Abuse RadarThreat Intelligence
Fleet

Sensors & collectors

10/10 healthy · 1 retired Connecting

Abuse Radar ingests from independent collectors. Sensor names here are anonymised labels — hostnames, addresses and log contents are never published.

Active sensors101 retired
Healthy10heartbeat arriving on schedule
Needs attention0missed check-in or never reported
Observing abuse6reported within their usual gap
Roster

Collector fleet

ordered by attention, then volume
Collector sensors and their reporting state, ordered with any sensor needing attention first
SensorHealthNew indicators / 14d ObservationsLast check-inLast observation
Leak Trap Sensor 02Leak Trap SensorHealthy11,9652026-10-09 18:55 GMT+81 minute ago2026-10-09 18:40 GMT+816 minutes agousually within 1.3 h
Web Sensor 03Web SensorHealthy5,8592026-10-09 18:53 GMT+83 minutes ago2026-10-09 17:37 GMT+81 hour agousually within 7.5 h
Web Sensor 02Web SensorHealthy4,2232026-10-09 18:53 GMT+83 minutes ago2026-10-09 13:01 GMT+85 hours agousually within 7.8 h
Web Sensor 01Web SensorHealthy2,3852026-10-09 18:55 GMT+81 minute ago2026-10-09 09:37 GMT+89 hours agousually within 10.5 h
SSH Sensor 01SSH SensorHealthy7292026-10-09 18:56 GMT+8just now2026-09-14 13:00 GMT+825 days agousually within 28.0 h
Mail Sensor 02Mail SensorHealthy5952026-10-09 18:55 GMT+81 minute ago2026-10-09 09:50 GMT+89 hours agousually within 6.1 d
SSH Sensor 02SSH SensorHealthy2922026-10-09 18:53 GMT+83 minutes ago2026-09-14 08:46 GMT+825 days agousually within 30.5 h
SSH Sensor 04SSH SensorHealthy2362026-10-09 18:53 GMT+83 minutes ago2026-09-17 00:11 GMT+822 days agousually within 32.8 h
SSH Sensor 03SSH SensorHealthy1982026-10-09 18:53 GMT+83 minutes ago2026-09-14 11:42 GMT+825 days agousually within 35.9 h
Mail Sensor 01Mail SensorHealthy1382026-10-09 18:55 GMT+81 minute ago2026-10-07 18:50 GMT+82 days agousually within 5.8 d
Leak Trap Sensor 01 (retired)Leak Trap SensorRetired20—2026-05-21 07:39 GMT+84 months agousually within 1.2 h
Each trace counts indicators this sensor observed for the first time on each of the last 14 days; a flat trace means no new indicators, not no traffic.
Reference

What the states mean

Health is collector liveness, proven by a heartbeat the collector sends every 5 minutes whether or not it has anything to report. Observation activity is reported separately and judged against each sensor’s own historical cadence — a leak trap that reports every few minutes and a mail sensor whose typical gap is days cannot share one rule. A collector that has seen no abuse is not unhealthy.

HealthyHeartbeat received within the grace window (4 missed 5-minute cycles = 20 minutes)
OverdueExpected heartbeat has not arrived — the collector, its host or its network needs checking
No heartbeatCollector predates heartbeat support. Liveness is unproven; observation cadence is used only to surface one that has clearly stopped
PendingKey registered, nothing received yet
RetiredKey disabled; ingest is rejected. The record is kept so historical observations stay attributable
Collectors check in with POST /api/v2/sensors/heartbeat, which updates last_seen_at only. Abuse observations update last_used_at. The two are never conflated, so a host running several sensors must check in once per identity.