I was tasked with adding daylight savings (and time-zone) support to an internal monitoring system. At first I grumbled mightily because I couldn't see the poi

I was tasked with adding daylight savings (and time-zone) support to an internal monitoring system. At first I grumbled mightily because I couldn't see the point to it, but after I'd plumbed everything through, I could definitely see the data points snapping in and out of alignment when turning on and off the DST adjustment. Assume a "moderate" ~45° slope or sine wave... typical diurnal traffic pattern. When shifting by 1hr, that's either 1/24 or 1/12 (call it 5-10%) of "deferred peak traffic"... definitely detectable even with moderate sensitivity. DST support gave us back confidence in alerting over an abnormally large time range too! Most alerts were set up with a 10-week lookback, discarding both the highest and lowest weeks, so DST changes twice per year potentially threw off alerting for up to 4 months! I have no doubt that I could detect DST with damn near anything, let alone some crazy calibrated laser inferometer machine!