Consolidate metric collection in orchestrator

[jpswinski]

Metric collection should not be maintained in the nodes because the nodes scale up and down and that information is lost. The metric collection should be consolidated in the orchestrator which lasts the lifetime of the cluster.

• Need metrics that count one per proxy request
• Metrics should be broken out by endpoint
• In the lock request there could be meta data that the orchestrator keeps track of (like region)
• Need a metric monitor that see the metric and posts to orchestrator - the orchestrator then just counts.
• Can the orchestrator keep track of unique ip addresses and maintain a metric that counts it
• Metric could should be removed/updated in the sliderule server code base

[jpswinski]

Metrics are now collected in the orchestrator (starting with v4.0.4). There are still some metrics that we might want to collect that aren't implemented yet

• unique ips
• region counts

But there is a robust set currently implemented with updated grafana pages. So new metrics will be handled by future GitHub issues.