According to @dang (https://news.ycombinator.com/item?id=28479595) via @sctb (https://news.ycombinator.com/item?id=16076041)

  We’re recently running two machines (master and standby) at M5 Hosting. All of HN runs on a single box, nothing exotic:
  CPU: Intel(R) Xeon(R) CPU E5-2637 v4 @ 3.50GHz (3500.07-MHz K8-class CPU)
  FreeBSD/SMP: 2 package(s) x 4 core(s) x 2 hardware threads
  Mirrored SSDs for data, mirrored magnetic for logs (UFS)

(beginner question) How do they store the data? is an SQL db on overkill for such a use case? what would be the alternative? an ad-hoc filesystem based solution? then how do the two servers share the db? and is there redundancy at the db level? is it replicated somehow?

petercooper

"ad-hoc filesystem based solution" is the closest of your definitions, I think. Last time I saw/heard, HN was built in Arc, a Lisp dialect, and use(s/d) a variant of this (mirrored) code: https://github.com/wting/hackernews

Check out around this area of the code to see how simple it is. All just files and directories: https://github.com/wting/hackernews/blob/master/news.arc#L16... .. the beauty of this simple approach is a lack of moving parts, and it's easy to slap Redis on top if you need caching or something.

There is a modern maintained variant at https://github.com/arclanguage/anarki/tree/master/apps/news as well if you want to spin up your own HN-a-like and have the patience.

File syncing between machines is pretty much an easily solved problem. I don't know how they do it, but it could be something like https://syncthing.net/ or even some scripting with `rsync`. Heck, a cronned `tar | gzip | scp` might even be enough for an app whose data isn't exactly mission critical.