Add simple Prometheus update metrics

[bbockelm]

We are observing some timeouts in the topology webapp logs around update intervals. Provide some clarity by adding a Prometheus endpoint and exporting the update timing.

[matyasselmeci]

LGTM though I don't have any experience with either prometheus-client and or DispatcherMiddleware. How do you test it?

Also, can you add a link to the /metrics endpoint to the front page (src/templates/homepage.html.j2, under "XML and other machine-readable data")

[matyasselmeci]

I can add it too.

[matyasselmeci]

I just realized this adds a dependency on prometheus-client to models.py. We can't do that -- models.py is also used by various scripts such as osg-scitokens-mapfile. Sorry, I have to revert.

[bbockelm]

@matyasselmeci - how should we resolve this? Create dummy objects if the import fails?

[matyasselmeci]

That's the approach I'm thinking of, too.

[brianhlin]

Create dummy objects if the import fails?

@matyasselmeci did you get around to this?

[matyasselmeci]

I didn't realize it was on my plate, so no.