How I learned to stop worrying and wrote my own memory-cache
Here I am, writing about performance optimization again. I'm a performance junkie. Constantly monitoring and investigating bottlenecks for our SaaS helpdesk webapp is my favorite thing ever. And I'm proud to say that with thousands of clients, even some really big ones, our app's backend process rarely goes higher than 5-6% CPU. Even during the peak load times which happen around 2-3pm UTC - the time when "Americas are waking up while Europe is still very active".