There’s a draft RFC proposing a standard format for health/liveness/service-status/whatever-term-you-like checks. I’ve implemented that format a couple times, and generally liked it, and it’s always nice to have an external standard so you don’t have to debate how to do it.
This is great, thanks for sharing! Looks like it’s expired and archived, but still good to know it exists. I may end up using some of the guidelines there as well, thanks!
I’m personally a fan of adding server uptime and/or start-time to healthcheck handlers, and/or the git hash of the running instance. There are some reasons not to do so, mostly security related, but on the balance having that extra information can help as a sanity check.