Transparency

Every place we check from.

Live status of our seven Cloudflare edge regions plus out-of-band sentinel nodes on independent ASNs (Hetzner AS24940). Updated continuously. Multi-ASN quorum verification guarantees that a single-cloud provider outage never triggers false customer alerts.

RegionCoversNetworkStatusLast check
πŸ‡ΊπŸ‡Έwnam
Western North AmericaAS13335 (Cloudflare Edge)ONLINEJust now
πŸ‡ΊπŸ‡Έenam
Eastern North AmericaAS13335 (Cloudflare Edge)ONLINEJust now
πŸ‡¬πŸ‡§weur
Western EuropeAS13335 (Cloudflare Edge)ONLINEJust now
πŸ‡΅πŸ‡±eeur
Eastern EuropeAS13335 (Cloudflare Edge)ONLINEJust now
🌏apac
Tokyo, Singapore, Hong KongAS13335 (Cloudflare Edge)ONLINEJust now
πŸ‡¦πŸ‡Ίoc
Australia, New ZealandAS13335 (Cloudflare Edge)ONLINEJust now
πŸ‡§πŸ‡·sam
Brazil, Chile, ArgentinaAS13335 (Cloudflare Edge)ONLINEJust now
πŸ›‘οΈext-sentinel
Multi-Cloud / Heterogeneous ASNAS24940 (Independent VPS (Non-Cloudflare))ONLINEJust now
Multi-ASN Quorum: 4 of 8 healthy regions plus independent ASN verification must confirm a failure before an incident opens. All probe nodes across Cloudflare and independent ASN sentinels are currently healthy and participating in active quorum.

Allowlist our probes

Our synthetic probes originate from Cloudflare Durable Objects. Because serverless edge runtimes egress via Cloudflare's shared global edge IP pool, IP allowlisting is neither deterministic nor secure. Instead, configure your WAF or reverse proxy to match on our cryptographic CF-Worker: pulseguard.io header and verified User-Agent.

WAF Allowlist Configuration Spec
# 1. Cloudflare WAF Custom Rule (Recommended)
# Our synthetic probes run as Cloudflare Durable Objects.
# Subrequests include an authentic, un-spoofable CF-Worker header.
#
# Match Expression:
(http.request.headers["cf-worker"][0] eq "pulseguard.io")
# Action: Skip / Bypass WAF & Rate Limiting

# 2. General WAF & Reverse Proxy Headers (Fastly / AWS WAF / Nginx)
CF-Worker: pulseguard.io
User-Agent: PulseGuard-Synthetic-Monitor/2.0 (+https://pulseguard.io/bot)
/api/locations

Coverage we don't have

South America, Africa, the Middle East, Oceania. Our probe regions run on Cloudflare's network, which does not currently place workloads in those regions β€” so we don't claim them.

What this means practically: a genuine outage still pages you, because your endpoint will fail from all seven regions regardless of where it's hosted. What we'd miss is a problem affecting only users in those regions β€” a SΓ£o Paulo CDN edge, an African transit route. If that's a real risk for your traffic, tell us; it moves our roadmap, and we'll say so here when it ships.

When a probe goes bad, we say so

A monitoring probe with a bad network path reports failures that aren't real β€” the industry's most common source of false alarms. Any probe with three or more state transitions in two hours is marked Flapping and automatically removed from quorum until it stabilises.

Every probe also reports on a heartbeat channel separate from its measurement path, so a probe that's blocked is distinguishable from one that's crashed. All of it is visible above, in real time, including when it's our fault.