Due to our internal requirements, we would like to force healt_check_interval to much shorter time (e.g. 1s), but only for some limited time during upgrade of our backends. In other words we would like to let to know Caddy that our backend is (or not) working asap.
Is there any way how to do it? We tried to look for plugins, but no luck. We also looked into source code and although we don’t know about go, it looks like there is no chance how to configure it.
If you change the health_check_interval to 1s when you want the more frequent health checks, you can send caddy signal USR1 and it will reload and apply that config without dropping connections.
However, I’d be happy to implement a custom solution for you. We’ll be revealing a commercial support option later this year. Could I message you with the details for now?