- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
cross-posted from: https://discuss.ntfy.sh/post/3200
I use ntfy (on another instance) + healthchecks.io to wake me up at night when ntfy.sh is down (crazy inception, right?). It’s the “poor man’s PagerDuty” if you will. It works amazingly.
Here’s how I set it up:
- I signed up healthchecks.io (free plan) and configured a project for “ntfy.sh” with a “ntfy” integration, i.e. publish to
ntfy.example.com/<secret>
with max priority- I have two different hosts execute small “integration ntfy.sh tests” and only ping healthchecks.io if they succeed. If they don’t healthchecks.io will publish to
ntfy.example.com/<secret>
- In the ntfy Android app, I subscribe to
ntfy.example.com/<secret>
, enable “Keep alerting on highest priority”, and make it override DND (do not disturb) for this topic.Now when ntfy.sh goes down, the integration tests in the cronjobs will fail, and so healthchecks.io will not be pinged, which will trigger it to publish to
ntfy.example.com/<secret>
and let my phone consistently ring until I acknowledge it.
(Disclamer: I am the maintainer of ntfy. Hope posting this is fine. Happy to answer questions; I also have a brand new ntfy community, feel free to join)
Except that you can selfhost both healthchecks.io and ntfy.sh; and you can’t selfhost reddit or any of the services they advertise, haha
it’s true :3 ntfy has been on my to-do to test out in my home lab. thank you for your service!
I thought the post was timely since the alert triggered today. The ntfy server was DDoS’d today and I got alerted pretty much instantly. It was quite nice.
must feel extra godly to see all your hard work perform exactly as hoped 💙