Concepts on ObserveRTC

Telemetry and glossary

Mon, 01 Jan 0001 00:00:00 +0000

Telemetry is the data you collect to understand the behavior of a running application. In ObserveRTC, it connects browser observations to endpoint and call-level analysis. Keep the observation separate from the conclusion you draw from it.

Connectivity, transport, and media quality

Mon, 01 Jan 0001 00:00:00 +0000

A call can connect successfully and still deliver poor media. Separate the stages to understand which measurements matter.

Dimension	Engineering question	Relevant evidence
Connectivity	Can the endpoints establish and maintain a usable path?	ICE gathering, connection state, selected candidates, DTLS state.
Transport quality	Is an established path delivering data well?	Loss, round-trip time, congestion, and path changes.
Pipeline disruption	Where does capture, encoding, delivery, decoding, or playout stop progressing?	Frame/counter progress and the relationship between adjacent stages.
Perceived quality	Does received media play continuously and with usable fidelity?	Freezes, concealment, playout behavior, and relevant application context.

Follow the symptom through the pipeline

For frozen video, first confirm whether bytes arrive, then whether frames decode, then whether frames render. This narrows the stage that lacks progress. Packet loss alone cannot establish whether the user sees a freeze, and successful ICE cannot establish that media is playing.

Detector evaluation and issue lifecycles

Mon, 01 Jan 0001 00:00:00 +0000

A detector checks a particular condition using available measurements and state. Some detectors report observations; others open an issue and later resolve it. Their output is evidence about that condition, not a universal diagnosis of the call.

Quality dimensions and scoring

Mon, 01 Jan 0001 00:00:00 +0000

A quality score summarizes measured penalties so you can spot affected endpoints and media components. Client Monitor reports values on a 0–5 scale. Use the score to choose where to investigate, then examine component reasons, issues, and metrics.