Recent searches
Search options
The fact that something as immense and complex as Twitter is still running after all the engineers were fired or provoked to quit really says something about their quality.
Every tech company should immediately offer them bags of money to work the same magic on their site.
@SFoskett As someone who works with infrastructure, it's not that surprising to me that the system works fine under normal loads. It's really when you start changing things is when downtime is most likely. A system left in a stable configuration could run until the electricity shuts off. Of course in this interdependent, cloud based, architecture that is the norm these days, engineers don't have control over what AWS is changing and so something will break eventually.
@theronconrey@fosstodon.org@Steve @SFoskett twitter runs on bare metal. They're wholly responsible for all of the widgets, app deployments and resiliency we see. Scale out architecture at this scale is pretty limited in terms of companies and teams supporting this. Going back to AMZ for twitter would be a wild direction to take, and would be a public ack they can't (or simply don't have the userbase requiring it) do it in house anymore.