1
Pick a scenario
One failure, one hypothesis. "If the payments API is slow, checkout degrades gracefully."
⚠ Guide
What is a GameDay?
A GameDay is a planned exercise where a team deliberately breaks part of its own system and practises the response: one failure, one hypothesis, a facilitator, and a blameless review afterwards. Amazon made them famous; any team with customers depending on uptime should run them.
Where GameDays came from
Amazon ran GameDays in the mid-2000s to prove its systems and teams could survive real failures. The idea spread through AWS, Netflix and the wider SRE community, and today GameDays are a standard part of reliability practice.
Plan your first GameDay
2
Assign roles
Facilitator, operator, observers, and someone taking notes.
3
Run it
Inject, observe, respond. Abort if limits are hit.
4
Review
Blameless write-up, action items, and a date for the rerun.
GameDay FAQ
+ How long does a GameDay take?
Two to four hours for the exercise itself, plus an hour of review.
+ Is an AWS GameDay the same thing?
AWS also runs competitive GameDay events for learning. The practice here is the internal exercise on your own systems.
Safe by design
Every fault Oh My GDaaS runs is aimed at infrastructure you have proved you own, through a DNS record, a scoped cloud role or a deployed agent. Every run has a blast-radius limit, a hard time limit and an abort button. We are not a booter, stresser or DDoS service and never will be.