I think HN's stack is the right choice for them and that it fulfills its purpose excellently, but I do seem to recall both of their hard drives failing more or

I think HN's stack is the right choice for them and that it fulfills its purpose excellently, but I do seem to recall both of their hard drives failing more or less simultaneously & HN going down for about 8 hours not that long ago. If that happened at the SaaS company I worked at previously, it would be a bloodbath. The churn would be huge. And our customer's customers would be churning from them. If that happened at a particularly inopportune time, like while we'd been raising money or something, it could potentially endanger the company. (I'd like to stress again this is not a criticism of HN/dang, but just to illustrate a set of requirements where huge AWS spends do make sense.)