World map

 

BM

World K5: Lunch Rush

Prometheus, Grafana and OpenTelemetry

You will learn: how numbers over time, charts and traces show what happened last Tuesday.

Checked with Kubernetes 1.37 on 31 Oct 2026

The problem

Last Tuesday at 12:30 the shop was slow. kubectl top only shows right now. How do you see last Tuesday?

Watch

Prometheus: A clerk who visits every stall on a timer and writes the numbers down, with the time. is a clerk with a clipboard. It visits every stall on a timer and writes the numbers down, with the time. Put its rounds in order.

First step at the top. Put these in order. Use the arrows, or drag.

  • It stores each number with its time
  • A chart or an alert shows the answer
  • Prometheus visits (scrapes) /metrics on a timer
  • Your app shows its numbers at /metrics
  • You ask questions in PromQL

Watch

Grafana: The big screen on the wall: charts and dashboards from your numbers. is the big screen on the wall. It draws charts from numbers that tools like Prometheus keep. Who does what?

Drag a card into a box. Or tap the card, then tap the box.

      Check

      A trace: One order's whole journey, step by step, with how long each step took. follows one order’s whole journey. It times each step. Here is order 42:

      gateway   2 ms
web      12 ms
api      30 ms
db      900 ms

      Where is the slow part?

      Watch

      Metrics, logs and traces are three kinds of clues. OpenTelemetry (OTel): One shared way to collect metrics, logs and traces, so you can switch tools later. (OTel) is one shared way to collect all three. Sort each clue.

      Drag a card into a box. Or tap the card, then tap the box.

            Check

            Good habit, or needs work?

            Swipe right for good, left for needs work. Or use the buttons.

            1 / 4

            Add OTel to the app once. Send its data to the tool you choose.

            You did it!

            You can now explain metrics, logs and traces, and who keeps them.

            Go deeper (optional)

            Many teams install Prometheus and Grafana together with a Helm chart. Kubernetes parts show their own /metrics too. An OTel Collector can receive, change and send on all three signals. Real rep (optional): try a ready lab on Killercoda. Or use a throwaway kind cluster.