Cutting Python Web App Memory Over 31%

tl;dr; I cut 3.2 GB of memory usage from our Python web apps using five techniques: async workers, import isolation, the Raw+DC database pattern, local imports for heavy libraries, and disk-based caching. Here are the exact before-and-after numbers for each optimization. Over the past few weeks, I’ve been ruthlessly focused on reducing memory usage on my web apps, APIs, and daemons. I’ve been following the one big server pattern for deploying all the Talk Python web apps, APIs, background services, and supporting infrastructure. There are a ridiculous number of containers running to make everything go around here at Talk Python (23 apps, APIs, and database servers in total). Even with that many apps running, the actual server CPU load is quite low. But memory usage is creeping up. The server was running at 65% memory usage on a 16GB server. While that may be fine - the server’s not that expensive - I decided to take some time and see if there were some code level opti