Skip to main content

Ad-Flow Dashboard Guide

Ad-Flow comes with a pre-configured Grafana dashboard that provide real-time monitoring and insights into your video stream processing.

Accessing the Dashboard​

  1. Access Grafana using your instance's Public IPv4 DNS, or Elastic IP: 
    https://<your-instance-public-dns>:3000
  2. Log in with your Grafana credentials (defaults to: admin/admin)
  3. You will be prompted to modify the password on first log in
  4. Navigate to Dashboards → "Capture Metrics Dashboard"

Dashboard Sections​

Summary Section​

The summary section provides key metrics at a glance:

Ad Detection Statistics​

  • Instances - Ad Detection: Number of active ad detection instances
  • Total Ad Breaks: Count of ad breaks detected in the last hour
  • Total Ad Duration: Total duration of ads in minutes over the last hour
  • Ad Breaks Chart: Visual timeline of ad breaks per channel
  • Ad Break Duration Chart: Duration of ad breaks per channel in minutes

Delay-only Statistics​

  • Instances - Delay-only: Number of delay-only instances

Details Section​

Detailed metrics for system monitoring and troubleshooting:

Connection Health​

  • Disconnects: Tracks SRT connection disconnects
  • Capture disconnects
  • Upload disconnects
  • Segment Missing Packets: Shows segments with missing data
  • Discarded Bytes: Monitors data loss
  • Total discarded bytes
  • Discarded due to no upload connection
  • Discarded due to low upload speed

Performance Metrics​

  • Transfer Rates: Shows capture and upload rates in Mbps
  • Process Metrics:
  • Number of handles
  • Number of threads
  • CPU Usage: CPU time per second (1000ms indicates full core usage)
  • Memory Usage: Process memory usage in MB

Setting Up Alerts​

Grafana contains an alerting system. Consult Grafana's documentation for details.

Key Metrics to Monitor​

Configure alerts for:

  • Connection disconnects
  • Missing packets in segments
  • High data discard rates
  • Abnormal transfer rates
  • Resource utilization spikes

Performance Thresholds​

Set alerts for:

  • CPU usage trends
  • Memory usage patterns
  • Process handle counts
  • Thread count anomalies

Best Practices​

  1. Regular Monitoring
  • Monitor connection stability
  • Track ad detection performance
  • Review resource utilization trends
  1. Troubleshooting
  • Use disconnect metrics to identify connection issues
  • Monitor discarded bytes for upload problems
  • Review process metrics for system health
  1. Dashboard Management
  • Default time range is set to last 1 hour
  • Use the time picker to adjust the viewing window
  • Data refreshes automatically