Create exam simulations with Killer Coda

Everybody loves practice exams. Here's how to create your own.

Hands-on vs learning theory

From anecdotal evidence, I’ve always suspected that the average engineer dislikes learning from theory and text books, and vastly prefers hands-on learning. Personally I prefer to learn the theory first, then hands-on, then fall back to theory when I’m going deeper.

I polled a field of 80 SE’s on a recent call.

  • About half (39/80) said that they learned AI by building something on their own.
  • When asked for most vs least effective learning methods:
    • the majority (44/69) said that building something themselves was most effective
    • the majority (38/66) said that instrustor-led training was the least effective

To that end, I want to train the field on Kubernetes with a mix of coursework (using Coursera), instructor-assisted learning with directions, and some hands-on labs.

Why Killer Coda?

In an ideal world, I’d find a pre-existing practice exam that was 100% aimed at what I think F5 SE’s need to know. But I haven’t, and since everyone likes practice exams, I think I can make my own.

Killer Coda is free for creators and end-users alike. Creators can create scenarios and courses and share them with the public, all without subscribing. I’d be happy to subscribe personally, and may do so in future. A PLUS Membership gets you faster load times and prioritized support, as well as a few other benefits.

How to create your own labs.

Long story short, they have good documentation here. I will just say:

  • you need:
    • a github repo
    • a Deploy Key (even if your repo is public, so that Killercoda can access your repo without getting rate limited as anonymous requests do)
    • a webhook set up (so GitHub can alert Killercoda when a commit is made)
  • for me, just getting started, I recommend copying a very basic example scenario directly into your repo and playing around
    • don’t try to make a perfect lab in your first commit (like I did). Waste of time, just iterate.

My first scenario:

Check this out at https://killercoda.com/michael-oleary/scenario/nginx-ingress-controller

Other platforms

I do plan to explore some other platforms after this too. Here’s a summary table of some free options. (Disclaimer: I generated this table with ChatGPT.)

Platform What It Is Free Hands-On Features Best Paired With Structured Learning Limitations
Killercoda Browser-based Kubernetes scenario labs Real Kubernetes CLI environments; task-driven scenarios; no local setup Kubernetes courses, docs, or certification study plans (CKA/CKAD/CKS) No built-in curriculum; scenario coverage varies
Play with Kubernetes (PWK) Temporary Kubernetes sandbox On-demand multi-node clusters; full kubectl access Tutorials, blogs, or instructor-led material that needs a live cluster Sessions are time-limited; no guidance or assessment
LabEx (Free Kubernetes Labs) Interactive Kubernetes labs & playground Browser-based clusters; guided hands-on exercises Intro/intermediate Kubernetes learning paths Less certification-focused; fewer advanced scenarios
IBM CloudLabs (Kubernetes) Guided, cloud-hosted Kubernetes labs Step-by-step interactive labs using managed Kubernetes Structured lesson sequences or workshops Requires free cloud account; limited lab selection
KodeKloud – Free Labs Concept-aligned Kubernetes labs Hands-on tasks mapped to Kubernetes topics Kubernetes courses, certification prep outlines Full lab catalog requires paid access
Google Cloud Skills Boost (Free Tier) Cloud-based Kubernetes labs (GKE) Free credits; guided labs and quests; real clusters GKE tutorials, cloud-focused Kubernetes learning Free access is credit-limited; GKE-specific focus

Next steps

Next, I’m going to finish a very nice example that includes a startup script, a validation script, and multiple scenarios grouped into a single course. This should make it easy to copy in future.

Updated: