I'm super curious as to the setup behind the URL. I work in web infrastructure and host numerous websites for automated bank processing. Just curious, is there any information anywhere about how Hubski runs and what you guys are using as far as servers/loadbalancers/solutions used? I've heard you talk about the Hubski infrastructure in bits and pieces over the place, but never seen the big picture. Mostly just curious about how many servers you have any their roles and how it's kind of logically separated. I understand if there isn't any information you feel like handing out. I'm just curious. :)
ATM it feels like we are running on hamster wheels. :) I'll share some stuff when I have a bit of time. The short story is that we are running an Arc server (one machine) behind an nginx reverse proxy. Hacker News pulls it off (mostly), and we are learning how. Our quickest fix since the influx was to drastically reduce data in memory, and resize. Now we are doing more targeted things.