CONTINUOUS INTERNET TELEMETRY24H DRIFT3,087 material changesacross 2,839 domains · -311 vs yesterdayDNS DRIFT15 domains changed DNS providertop destination cloudflare.com · -12 vs yesterdayEMAIL DRIFT5 domains switched email providertop destination google.com · -1 vs yesterdayCERT DRIFT5 domains switched issuing CA24h · -2 vs yesterdayNOW3,322 curated domains not answering-6,396 vs yesterdayERRORS10,913 responded with an errorlast probe · 5xx / 404 / 429 / 40324H DRIFT3,087 material changesacross 2,839 domains · -311 vs yesterdayDNS DRIFT15 domains changed DNS providertop destination cloudflare.com · -12 vs yesterdayEMAIL DRIFT5 domains switched email providertop destination google.com · -1 vs yesterdayCERT DRIFT5 domains switched issuing CA24h · -2 vs yesterdayNOW3,322 curated domains not answering-6,396 vs yesterdayERRORS10,913 responded with an errorlast probe · 5xx / 404 / 429 / 403

Registered here. Served from there.

We hold both halves of a domain's geography: the country it was registered from, and the country it is actually served from. The interesting part is not either map. It is where they disagree.

The country the network serving the domain is registered in (from the ASN).

Computed on 18,825 of 49,853 active domains (37.8% coverage). The rest is unknown, not zero: an ASN or RDAP lookup that has not succeeded yet is an absence of evidence, never evidence of absence. Coverage climbs as enrichment sweeps the catalog.

🇺🇸 US 15,14280.4% 🇩🇪 DE 5102.7% 🇳🇱 NL 4182.2% 🇷🇺 RU 3291.7% 🇨🇳 CN 2921.6% 🇨🇦 CA 2811.5% 🇫🇷 FR 2791.5% 🇬🇧 GB 2301.2% 🇸🇬 SG 1300.7% 🇸🇪 SE 990.5% 🇯🇵 JP 770.4% 🇰🇷 KR 650.3% 🇮🇳 IN 530.3% 🇨🇿 CZ 510.3% 🇪🇸 ES 510.3% 🇧🇷 BR 490.3% 🇦🇺 AU 440.2% 🇨🇭 CH 430.2% 🇮🇱 IL 410.2% 🇵🇱 PL 410.2% 🇦🇪 AE 350.2% 🇹🇷 TR 350.2% 🇮🇹 IT 310.2% 🇻🇬 VG 270.1% 🇧🇬 BG 260.1% 🇱🇺 LU 260.1% 🇦🇹 AT 230.1% 🇮🇷 IR 220.1% 🇹🇼 TW 220.1% 🇳🇴 NO 200.1%

Country grain, on purpose. We do not hold lat/lon and we will not invent it: IP-to-city geolocation is a licensed dataset and it is routinely wrong for anycast and CDN addresses, which describes most of this catalog. A pin on a building would look more impressive and mean less. Every figure above comes from a signed observation you can re-verify yourself.