Bring Servers Down
Reported by candidates from Amazon's online assessment. Pattern, common pitfall, and the honest play if you blank under the timer.
Amazon's "Bring Servers Down" problem showed up in June 2024 OAs and it's a simulation problem disguised as infrastructure chaos. You're managing a cascade of server failures, and the trick is understanding the order and dependencies. Candidates often miss that you need to track state changes as servers go down, not just count them. StealthCoder is your safety net if the state machine logic blanks you during the live assessment.
Pattern and pitfall
This is a simulation problem where you track servers falling in sequence and model what happens at each step. The pattern isn't brute force; it's about maintaining the right data structure to track which servers are still up and which have cascading failures. Most candidates try to simulate naively and timeout or get the order wrong. The trick is recognizing that you need to process events in the right sequence and update state efficiently. If you're stuck during the OA, StealthCoder can walk you through the state transitions so you don't lose 30 minutes on debugging.
If this hits your live OA and you blank, StealthCoder solves it in seconds, invisible to the proctor.
You can drill Bring Servers Down cold, or you can hedge it. StealthCoder runs invisibly during screen share and surfaces a working solution in under 2 seconds. The proctor sees the IDE. They don't see what's behind it. Built by an Amazon engineer who would have shipped this the night before his JPMorgan OA if he'd had it.
Get StealthCoderRelated leaked OAs
You've seen the question.
Make sure you actually pass Amazon's OA.
Amazon reuses patterns across OAs. Built by an Amazon engineer who would have shipped this the night before his JPMorgan OA if he'd had it. Works on HackerRank, CodeSignal, CoderPad, and Karat.
Bring Servers Down FAQ
Is this actually a hard problem or just weird naming?+
It's medium. The naming is deliberately vague to throw you off. Once you recognize the state-machine pattern, it's a clean simulation. The gotcha is handling cascading effects correctly and not missing edge cases around simultaneous failures.
Do I need to track individual server IDs or just counts?+
Individual tracking. You need to know which specific servers are down to determine what failures cascade next. A simple counter won't cut it. Use a set or boolean array to track state.
What's the most common wrong approach?+
Trying to compute the final state without simulating the sequence. Servers going down can trigger other servers to fail, so you must process events step by step in the correct order, not jump to the end.
How do I prepare in 24 hours if I've never seen simulation problems?+
Walk through a small example by hand. Write out each step, each state change. Then code the exact steps you wrote. That's the pattern. Simulation is just "do what the problem says, in order."
Will Amazon ask this exact problem again?+
Unlikely word-for-word, but the simulation-cascade pattern is a favorite at Amazon. If you understand how to track state and process events in order, you'll handle whatever variant they throw at you.