Could Your Home Router Be Your First Service Monitor?

If you work with services online, you’ve probably thought a lot about uptime from the server’s point of view. But recently, I found myself considering a more fundamental, almost primal form of monitoring—the little lights on my home router. That blink-and-you’ll-miss-it flicker of the WAN light, which had gone solid amber, was my first alert. Not PagerDuty, not a dashboard. Just a sad, static glow telling me my link to the outside world was dead. It got me thinking: what if our home networks are, in their crude way, the most relatable observability tools we have?

We design sophisticated health checks that probe database connections and validate API responses. We set complex latency budgets and trace requests across continents. Yet, the initial “is it working?” question for most end-users—and, in a way, for our own services—often boils down to a binary state visible on a cheap plastic box on a shelf. That router doesn’t measure p99 latency. It doesn’t differentiate between a DNS failure and a BGP route leak. It has one job: signal the presence or absence of a viable connection. Its “observability” is laughably shallow, but its signal is incredibly strong.

The Unfiltered End-User Signal

This is where the lesson lies. Our routers give us an unfiltered, end-of-the-line signal. They don’t care if our Kubernetes cluster is healthy or if our load balancer is passing probes. If the packet doesn’t make the final hop, the light changes. In our professional monitoring, we can create a bubble of perceived health. A service can pass all its internal health checks while being utterly unreachable from specific networks, behind a misconfigured firewall, or suffering from an ISP-specific routing issue. The router’s perspective—the “can I even talk to you?” check—is the most forgiving health check we often forget to implement from the outside in.

So, what does a “router light check” look like in practice for a service? It’s the synthetic check that runs from a residential ISP, not just a cloud provider. It’s the ping from a different continent that isn’t just measuring latency, but also reachability. It’s acknowledging that the network path is part of your service’s health, not just a transparent pipe you can ignore. When your monitoring can only see from inside your own infrastructure, you’re missing the entire last mile of the user’s experience.

My amber router light was a failure of my ISP, not my code. But it perfectly simulated a total service outage from my user’s location. My professional dashboards, hosted in the cloud, were blissfully unaware. That disconnect is the gap we need to mind. Reliability isn’t just about your servers being up; it’s about the path to them being clear. Sometimes, the most sophisticated observability platform needs to remember the simple question embodied by a blinking light on a plastic box: can you even see me?

Notes & further reading

A few pages I came back to while writing this: