It’s not infinite, it better be timeout but frequently isn’t and yes there is variance to it but usually (handwave) not any more variant than the tail percentil

It’s not infinite, it better be timeout but frequently isn’t and yes there is variance to it but usually (handwave) not any more variant than the tail percentile that teams choose. I’ve had the experience multiple times where simply shifting a graph to max shows that timeouts/load shedding is t working, then teams get to the point where they are hitting timeouts way more than they thought. Only after working through those issues do you get to actually improving latency. The upside of simplicity in the number is only overtaken by the downside when you start chasing real time constraints in systems that don’t need it.