1 min lesson
Sensitivity vs alert fatigue
Walk through each part of "Sensitivity vs alert fatigue", then explain what each one does.
Step 1 of 2
Sensitivity vs alert fatigue
Every detector has a knob between catching real regressions and crying wolf. Tie that knob to economics, not vibes. The cost of a missed regression is its blast radiusHow much breaks if a change goes wrong; the scope of potential damage. Press Enter for the full definition. in frustrated users; the cost of a false positive is an engineer's interrupted afternoon and slow erosion of trust in the alert. Calibrate against your error budget: page when burn rate threatens the SLO, file a ticket when it merely nicks it.
Calibrating the alarm
- Page (wake someone)
- Fast burn of the error budget; user-visible and spreading
- Ticket (next business day)
- Slow burn; real but bounded, no spreading
- Dashboard only
- Inside noise or within budget; watch, don't interrupt
- Suppress
- Known maintenance window or expected deploy blip
The threshold is a business decision about false-positive cost, dressed up as statistics.