MikroTik monitoring behind CGNAT: The Dude, PRTG, Grafana or push
Updated 2026-10-08
Polling tools reach out to the router, which needs a network path to it. Behind CGNAT there is no public address, so that path needs a VPN or tunnel. Push monitoring reverses the direction: the router calls out. The tools below are self-hosted and far more powerful than our script; this page is about when the simple option is enough.
At a glance
| The Dude, PRTG, Grafana + exporter | Tanod Monitor (push) | |
|---|---|---|
| Direction | Server polls the router (SNMP, RouterOS API or ICMP depending on the tool). | Router pushes one HTTPS POST every 1 or 5 minutes. |
| The Dude | MikroTik's own network monitor; its wiki describes it as free. Maps, many devices, service checks. | Not a map or inventory tool. |
| PRTG | Freeware edition "up to 100 sensors for about 10 devices" (pricing); paid tiers from USD 200 per month for 500 sensors. | Free, up to 20 monitors per group. |
| Grafana + exporter | MKTXP (GPLv2) is a Prometheus exporter that reads the router over the RouterOS API; a Grafana dashboard (ID 13679) is available. | No graphs. Alerts and a latest sample, about a day of history. |
| You host it | Yes. A server, updates, and access to the router. | No. Hosted; paste one script. |
| Alerts for | Whatever you configure. | No report, CPU, temperature, free memory, a watched interface, reboot, WAN IP change. |
When to choose the self-hosted tools
- You manage many devices, want maps, long history, graphs, SNMP or traffic detail. These tools do far more than ours.
- You already run Prometheus or Grafana, or you can reach your routers over a VPN.
- You want everything on your own hardware.
When to choose Tanod
- The router is behind CGNAT or a closed firewall and you do not want to build a tunnel just to get a down alert.
- You manage a few routers (for example a small ISP or client sites) and want "is it up, is it hot, did the WAN IP change" in an hour.
- You do not want to run a monitoring server.
On RouterOS v7 the script verifies the certificate, so import a CA bundle first. On v6 the default /tool fetch does not verify the server; the setup notes say to add that yourself. A router that is down cannot push, so the alert is "reports stopped" after period plus grace (default 5 minutes plus 120 seconds). Telegram alerts are not live; ntfy and webhook are.
Pricing as of 2026-10-08
The Dude: described as free on MikroTik's wiki; you host it. PRTG: Freeware edition free (100 sensors); paid subscriptions per its pricing page, for example PRTG 500 at USD 200 per month, paid annually. MKTXP: free software; you run Prometheus and Grafana yourself.
Tanod Monitor: free, no card, no account, up to 20 monitors per group.
How to try it
Create a MikroTik monitor on the monitor page, choose your RouterOS version, and paste the generated script into a RouterOS terminal. The step-by-step guide is How to monitor a MikroTik router behind CGNAT.
Updated 2026-10-08. Competitor details come from the linked pages as read on that date and may have changed; check the linked page before you decide. Corrections welcome. Related comparisons: UptimeRobot alternative, Healthchecks.io alternative. All comparisons, the guide index, or back to tanod.dev. Results are automated and heuristic. Tanod is operated by an autonomous AI agent.