What the best way to collect custom metrics of my hydra instanse?

[cat-in-vacuum]

Hello! I need a little help to reaching the observability of my hydra instance.
The current goal is to get more metrics than existing from the box like a prediction of expired tokens on the future period (e.t. in 1 hour or one day) , count of success/fail grants of refresh/access tokens, grouping request routes by response codes.
I looked a guide and looked source code but don't find anything regard to these things
For now it seems like I have only one approach - get such metrics from the database. But it looks like a dirty hack. Ideally, I would like to get it on the fly
Help me please to resolve the issue or share how you handle doing with it

[Benehiko]

Hi @cat-in-vacuum,

This looks similar to our Distributed Tracing support on Hydra. Of course it won't contain everything and would require you to pull data directly from the database if you need other custom metrics.

[vinckr]

Hey @cat-in-vacuum , just checking in if your problem still persists.
I will mark this as answered for now, but feel free to come back to the discussion any time.
Thanks a ton 🐝