OA: Not LeetCode. You're given an actual scenario and you live-code a Service class on HackerRank right there — the logic is fairly simple, basically a bunch of if-else checks just to see whether your logic is clear, not really algorithm-based at all. But there were three parts. My advice: once you've nailed down the requirements, write fast, otherwise you definitely won't have time left to write tests for the third part.
Onsite design round: the interviewer asked about distributed Cron. It's a classic problem — the Google SRE book has a really good explanation of it, and it's also been covered in an ACM Queue article.
Debug round: DasherPicker. The interviewer would not accept your approach at all — you had to do it his way. Advice: he seemed to have a checklist in hand, and every time you called out a bug you'd get a point. Tip: there were a lot of code structure issues to catch — naming convention, exception handling, validation, multithreading, and so on. The bulk of the score is definitely for getting the test cases to actually pass, and production code style seems to be the second priority. For things like the multithreading issue, you can just drop in a #TODO: multithread comment first — the interviewer will quietly note that down and give you credit for it, then ask you at the end how you'd change it to be multithreaded.
Craft round: Bootstrap. The business logic was simple — fetch three external APIs, combine their results, and output them as a new API. It's mainly testing code style. I'd recommend using CompletableFuture to wait for all the calls to return or time out. For retry, I'd recommend writing your own RetryTemplate(Callable<T> call) — that part isn't strictly necessary, it only came up in closing small talk. Getting the code to actually run is still the main thing.
HM round: I didn't answer well. The HM clearly had a checklist in hand — every time you hit a point (like showing stakeholder management skills), they'd check it off, and if you flubbed something they'd mark an X. I felt like I was talking the whole time while the HM was marking X's. They also asked some values questions, like integrity vs. efficiency.
Discussion
Loading comments…