[oneepic]

On the topic of 'looking around for potential improvements/optimizations': I pretty much eliminated my team's several-GB memory spikes we got from time to time. We all knew they were a problem from our dashboards, but no one investigated it, or wrote a ticket for ourselves. I felt curious one day during some downtime @ work, checked logs, debugged locally, and submitted a tiny patch with pictures of the memory savings during a single test run.

[silicon2401]

Great idea. I had assumed this kind of thing was just due to AWS infra; didn't even think it could be coming from the code itself.