Should every alert wake someone up at night?
Q: Right now everything pages the phone, day or night. People are burning out. How do I decide what's worth a 3am call?
A: A reader fixed this by routing on two axes: severity and time of day. Before, all 100% of alerts went to phones around the clock. They measured that only about 15% of nighttime pages were for things that couldn't wait until morning.
They defined three tiers. Critical (site down, payments failing) pages 24/7. Major (one region degraded, elevated errors) pages during business hours but goes to a queue overnight. Minor (a single slow check, a cert at 30 days) never pages — it lands in a daily digest.
Result: overnight pages dropped 78%, on-call sleep improved, and not a single real emergency was downgraded, because the critical tier was defined narrowly and tested.
The follow-up: get the team to agree on tier definitions in writing before launch. The arguments happen at definition time, not at 3am — which is exactly where you want them.
Got a question? Drop it in the comments.
Pingback Clinic
@PingbackClinic
Should every alert wake someone up at night?
Этот пост опубликован в Telegram-канале Pingback Clinic. Подписаться можно по ссылке: @PingbackClinic.