AWS S3

Each row below is one probe-side measurement we run from every vantage point. The availability number on this page is the consensus across these checks — not a single endpoint hit.

About this data

Availability is measured independently by probes in five locations. Each probe runs the appropriate mix of HTTP, DNS, TLS, and storage checks at per-check cadences and records timing, status code, and any error. A probe votes "up" for a given minute if any of its checks for this service succeeded; the minute is then counted as "up" on the leaderboard only when a strict majority of reporting probes voted up. Minutes where fewer than three probes reported are excluded from the denominator. Full methodology →