Your Java service has to respond with a p99 of no more than ten milliseconds. How would you design it with the JVM's characteristics in mind, and how would you find the source of latency if the target isn't being met?
~10 min by voice
How to approach it
The interviewer wants to see whether you understand where tail latency comes from in a JVM service (GC, JIT, safepoints, the container, locks, the network), whether you can design to avoid it, and whether you can investigate the problem systematically with tools rather than guesses.
The interviewer will ask this question and listen. Think out loud and ask for a hint if you need one. When you're done, pause or tap “That's my answer”. In about a minute, your score and the points you covered show up here.