Trust & Transparency
How rimetry detects, scores, and reports internet outages — the methodology, regions, and signals behind every status.
Technical monitoring
Every service is probed from four global regions on a 1–5 minute cadence. We require persistence across multiple cycles before flipping a service to Down, and we cross-check vendor status pages, DNS health, and TLS validity.
Community signal
Verified user reports contribute 20% to status confidence, with anti-spam scoring, IP fingerprinting, and rate limits. Reports never override technical signals on their own.
Status tiers
We use five tiers instead of binary up/down to reflect real-world performance.
- OnlineAll probes succeed, vendor and community signals confirm.
- SlowResponse time exceeds baseline by 2× across multiple regions.
- Degraded≥1 region failing or vendor reports partial issue.
- Partial outageMultiple regions failing or major feature down.
- DownMajority of regions fail repeatedly, persistence gate cleared.
Probe regions
Active probe locations contributing to consensus.
Scoring formula
Health score and confidence are deterministic — no hand-tuning, no hidden weights.
Health = 0.5 × Uptime
+ 0.3 × (1 − IncidentWeight) × 100
+ 0.2 × CommunityConfidence
− LatencyPenalty
Confidence = 0.4 × Technical + 0.4 × Regional + 0.2 × CommunityPrivacy & data
Reports are stored with a hashed IP only. We don't sell data and we expose internal anti-abuse fields only to admins.
Live accuracy
We publish live numbers on every probe, vendor agreement, and mean time to detect.
View live accuracy →Maintained by rimetry
This page is maintained by rimetry to answer common questions about how our uptime monitoring works. It describes rimetry's enabled controls, not a third-party certification.