Test Execution Lifecycle

Every test run follows a deterministic lifecycle with clearly defined phases. Understanding this lifecycle helps you interpret test results, diagnose failures, and design effective test configurations.

Test Run Phases

Created → Queued → Dispatching → Running → Completed
                                    │
                                    ├── Registration
                                    ├── Call Setup
                                    ├── Media Active
                                    └── Teardown

Created

The test configuration is saved. No resources are allocated yet. The test run is visible in the dashboard with a "Created" status.

Queued

The platform evaluates worker capacity and queues the test run for execution. If sufficient workers are available, this phase is near-instantaneous. If workers are at capacity, the test waits in the queue until resources become available.

Dispatching

The platform sends the test configuration to the assigned workers. Each worker receives:

• The number of endpoints to create
• Registrar credentials and SIP configuration
• Media track settings (codecs, bitrate, DSCP)
• Scenario actions (hold, resume, DTMF, media updates)
• Encryption settings (TLS, SDES, DTLS)

For cross-group tests (Group A calls Group B), dispatch is phased:

1. Receiver endpoints are dispatched and registered first
2. The platform waits for receiver registration confirmation
3. Caller endpoints are dispatched with the resolved target URIs of the receivers

This phased dispatch ensures receivers are ready to accept calls before callers start dialing.

Running

Workers execute the test. Each endpoint follows its own lifecycle independently:

Registration Phase

Each endpoint sends a SIP REGISTER request to the configured registrar. The platform tracks:

• Registration success or failure
• Server-granted Expires value
• Time to register

If registration fails (timeout, 401 authentication failure, 403 forbidden), the endpoint transitions to a failed status and reports the error.

Call Setup Phase

After registration, caller endpoints send SIP INVITE requests. The platform measures:

• Time to Trying — INVITE to 100 Trying
• Time to Ringing — INVITE to 180 Ringing
• Post-Dial Delay (PDD) — INVITE to first ringback indication
• Call Setup Time — INVITE to 200 OK

If using 100rel/PRACK (RFC 3262), the provisional responses are acknowledged reliably.

Media Active Phase

After the call is established (200 OK + ACK), media flows between endpoints. This is where the bulk of metrics are collected:

• 150+ metrics per second per endpoint (quality, network, audio, video, jitter buffer)
• Scenario actions execute at their configured triggers (hold, resume, DTMF, media updates)
• Session timer refreshes occur at the configured interval
• DTMF digits are sent and received per scenario configuration

The media phase continues for the configured call duration.

Teardown Phase

At the end of the call duration, the endpoint sends a SIP BYE request to terminate the call. If session timers expire before the configured duration, the endpoint may terminate early (unless soft expiry is enabled).

After BYE, the endpoint sends a final metrics report and de-registers from the SIP registrar (Expires: 0).

Completed

All endpoints have finished their lifecycle. The test run transitions to "Completed" and full results are available in the dashboard.

Endpoint Lifecycle Timeline

Each endpoint produces a timeline of SIP events that you can inspect in the endpoint detail view:

INVITE → 100 Trying → 180 Ringing → 200 OK → ACK → First RTP → ... → BYE → 200 OK

The timeline shows absolute timestamps for each event, enabling precise measurement of signaling delays and identification of bottlenecks in call setup.

For cross-group tests, the paired endpoint's timeline is also visible, allowing side-by-side comparison of caller and callee perspectives.

Group Topology

The test run overview page shows the group topology diagram — a visual representation of which groups are calling which other groups. This includes:

• Group cards with endpoint count and codec summary
• Directional arrows showing caller → callee relationships
• Scenario action summaries per group

Failure Handling

Individual endpoint failures do not affect other endpoints in the test. The test run completes when all endpoints have either finished or failed.

Next Steps

• Creating a Test — Configure your first test
• Analyzing Results — Interpret test run data
• Metrics Collection — How metrics are captured per second