Skip to main content

Ad-Flow Troubleshooting Guide

This guide helps diagnose and resolve common issues encountered when running the Ad-Flow application, a Docker container deployed on EC2 that captures and re-streams video with SCTE-35 ad markers.

Quick Diagnostics​

Check System Status​

Check CloudWatch Logs

Check Grafana Dashboard for any anomaly

Common Issues and Solutions​

General Issues​

Cannot Connect to Grafana​

Symptoms:

  • Unable to access Grafana dashboard remotely.

Possible Causes:

  1. Security Group does not allow inbound traffic for TCP port 3000.
  2. Use incorrect IP address

Solutions:

  1. Double-check the CloudFormation cluster template.
    • Make sure there is an Inbound Rule with:
      • Protocol: TCP
      • Port Range: 3000
      • Source: 0.0.0.0/0 (or restrict to specific IPs for security).
  2. Access Grafana:
    • Use the Public IPv4 DNS of the EC2 instance in your browser, or its Elastic IP (e.g., https://Public_IPv4_DNS:30000).
    • make sure to use 'https' prefix, and not 'http' prefix
    • Accept the security risk from the self-signed certificate.
  3. Verify Docker logs in CloudWatch

Ad-Flow Container Not Running​

Symptoms:

  • Dashboard does not show any Ad-Flow instance

Possible Causes:

  1. Incorrect Configuration

Solutions:

  1. Check the Ad-Flow CloudWatch logs
  2. Check other logs for the other containers running in the same ECS cluster

Stream Reception Issues​

No Incoming RTP Stream​

Symptoms:

  • Zero receiving rate in Grafana.
  • No stream data in logs.

Possible Causes:

  1. Security Group missing inbound rules for the RTP port.
  2. Incorrect configuration of IP, ports, or protocol in the Docker Compose file.
  3. Issues with the RTP sender setup.

Solutions:

  1. Make sure your CloudFormation template contains a Security Group inbound rule:
    • Type: Custom UDP
    • Port Range: INPUT_PORT (e.g., 5004)
    • Source: 0.0.0.0/0 (or restrict to specific IPs).
  2. Verify Ad-Flow container settings:
    • INPUT_IP: Should be set to 127.0.0.1.
    • INPUT_PORT: Matches the port used by the RTP sender.
  3. Confirm RTP sender setup:
    • Ensure the sender is streaming to the Elastic IP of the EC2 instance on the correct port.
    • Check sender logs for any errors.
  4. Check CloudWatch logs for errors

No Outgoing RTP Stream​

Symptoms:

  • No outgoing RTP stream detected by the video encoder.
  • Sending rate is 0 on the dashboard.

Possible Causes:

  1. Incorrect output IP or port configuration.
  2. Firewall or network issues blocking outgoing traffic.

Solutions:

  1. Verify Ad-Flow container settings:
    • OUTPUT_IP: Set to the public IP of the RTP receiver.
    • OUTPUT_PORT: Matches the input port of the RTP receiver.
  2. Ensure the RTP receiver is correctly configured to accept the stream from the EC2 instance.
  3. Check CloudWatch logs for errors

SRT Troubleshooting​

No Incoming SRT Stream​

Symptoms:

  • Zero receiving rate in Grafana.
  • No incoming SRT stream data.

Possible Causes:

  1. Incorrect sender or container configuration.
  2. Firewall issues.

Solutions:

  1. Verify sender configuration:
    • Sender must be in Listener Mode.
    • Stream to 0.0.0.0:<INPUT_PORT>.
  2. Verify container configuration:
    • INPUT_IP: Set to the IP of the sender.
    • STREAM_PROTOCOL: Set to srt.
    • INPUT_PORT: Matches the port used by the sender.
  3. Check sender logs for connection errors.

No Outgoing SRT Stream​

Symptoms:

  • No outgoing SRT stream detected by the receiver.
  • Sending rate is 0 on the dashboard.

Possible Causes:

  1. Incorrect receiver or container configuration.
  2. Firewall or network issues.

Solutions:

  1. Verify receiver configuration:
    • Receiver must be in Sender Mode.
    • Stream from <OUTPUT_IP>:<OUTPUT_PORT>.
  2. Verify container configuration:
    • OUTPUT_IP: Set to 0.0.0.0.
    • OUTPUT_PORT: Matches the receiver's input port.
  3. Check receiver logs for connection errors.

SCTE-35 Marker Issues​

Markers Not Being Inserted​

Symptoms:

  • No markers in output stream.
  • Missing ad break signals.

Possible Causes:

  1. Incorrect or incomplete SCTE-35 analysis.
  2. Configuration or buffering issues.

Solutions:

  1. Check buffer settings:
    • Ensure DELAY_SECONDS exceeds longest ad break.
    • Verify stream analysis is completing.
  2. Validate stream format:
    • Confirm PAT/PMT structure.
    • Check for existing markers.
  3. Check CloudWatch logs for errors

Advanced Troubleshooting​

Preventive Measures​

  1. Regular health checks:

    • Monitor Grafana dashboards.
    • Review log patterns.
    • Track resource utilization.

  2. Backup procedures:

    • Maintain configuration backups.
    • Document custom settings.

Getting Support​

If issues persist:

  1. Collect diagnostic information:
    • Container logs.
    • CloudFormation templates used for deployment.
    • Grafana screenshots.
  2. Document the issue:
    • Exact symptoms.
    • Timing of occurrence.
    • Recent changes.
  3. Contact support with collected information.