<b>Q: Everyone says "check every 60 seconds." Is faster always better?</b>

<b>Q: Everyone says "check every 60 seconds." Is faster always better?</b>

A: No. Your check interval should match your time-to-recover, not your ego. A faster interval only helps if you can act faster. If your on-call needs 15 minutes to respond, a 10-second check just buys you 50 seconds you can't use.

What faster intervals actually cost you:
— More requests hitting your origin (a busy status endpoint can add real load)
— More chances for a single network blip to fire a false alarm

Practical rule: critical revenue paths at 30-60s, supporting services at 2-5 min, low-priority internal tools at 5-10 min. Then require 2-3 consecutive failures before alerting, so a single dropped packet never wakes anyone.

Got a question? Drop it in the comments.

—
Доп. контекст по lcp basics — @CoreVitals101