Improve notifications metrics dashboard

• Make endpointName a drop-down option in Grafana with default value "ALL"

• Overview Row

  • Rename stat-panel "Pending events" to "Queue size"

• Events Delivery Row

  • Move "Events queued per second" to this row
  • Update "Events queued per second" description to: "The per-second rate of all events generated by the registry and scheduled for sending."
  • Add new metric: "Events delivered per second"
  • Move "Event delivery failure rate" to this row
  • Rename "Events per second (by Status Code)" to "Event delivery failure rate (by Status Code)"
  • Move "Event delivery error rate" to this row
  • Fix "Event delivery error rate" metrics to always show at least zero (currently shows zero or very low)
  • Create new metric: "Event delivery success rate"
    • Calculate as percentage of delivered events: successful / (successful + failed + err)

• Queue Row

  • Move graph "Pending events" to this row
  • Rename "Pending events" graph to "Queue size [whole cluster]"
  • Add min and max queue sizes (per pod) to show min/max queues on pods