We poll each provider's official public status feed every few minutes — statuspage.io APIs
(/api/v2/summary.json), Instatus feeds, RSS/Atom feeds, and vendor JSON feeds
(AWS Health, GCP Service Health, Azure RSS, Stripe). Status values are normalized into
operational / degraded / partial / major / maintenance / unknown.
Independently of what providers report, we ping each provider's public endpoint from multiple regions and record reachability + latency. A 401/403 still counts as reachable (the service answered) — timeouts and 5xx do not. Probes are what power our latency leaderboard and "down for everyone or just you?" region data.
Anyone can file a "report a problem" on a provider page. Reports are rate-limited, hashed (no raw IPs stored), and shown as unverified crowd signal — never merged into official status.
cloudstatus.io is an unofficial aggregator and is not affiliated with any provider. Names and statuses belong to their owners; we link to the official status page on every provider page. Something wrong? abuse@cloudstatus.io