Is it normal to see high variance in response times, and if so, are there recommended steps for mitigating that? I haven't set up the telemetry yet, just looking for advice if y'all have any. I'm running a read/check-heavy operation that sometimes takes 30ms and sometimes takes 300ms, which falls in and out of "a user will notice this latency" territory. It's on localhost with only one user, so not a contention thing