Network monitoring for ISPs
Maps your NOC can read at a glance, alerts the right people actually answer, and every remote site in one standalone system. A modern replacement for The Dude.
Runs on one VM · Prometheus probing · No outside services
Why we rebuilt it
The Dude still holds the shape of many ISP networks: maps, devices and the parents between them. What it no longer holds is trust. Maps were drawn for a different decade; a failed parent turns forty devices red at once; the NOC keeps three tools open to answer one question.
Dr Monitor was rebuilt from that experience — the Dude's model, with confirmed status, alerting people act on, VPN sites, and the routine RouterOS work in the same place.
The map's job is to show the blast radius. The alert's job is to wake one person.
The product
Plain names, because a NOC should not need a glossary. Each module is a page in its own right.
Status that shows the cause, not just the symptom — on maps your team already knows.
Read moreEmail, Telegram, WhatsApp, SMS and phone calls, with acknowledgement and escalation.
Read moreWireGuard and L2TP/IPsec tunnels, site alerting and agents — monitored like local kit.
Read moreSNMP health, interfaces and reports: more than a ping, from the hardware itself.
Read moreVersion checks, staged updates, config backups and a read-only RouterOS view.
Read moreIXP Manager, UniFi, CAPsMAN and remote agents — optional, read-only connectors.
Read moreHow it works
The short version. The technical walkthrough is a page of its own.
Prometheus and blackbox_exporter check every address on its own schedule — ICMP or TCP, with per-profile intervals, timeouts and failure counts.
A failed probe is not an outage until the rules confirm it. Parents, tunnels and maintenance explain the rest, so the map shows cause and the alert stays quiet.
One engine sends the message, reminders, escalation and recovery — and stops when someone acknowledges, including from a phone keypress.
If almost every probe fails at once while Prometheus still reaches the probing engine, the monitor can now pause per-device messages, send one MONITORING PATH FAILURE message instead, and release held alerts in grouped messages when it clears. It starts in shadow mode so its judgement can be reviewed first.
Member traffic from one or more IXP Managers, with current/average/peak readings, a 24-hour or 7-day chart, and warn/peak alerts on the busier direction.
Gauges, graphs and readouts on any map for CPU, memory, temperature, interface traffic and more — each with its own warn/peak limits, reminders, acknowledgement and quiet hours.
Pricing
Final pricing and the free-edition limits are being settled before launch; the shape is decided.
Free forever
For a small ISP, a lab or a school. Core monitoring, maps and alerts without a key or registration. Limits will be announced before launch.
Edition detailsPer device band
For production fleets: larger device counts, advanced modules and the support that goes with them. Pricing to be announced.
Edition detailsEvery install starts with a 30-day full-feature trial. Monitoring never stops when a licence lapses — it softens.
See it for yourself
Made-up network, real probing, guest read-only sign-in. No sales form in the way.