Just wondering if you can share your stack and what kind of traffic did you get (in terms of requests/second)? Always curious as to how much traffic HN really brings and why some sites stay up and others die.
- Time on front page: 15 min
- Views in first 15 min: ~5000
- Views in first 60 min: ~7000
- Views in first 600 min: ~11,000
- Views over the next 5 days: ~5000
HN front pages bring a bunch of users quickly, but it isn't a scale that will cripple moderate hardware with some basic performance tuning (directly serving assets, for example). The tail is also great, fully 3/4 of my post-front-page views come from click-throughs on social.
I saw similar numbers on my server. About 10,000 views over the course of an entire day (derived as total data transmitted divided by the size of one full page load), with the bulk concentrated in the first two hours. My nginx served it without breaking a sweat. Peak CPU usage was at 5% and network load peaked at 1 Mb/s (data at 60-second resolution).
My stack was just nginx serving static files, on a 1/1 VM that also hosts half a dozen other services (XMPP, Gitea, Matrix, Mumble etc.).
It is my belief that everyone who encounters the HN hug of death is doing something wrong.
I used to have a blog that ran on Wordpress (3.0 was I think the current version at the time, this was a while ago) with the APC plugin, served via Apache. An article hit HN and it fell over.
Some time later I had migrated to blogofile and nginx and made it to HN again, the server barely noticed.
Generally two possibilities for managing a large load - either a third party blogging provider that can deal with it, or a static site.
Why anyone hosts small blogs on self-hosted VPNs is beyond me - all sorts of problems including security maintenance and peaky load. You can go one further with caching, but that’s yet another moving part that’s not necessary for a simple site.
Here's the Google Cache link: http://webcache.googleusercontent.com/search?q=cache%3Ahttps...