I've never configured autoscaling. Can I play?
Yes. Your control room starts with working defaults, and the practice wave shows what each lever changes as you move it.
Can I play before ranked sign-ups open?
Yes. Practice mode needs no sign-in: same game, same levers, against sample teams, with your runs kept in this browser. Ranked runs need a Qovery sign-in and a real environment for your team; you can ask for one email when sign-ups open.
What else is there between ranked runs?
A daily challenge: one scenario a day for everyone, with three levers and its own ranking, and a streak if you come back every day. Achievements for things like zero lost customers, half the budget or a personal best. And practice waves, as many as you want.
When can I play?
Any time. Every calendar month is a season, from the 1st to the last day (UTC). Your team and your app stay from one month to the next; the leaderboard starts fresh on the 1st, and past months stay public.
Who pays for the instances?
Qovery does. The infra bill comes from public instance prices and the money lost from your slow and lost requests. Both are game numbers, and nobody gets a bill.
Do I need a Qovery account? Do I need my own cloud?
Yes, you need a Qovery account. It keeps your session, your team, your dedicated app and your scores from one day, and one month, to the next. Sign in with GitHub, GitLab or Google; if you don't have an account yet, signing in creates one, free and without a card.
You don't need any infrastructure, cloud account or credit card. Your team's app runs on Qovery's infrastructure, and Qovery pays for it.
Do harder leagues earn more points?
Not in the monthly ranking: each league has its own, since the scenarios and budgets differ. In the yearly championship, yes: a place earns 25, 18, 15, 12, 10, 8, 6, 4, 2 or 1 points, times ×1 in Beginner, ×2 in Intermediate, ×3 in Expert and ×4 in KEDA. Winning Expert is worth three times winning Beginner.
How is the score computed?
Each request is fast (under 300 ms), slow (300 ms to 5 s) or lost (an error, or no answer within 5 s). A scenario scores fast − slow − 5 × lost, divided by what your instances cost at public cloud prices. Under 99% availability, or over the p95 latency target, the scenario score is halved. Over budget in any scenario and the run is disqualified. Your league score is half your average scenario and half your weakest one.
What runs behind the game?
Each team gets its own environment on Qovery, on a Kubernetes cluster managed by Qovery. Your app is a small Go service, and the levers are its Kubernetes autoscaling settings: the Horizontal Pod Autoscaler (CPU or memory target, min and max replicas, scale-down window, scale-up policy), the readiness probe, spot capacity, and KEDA scalers (request rate, cron, scale to zero) in the KEDA league. The control room applies them through the Qovery API in about 15 seconds.
The load generator runs inside your environment and calls your app on its internal address, so internet latency never counts against you.
What do winners get?
Each month, the podium team gets three months of a Qovery plan for its company and a gift card for the players to keep: Scale and $500 for first, Growth and $300 for second, Starter and $100 for third. If your company already pays for Qovery, you get the gift card but not the plan months. Every player also gets a result card to share after each ranked run. The plans are described on qovery.com/pricing.
Can I see how others did it?
At the end of each month, the top 10 settings of the season are published on the leaderboard.
10x in 15 seconds: how fast do you react?
A calm, then it comes back: did you scale down too far?
Ranked waves shift the timing, so you can't hard-code a schedule
Beginner: learn how your app reacts to a rush
Intermediate: trade speed against your bill
Expert: cheap servers that can vanish mid-wave, and every second of startup
Blow the budget and you're out
