Hacker News new | ask | show | jobs
by pottspotts 924 days ago
After some checking, it seems to be an issue with the free Heroku instance they are running.
2 comments

Underrated comment of the week.

Amazing what they squeeze out of a single process architecture.

It could be impressive, but the website frequently goes down when big news hit.

I think they could move it to a more powerful machine and the remaining issues would be fixed. It wouldn't be as big of a flex as running on an ancient Intel CPU though.

Edit: I got "We're having some trouble serving your request. Sorry!" while trying to send this comment :)

Looking at that status page, unless it's completely wrong, there's outages multiple times every month. They really could just upgrade their instance and get more mileage out of it and it wouldn't require any "refactoring". Site's mostly fine as is
Amazing... yes, but there are also obvious drawbacks. Take a look at any big thread that has manual and clunky pagination on it, or the site being down because a single provider or disk crapped out. Totally acceptable and fine things depending on the scale of what you're trying to achieve, and especially so for a free website... but it can definitely be better (no, I'm not talking about microservices in Kubernetes on the Edge in the Cloud powered by AI).
Best site on the net and a bunch of google and apple techs want to 'fix' it.
What's the reason for the single process arch, just don't have time/money to scale up?
Would like to know their compute and bandwidth usage.

The front page is ~7 kilobytes on reload, and ~20 on first load - such a good example of "less is more".