|
Something I noticed this week
I was reviewing some work from our current DevOps cohort this week.
One of the tasks was around building the Azure infrastructure for our application.
And honestly, creating the resources is not the difficult part anymore.
You can search how to create Azure Container Apps.
You can search how to create PostgreSQL with Terraform.
You can even ask AI to generate a lot of the code.
|
Then I ask one question:
Why did you do it this way?
|
|
And that's where things become interesting.
|
Why is the frontend public?
|
|
|
Why is the backend internal?
|
|
|
How will the frontend reach the backend?
|
|
|
Why are you using a Service Principal here?
|
|
|
Could Managed Identity work instead?
|
|
|
What happens if the deployment fails?
|
Suddenly this is not just a Terraform exercise anymore.
Now you actually have to think like an engineer.
|
|
|
|
This is what I see again and again
We spend too much time learning tools.
Docker. Kubernetes. Terraform. Azure. GitHub Actions.
Nothing wrong with learning them.
But then you go into an interview and someone says:
|
“We have a frontend, backend and database. Design the infrastructure.”
|
|
Now there is no tutorial telling you exactly what to do.
You have to decide.
And more importantly, you have to explain
why you made that decision.
|
Why I started the cohort
I didn't want to create another course where I share my screen, build something and everyone copies me.
There are already thousands of tutorials like that.
Instead, I give you a ticket.
|
|
How the work actually happens
|
Ticket
Understand
|
→ |
Design
Decide
|
→ |
Build
Implement
|
→ |
Review
Explain
|
|
|
Sometimes your solution works.
Sometimes it doesn't.
That's okay.
Then we review it together.
|
And yes, I ask a lot of questions :)
Why this architecture?
Why is this service exposed?
What happens if this dependency goes down?
How will you deploy without affecting users?
How would you roll it back?
|
Because production doesn't give you step-by-step instructions.
Production gives you something like:
Production issue
“The application is returning 400 errors after deployment.”
That's it.
Is it the application? Database? Networking? Configuration? Authentication? Deployment?
You collect evidence, form a hypothesis and start eliminating possibilities.
|
This is the skill I want people to build.
Not just:
“I know Kubernetes.”
“Give me the problem. I'll figure out what is happening.”
|
|
Next cohort
Want to practice DevOps this way?
I'm preparing the next hands-on DevOps cohort now.
This is not for complete beginners.
If you already understand the basics of Linux, Git, Docker, cloud, Terraform and CI/CD, but you feel like you're missing
real project experience, this is exactly what we're working on.
✓ Real engineering tickets
✓ Architecture decisions
✓ Production-style incidents
✓ Live technical reviews
✓ You explain why you made each decision
Check the program first. Then decide if it's right for you.
|
Not sure if you're ready?
Just reply to this email with COHORT.
Tell me briefly what you're working on right now.
I'll tell you whether I think the cohort makes sense for you.
If you're not ready yet, I'll tell you what I'd learn first.
|
One thing to remember:
Learning Terraform teaches you Terraform.
Being responsible for infrastructure teaches you engineering.
|
|
Learn With DevOps Engineer
Arbaz
YouTube:
youtube.com/@learnwithdevopsengineer
Newsletter:
learnwithdevopsengineer.beehiiv.com
Instagram:
instagram.com/learnwithdevopsengineer
|
