The ping reboot cycle of death is what I call the most common self-inflicted outage in M2M. A router is given a ping target it can never reach. Every check fails, so it reboots. It comes back, connects perfectly, fails the check again and reboots again. It never stays up long enough for anyone to work out why.
The cycle of death at a glance
| Symptom | Router online for a few minutes at a time, then gone, all day |
| Usual cause | Ping target unreachable by design, for example 8.8.8.8 on a closed private APN |
| Looks like | Bad coverage, a faulty SIM or a faulty router |
| Fix | A target the router can actually reach, and a gentler first action |
| Prevention | Test the target by hand before enabling the rule |
How it happens
Most often, a router configured for an internet SIM is reused with a private APN SIM, or a configuration template with 8.8.8.8 in it is pushed to a fleet on a closed network. Sometimes the target is a company server that was moved, or a firewall that stopped answering ping. The router is fine. The connection is fine. The yardstick is wrong.
It is not a new problem. Teltonika’s 2012 RUT500 manual warned that if your server does not answer pings, the router keeps rebooting. MikroTik’s documentation says the same about its ping watchdog: with an unreachable address and default timings, the router reboots roughly every six minutes.
How to recognise it
- The router is reachable for a few minutes, then drops, on a regular rhythm.
- The uptime counter never gets past the length of one cycle.
- Signal and registration look healthy every time you catch it online.
- The event log, if it survives the reboot, shows ping reboot actions over and over.
- Swapping the SIM or the router makes no difference, because the configuration goes with it.
How to break it remotely
- Use the window. Each cycle gives you a few minutes online. Log in during one and switch the ping reboot rule off, or change the target.
- Use SMS if you have it. On a SIM with SMS, router SMS commands or a management platform can change the configuration without waiting for a window.
- Fix the target. Ping something the router can reach. On a closed APN, that is an address inside the APN. See which IP address to ping.
- Soften the action. Restart the mobile connection first rather than rebooting the router. A gentler first action does less damage while you work it out.
How to never see it
Test the target by hand from the router before you enable the rule, through the interface the rule will use. Then test the rule itself by pointing it at an address that will not answer, and watching it act once. The recovery and reboot loop calculator shows how often a bad target would reboot your router with your timings, which is a sobering number to look at before you deploy.
The wider story, including where the warning first appeared, is in the history of ping reboot.
Questions
Why does my router reboot every few minutes?
Usually because ping reboot is targeting an address the router cannot reach, such as 8.8.8.8 on a closed private APN. Every check fails, so it reboots in a loop.
How do I stop a reboot loop remotely?
Log in during one of the online windows and switch the ping reboot rule off or change its target. SMS commands can help if the SIM supports SMS.
How do I prevent the cycle of death?
Ping a target the router can actually reach, test it by hand before enabling the rule, and use restart mobile connection as the first action.