Prevent re-definition of prometheus metrics

[blink1073]

Fixes #209

[blink1073]

Terminal cwd failure is unrelated.

[kevin-bates]

Thanks Steve. I haven't worked with Prometheus and know hardly anything about it (other than what I've just read glossing over some of the docs), but the comments and changes imply that the metrics are scoped at a broader level. Just wondering if we should adjust (namespace) the metric names in juptyer_server so that the notebook-specific metrics are not included should there be a notebook server running on the same node (or network for that matter - since it appears Prometheus performs distributed rollups). I'm assuming we'd want these separated.

As it says on the tin, re-defining these metrics results in an error.

Could you please provide a reference to the tin label? This may help my understanding here - thank you.

If what I said above is how things work (for the most part), but we do want stats cumulated across notebook and js instances, then these changes seem appropriate. My vote would be to keep these separated.

[blink1073]

Hi @kevin-bates, I haven't looked deeply into Prometheus either, I ran into this bug while working an Jupyter Telemetry, since I ended up importing the log.py file twice. I do think it makes sense to scope the parameters down per-server. I don't have the bandwidth to do much more than open this PR at the moment, I'm in the midst of a major refactor of the JupyterLab test suite.

[kevin-bates]

Looks good - addresses the immediate issue. We can revisit scoping via another issue/PR.