idea: communal status page, users make a merge request with info on what endpoints to monitor (with the option of specifying something like an email address for alerts) and then one gets a nice status page hosted off of their own infra, for free!
ofc mainly meant for simple checks like ICMP echo reply or checking for 200 ok from an unauthenticated http endpoint
thermia has thought about building something like this for a while now. i could be very simple it thinks
but it could also be over engineered like crazy lol
@thermia I once spent an entire night in a student pub (20:00 - 02:00) with a group huddled around a table trying to design a distributed health check system with failovers.
- Many nodes check many hosts (but not all, avoid DoS)
- uptime information from the network is freely shared and distributed (no hierarchy)
- option for new hosts to be introduced and shared from one node to others to maintain redundancy (discovery)
We concluded nothing and went home to continue the discussion on irc
@nyastrid@computerfairi.es that's really neat ^^