I'm setting up simple uptime monitoring and prefer Healthchecks.io, but I'm still figuring out how to monitor appliance-type devices that don't have a normal operating system or an agent I can install. Ideally, I'd like a monitoring machine to ping each appliance and report the result to Healthchecks. Some of the devices expose web interfaces, and SNMP might also be available. For those using Healthchecks, what approach works best for monitoring devices like this?
3 Answers
A basic ping only confirms that the device’s network stack is responding; it doesn’t prove the actual application or service is healthy. If an appliance has a web management page, checking that endpoint is usually a better test. SNMP can also provide useful health and status information if the device supports it.
Keep in mind that some devices don’t answer ICMP ping at all, and others may only respond while a particular network service is running. A failed ping can therefore mean either the device is down or simply that ping isn’t supported.
For this use case, a monitoring tool that actively performs checks may be simpler. Uptime Kuma can test availability directly, and its Push monitor can also provide a Healthchecks-style workflow. With Healthchecks, failures in the checker script, scheduler, monitoring host, or internet connection can all look like the appliance failed, so you’ll need to account for those extra dependencies.
That’s a good point. I may run both tools in a test VM and compare which one is easier to manage for these appliances.
Healthchecks is mainly a receive-only service, so it won’t directly go out and ping an appliance. You’d need a separate machine running a script or scheduled job that checks the device and then sends a success or failure ping to Healthchecks. A systemd timer, cron job, or similar scheduler would work.
That makes sense. I’ve been reading more and it looks like a local script that checks the device and then reports to Healthchecks is probably the route I’d need.
Most of my appliances appear to have web interfaces, including an Avaya IP Office appliance, so I’ll probably start by checking those before looking into SNMP.