Degraded status?

[nabeelio]

Hi all. Aside from up/down, is there a degraded status? Red, yellow and green?

[jemand771]

yes (soon) actually, but probably not in the way you think:

there's a pending PR that's almost good to go over at #257. it introduces different levels of "down-ness" depending on how many pings for a service have failed. (so e.g. only an hour of downtime isn't as bad as a fully day)

I assume you're asking more about something like "service still works but has increased latency" or some other kind of condition that doesn't result in a service straight up not responding. sadly, that isn't really possible with the way things are set up at the moment.

did you have anything particular in mind? we're always open to community submissions, although we [as maintainers] don't usually implement big features ourselves

[nabeelio]

Yup, that's right. I'll be having some synthetic monitors that look at several Prometheus metrics, and then derive a simple red/yellow/green for end users, for about 8 different stages of a pipeline

Is it a difficult change? I haven't looked at the source yet