Adaptive Monitoring

Adaptive Monitoring reduces compute costs by scaling monitoring intensity per stream, without losing error detection.
What is it?

Adaptive Monitoring lets you dynamically run each channel in Full, Light, or Extra Light mode so you keep full probing and alarms while cutting video decoding and CPU usage

What can I do with it?

With Adaptive Monitoring, you can manage more streams with less CPU by moniotirng in Light/Extra Light modes (full probing and alarms, reduced decoding). Streams with exceeded thresholds will automatically switch to full monitoring.

What are the benefits?

Adaptive Monitoring lowers costs by keeping full monitoring and alarms on every service while dramatically cutting CPU usage, so you can scale more channels on the same hardware and only spend full resources where they’re really needed.

Adaptive Monitoring lets operators define monitoring intensity per source, optimizing CPU resources while keeping full visibility. Each stream can run in one of three modes:

  • Full Mode: continuous decode and multiview visualization.
  • Light Mode: partial decode every ~3 seconds (~30% CPU).
  • Extra-Light Mode: minimal decode every ~5 seconds (~20% CPU).

Regardless of mode, all streams are continuously probed by TAG’s 500+ video, audio, and metadata checks. If an error is detected, the stream is automatically promoted to Full Mode and visualized for operator attention.

This approach allows monitoring of far more channels with the same infrastructure, cutting compute requirements while maintaining full error coverage. Adaptive Monitoring supports all formats and workflows- from ST 2110 and MPEG-TS to HLS, DASH, Zixi, and SRT- across on-premises, cloud, and hybrid environments.

With built-in API and dashboard integration (Grafana, Kibana, NMS), operators gain complete situational awareness. Combined with tools like Penalty Box and Smart Recording, Adaptive Monitoring helps teams reduce operating costs, accelerate troubleshooting, and scale monitoring without compromise.

FAQ

What’s the difference between Full, Light, and Extra Light modes?

Full Monitoring - Fully decodes every video frame, can be displayed in real time on a mosaic tile, and performs full probing, analysis, and alarms (typical for premium or problematic channels).

Light Monitoring - Fully probes and alarms transport and audio but only decodes with ~3 second gaps, enough to verify video presence but not real-time display; uses about ⅓ of Full CPU.

Extra Light Monitoring - Same behavior as Light but with longer gaps (~5 seconds) between decodes, using about ⅕ of Full CPU.

Do I lose any alarms or error detection in Light / Extra Light mode?

No. In both Light and Extra Light modes the system continues full-time probing, analysis, and alarming on all aspects of the source-including transport, timing, and audio-while video is decoded intermittently, so you do not sacrifice error detection when cutting CPU usage.

How does Adaptive Monitoring help me reduce costs and scale?

By running lower‑priority or stable OTT/compressed channels in Light (0.33 hardware usage) or Extra Light (0.2 hardware usage) while keeping full alarms, you can reduce system resource utilization by up to ~80%, monitor many more services on the same hardware, and only pay the CPU cost of Full decoding for the streams that actually need to be seen live on the mosaic.

Subscribe to TAG news

Stay in the loop of product updates, company announcements and more