Details for the monitoring / metrics section

[mmattel]

References: #695 ([5.0] Add metrics to the proxy service description)

While working on the referenced PR, I identified that we have metrics not only in the proxy but also in other services but only proxy currently has a description. This leads to the point that it is much better to have an own monitoring section and describe all stuff there and reference in the services affected that new monitoring section. This makes such a documentation future proof and readable.

The following services have a metrics implementation:

grep -rn metrics.Metrics

web
thumbnails
webfinger
eventhistory
store
ocs
idp
graph
webdav
settings
search
userlog
invitations
proxy

We also need to add the metrics requirement part of the new service documentation in the dev docs. (done see referenced PR below)

Questions:

• How does one know which services have a metrics endpoint (except parsing the code)
• Not all services do have a metrics endpoint which is most likely ok, but do all which should have one?
• Are the results queried on the endpoint on each service identical to that what we have in proxy or are they differnt?

[mmattel]

Closing, all services where metrics are implemented are gathered via the proxy service.

[mmattel]

Reopening as we have to think about what to do in Standalone Mode (vs single binary).