Exposing more data via /metrics

[timp (Tim Portmann)]

Hello Everyone

I’m not sure whether this post belongs here, or into the help category. Please let me know should this be the wrong place, and I’ll make sure to move it.

When implementing monitoring for my caddy instances, I noticed that the three currently exposed metrics categories (via the /metrics endpoint) might not be enough for me. While combing through the forum and GitHub, I saw that this is a somewhat reoccurring discussion, which is why I am a bit confused about the current state of this topic. Are there any plans on including other categories to the /metrics endpoint (for example certificate specific data like number of certs issued, tls errors, etc.)? Is it thought that if people need more data specific to their use case, that they write a plugin for that or adapt their caddy build accordingly? In a discussion about this, it was mentioned that this cloud be done by monitoring events (Exposing TLS certificate metrics · Issue #1683 · caddyserver/caddy · GitHub). Is this still up to date, or should we approach gathering these metrics differently?

Appreciate your help, cheers

[matt (Matt Holt)]

/cc @hairyhenderson – might be able to answer!

[francislavoie]

The issue is Matt and I don’t use metrics, and we have a very surface-level understanding of how it works, so we have trouble maintaining it. The underlying libraries we’re using seem to have limitations with changing config after being loaded. So it’s a struggle. We wouldn’t mind help.