Proving SQLx’s Statement Cache with bpftrace

Mon, 24 Aug 2026 11:39:24 +0200

During the blissful one week off I spent without touching any AI assistants, I realized I’d stopped learning new things during regular paid work.

Now my job requires me to work with the clankers, so I don’t get to observe the APIs or hit my head against the invalid imagination of some gnarly problem.

I decided to spend some time restarting the habit of writing down what I discover and, more importantly, how I prove it to myself. Let’s start with some surprising optimization in sqlx

TIL on FlakM blog

Proving SQLx’s Statement Cache with bpftrace