Two phone screens for Meta PE (the SRE-equivalent role there), 45 minutes each.
Round 1, System (full-loop troubleshooting): same prompt as another thread posted on this forum, but the interviewer didn't seem to have a specific answer in mind. They asked a bunch of pretty open-ended things — how do you make sure everything keeps running smoothly during the days you're on call? If something breaks, how do you troubleshoot it? If a user's requests keep not getting a response, what could be wrong? (I said maybe a bad DB query, and you'd check the DB logs.) How do you quickly figure out what's running on a server? (I said you could check /etc/configs, since anything beyond the default services usually has its config there.) What limited resources does a server have? And finally, if you had to give your uncle some advice on running a website, what would you tell him? For every single question, the interviewer moved on to the next one before I'd really gotten into it. I still don't know what they were actually looking for. My gut says this is the round that got me.
Round 2, coding: (the original post here linked a couple of practice problems with garbled, homophone-coded labels, so I can't tell exactly which ones they were). Got lucky with the interviewer this round — we jumped straight into the problem, and I finished with 10 minutes to spare. Going into the interview, and even right after finishing, I felt pretty confident, because I've spent the last 6 years doing SRE and platform work, and every time I was on call I managed to stop the bleeding, and I've also helped other people on the team with their on-call issues. I always figured troubleshooting was my strength, and somehow I still got rejected out of nowhere... I guess job hunting really does come down to luck. Thanks to the interviewer for the help — wishing him a good, safe life. 🙏
Discussion
Loading comments…