Monitoring a Self-Hosted HealthChecks.io Instance (without modifying containers)

Introduction

In the exciting (but sometimes tricky) world of self-hosting, you might find yourself wanting to monitor the health of your various services. A popular choice for this is HealthChecks.io. But what happens if your self-hosted HealthChecks.io instance goes down? How do you ensure you get notified?

This post details the solution I found during my own self-hosting journey, allowing you to monitor your self-hosted HealthChecks.io using the hosted service at HealthChecks.io – all without modifying any containers!

Why Self-Host HealthChecks.io?

The free and supporter plans of HealthChecks.io offer a generous 20 job limit, which might be sufficient for some. However, if you anticipate needing more jobs in the future, self-hosting allows for easier scaling without migration headaches. Additionally, setting it up with tools like Portainer can be quite straightforward.

The challenge arises when you consider that a downed self-hosted HealthChecks.io container would leave you blind to notifications. Here's how to use the hosted service to monitor your on-premise instance.

Beyond Cron and Curl:

While a simple cron-based curl command might seem like the answer, the official HealthChecks.io Docker image doesn't include curl, wget, or nc. Modifying the container itself is not ideal due to potential update/restart conflicts.

The Solution: Using echo and /dev/tcp

Research revealed a clever way to send HTTP requests using just echo and redirection to the /dev/tcp file descriptor. This allows you to construct an HTTP request and "send" it without additional tools.

Here's the command that sends a GET request to the HealthChecks.io server:

echo -e "GET /<uuid> HTTP/1.1\nHost: hc-ping.com\n\n" > /dev/tcp/hc-ping.com/80

(Replace <uuid> with your actual HealthChecks.io ping URL)

Combining Forces: Cron and Docker Exec

We can combine this technique with docker exec to schedule the request execution. Here's the final crontab entry:

*/15 * * * * docker exec HealthChecks.io bash -c 'echo -e "GET /<uuid> HTTP/1.1\nHost: hc-ping.com\n\n" > /dev/tcp/hc-ping.com/80'

Important Note: Ensure that the user to whose crontab you add this to has permission to run docker

Conclusion

By utilizing this approach, you can effectively monitor your self-hosted HealthChecks.io instance using the hosted service, ensuring you stay informed about its health even if it encounters issues.