feat(grafana): Dashboard for JVB and jicofo monitoring

[24kushang]

Dashboard for monitoring JVB and Jicofo services

Summary:

This PR concerns the monitoring dashboard for JVB & Jicofo services. It uses Prometheus as the data source for time-series data of various metrics required for monitoring JVB and Jicofo services.

Charts Overview:

• Health Checks / Up:
  • It checks JVB's health check metrics
  • Plot of time series for jvb_stress metrics
• JVB:
  • Plots for JVB endpoint counts for total, current, and active
  • Plots for endpoints sending audio and video
• Participants and Conferences:
  • Participants per video bridge
  • Largest Conference, Count of Conferences and participants
  • Count of both JVB and Jicofo threads
• JVM:
  • Time series on - GC and Open FD
• Packets:
  • Packets Sent and received
  • Relay Packets sent and received
  • Round Trip Time (RTT)

Screenshots of the Dashboards:

Looking forward to more insights for expanding the dashboard for better monitoring

---

Contributor: @24kushang

[bgrozev]

Review based on the talk we had:

• Dashboards don't load data until you click "run queries"
• Rename the "Health Checks / Up" section to "System"
• Rename "JVB Health Check" to "Healthy"
• Rename "Participants per bridge" to just "Participants". Add "Jicofo particiapnts" and "JVB Participants" as two separate lines
• Remove "Restart Requests Created" (don't need to graph any of the _created metrics)
• Move the graphs from the "JVB" section to the "Participants and conferences" section
• Remove "Total Participants", change the "Participants" gauges to show "JVB Participants" and "Jicofo Participants"
• Move "Count of Threads" under the "JVM" section, rename to "Thread Count"
• Packets section:
  • Rename to "Network"
  • Remove existing graphs
  • Add one graph for "Packet rate" and one graph for "Bitrate"
  • Each graph has 2 lines, one for "up" and one for "down"
  • Graph the change of the packets/bytes metric
  • Document the units (bytes/sec or bits/sec?)

[saghul]

What's the status of this?

[24kushang]

I would attach the screenshot of the updated dashboard in the comment below this. If there are any comments or reviews than we can discuss it later as well