Replication is one of those things that works perfectly — right until someone actually needs fresh data. That’s when your replica decides to reflect on life choices and drift a few seconds (or minutes) behind reality. Suddenly “eventually consistent” feels less like an architectural decision and more like a personal attack. In this post, we’ll break down why replica lag happens, how to measure it properly, and most importantly, how to reduce it without just throwing bigger instances at the problem and hoping for the best.