Live Threat Map.
A world map of attack traffic built only from feeds that publish real observations, not animated guesses.
PublicMost attack maps animate invented traffic. This one draws four things that can each be traced to a named feed: the networks SANS DShield sensors report most, the ports they probe, the claims ransomware crews post on their leak sites, and the botnet command servers abuse.ch tracks. Newly exploited CVEs from the CISA KEV catalog sit alongside.
A small Node relay polls each feed on its own timer, keeps the last good copy when one fails, and serves a single cached summary, so the page makes one request and no feed sees visitor traffic. Source addresses are geolocated offline with DB-IP Lite and cut to their /24 before they leave the relay.
Two things are left out on purpose. Organisations named on leak sites are not repeated, because those posts are claims made by the attackers. And the map draws a path between two countries only when a feed reports both ends of an attack; a source country alone gets a mark, not an arrow to somewhere made up.
The relay reads the anonymised stats file from my SSH honeypot when it is given one, and refuses the generated sample files that share its format, so test data cannot be published as attacks.