Practise fixing outages.
Before it's your pager.
Realistic incidents in a simulated terminal. Investigate, find the cause and restore service, with nothing real at stake. Try a free lab from each track now. Enrolled students unlock every lab in their course.
No sign-up needed for free labs · Works in your browser
Follow the evidence
Read logs, check what changed, and question the dashboards, just as you would on a real on-call shift.
Make the fix
Roll back, reconfigure, contain or patch. A wrong change costs time and points, the same way it would at work.
Learn from the debrief
See the real root cause, replay every decision you made, and take away what to do next time.
Cloud & DevOps
Outages on AWS: bad releases, expired certificates, exhausted databases.
The Friday release
Customers could pay ten minutes ago. Then a new checkout release went out and payments started failing. Find the cause and get checkout working.
AWS · ECS · ALB05:00 timed
The midnight expiry
Every dashboard is green, but customers can't open the site. Look into a security certificate that expired overnight.
AWS · ALB · ACM10:00 timed
Part of AWS Cloud Computing →
Too many connections
A marketing campaign is live, the API is timing out, and the database CPU is low. Find what is blocking requests and get the API responding.
AWS · ECS · RDS PostgreSQL10:00 timed
Part of AWS Cloud Computing →
Security operations
Alerts from a real SOC queue: brute force, phishing, and malware that won't stay dead.
Knocking at the door
Wazuh raised a high-severity SSH alert on the bastion host. Work out whether anyone got in, contain the source, and classify the alert.
Wazuh · Linux · SSH08:00 timed
One click too many
A finance user signed in from Lagos and the Netherlands six minutes apart. Trace the phishing email, find out what the attacker set up, and lock them out.
Microsoft 365 · Entra ID · Proxy12:00 timed
Part of SOC Analyst →
The quiet miner
web-01 is at 99% CPU and checkout is slow. Something is running that shouldn't be, and it comes back when you kill it.
Wazuh · Linux · nginx15:00 timed
Part of SOC Analyst →
Software development
Broken builds, breaking migrations, and scripts that run out of memory.
Green locally, red in production
The Black Friday page works on everyone's laptop and in the preview deploy, but the production build keeps failing. Launch is in an hour.
Next.js · GitHub Actions · Hosting07:00 timed
The column that vanished
The orders API started returning 500s at 10:02, but nobody deployed it today. Find out what changed underneath it and restore order history.
FastAPI · PostgreSQL · Alembic10:00 timed
Part of Python Programming →
The script that ate the server
The nightly sales report has worked for a year. Last night it died without a traceback, and Finance needs it by 9am.
Python · Linux · cron10:00 timed
Part of Python Programming →
Get trained for the job, not just the exam.
Every Academy course in cloud, security and software development includes hands-on incident labs, instructor-reviewed projects and live sessions.