# Updated via Autotest Editor - 2026-05-25T21:52:12.017Z
# Source of truth: this YAML file

id: pbx-availability-check-automatic_failover_and_call_preservation
module: pbx
group: pbx-availability
title: Verify automatic failover preserves active calls (no drops, no audio interruption > 200ms)
description: Verify automatic failover preserves active calls (no drops, no audio interruption > 200ms)
priority: Medium
tags: []
references: []
dependencies:
  - label: Admin Portal
    value: https://pbx.local:4443 (configure via $BASE_URL)
  - label: Repository URL
    value: ${BASE_URL}/downloads or $REPO_URL
  - label: Protocol
    value: HTTPS + SIP (UDP/TCP/TLS) + RTP/SRTP
  - label: Load Gen
    value: SIPp or custom load generator + monitoring (Prometheus/Grafana recommended)
  - label: Credentials
    value: Admin user with sufficient privileges (portal + SSH where needed)
  - label: Test Data
    value: Clean test tenant or lab PBX instance (recommended isolated environment)
preconditions: []
postconditions: []
steps:
  - action: Open admin portal or repository URL in browser context
    expected: ''
  - action: Execute the main verification or configuration action described in the test
    expected: ''
  - action: Capture evidence (screenshot, log, API response)
    expected: ''
  - action: Assert expected outcome
    expected: ''
automation_status: manual
file: tests/pbx/availability.spec.ts
custom_fields:
  original_action: check
  manualReason: Requires active call sessions during failover and simultaneous call-state monitoring
automatable:
  - Most UI-driven configuration and verification steps via Playwright
notYet:
  - Requires active call sessions during failover and simultaneous call-state monitoring
  - Requires external SIP endpoints or physical phones for media/audio validation
  - Hypervisor-level operations (vCenter, libvirt, vSphere) not reachable from browser automation
  - Long-running load or HA failover scenarios need dedicated lab infrastructure