need your own GitHub repo for later stages of this hands-on) and clone it locally.

2. Run mise install at the root of the repository.

3. Depending on your OS, run the according script to setup the environment (setup.sh or setup.ps1) . The script check an install pre-requisites for the TP.
   It also setup a local k3d cluster with the necessary config.

4. Configure your hosts

As we are all working locally, and to make this hands-on easier for all, we need to all use the same urls. To do so, modify your hosts file with the DNS entries for each microservice:

• On MacOS or Linux, open "/etc/hosts" with any text editor, as "sudo".
• On Windows, open "%SystemRoot%\System32\drivers\etc\hosts" with any text editor, launched as an administrator.
  • Either enable sudo, and then run sudo notepad
  • Or right click on Notepad and choose "Run as admnistrator"
• Add the following line to "hosts":

127.0.0.1 order.k3d.local product.k3d.local shopping-cart.k3d.local stock.k3d.local signoz.k3d.local

Check in Headlamp that everything is fine.

You can reach the APIs documentation at the following URLs: order, product, shopping-cart, stock.

Restart

If you need to relaunch the cluster (after exiting down Docker for example):

k3d cluster start otel-hands-on

Cleaning

If you'd like to clean your computer after the lab:

• Remove the line added to your hosts file
• Uninstall Headlamp
• Uninstall mise: mise implode (more detail on the official documentation)
• Remove the cluster and its content:

k3d cluster delete otel-hands-on # should remove everything, except the images

docker image ls # list all Docker images
docker container ls # list all Docker containers
docker system prune # remove all unused data
docker image rm <IMAGE ID> # remove a specific image
docker container rm <CONTAINER ID> # remove a specific container

Labs

• Lab 1: Work without observability
• Lab 2: Let's add some observability
• Lab 3: Add some business value
• Lab 4 (Bonus): Filter traces