@dang posts it from time to time, I think this is most recent one: https://news.ycombinator.com/item?id=28479595

Last month HN had issues with SSD's suddenly dying after 40000 hours and moved to AWS temporarily: https://news.ycombinator.com/item?id=32031639