Set up initial infrastructure for Grafana metrics capture with docker monitoring stack

[kylehuntsman]

This sets up the initial infrastructure needed to capture boost and booster-http metrics via the docker monitoring stack.

• Modified the metrics exporter to work with any boost process outside of boostd, instead of just for boostd
  • Instantiates an exporter for booster-http to capture request counts and durations
• Dashboard for metrics are saved in /docker/monitoring/dashboards/exported_dashboard.json
  • The dashboard panels can still be modified in browser, but when saving you will be prompted to copy the JSON model of the dashboard to your clipboard. PRs will need to be made to modify the dashboard panels.
• Prometheus was told to watch the booster-http process via the /docker/monitoring/prometheus.yaml
• Additionally, adds a trace to booster-http's /piece/<piece-cid> endpoint

When running both the devnet and the monitoring docker stacks, the command docker network connect devnet prometheus must be ran after starting both stacks to ensure prometheus can actually talk to the processes running in the devnet stack.

Updates #772

[nonsense]

Does this work? I thought we need to have /etc/grafana/grafana.ini.

[kylehuntsman]

I'm fairly certain this is okay, but let me double check the tracing functionality

[kylehuntsman]

Yeah, this is fine. I discovered that the only things that are important are where the file is located and what file extension it has. As long as those two things are right, Grafana will read the file and determine what it is.