I can’t seem to figure out why the screen recording metadata returned by the /v2/quality/recordings endpoint shows a status of completed but lacks the media_url field when calls route through secondary BYOC trunks in the Asia/Singapore region.
Background
Managing 15 BYOC trunks across the Asia/Singapore region. The environment uses Genesys Cloud Architect flows to trigger screen recordings for QA evaluation via the /v2/quality/evaluations endpoint. Primary trunks handle 90% of traffic and record metadata correctly. Secondary trunks activate only during primary SIP registration drops or high-latency events.
Issue
When traffic fails over to secondary BYOC trunks, the screen recording initiation succeeds. The Architect flow logs confirm the recording start event. However, querying the recording status via /v2/quality/recordings/{recordingId} returns a 200 OK response with status: "completed". The critical media_url field is null or missing entirely. This prevents downstream QA tools from fetching the actual video file. Primary trunk recordings populate media_url correctly.
Troubleshooting
Verified SIP credentials for secondary trunks are valid and active. Confirmed outbound routing rules prioritize secondary trunks only when primary health checks fail. Checked the /v2/voice/trunks endpoint for registration status; secondary trunks show registered: true. Inspected Architect flow logs for errors during the recording start action; no errors logged. Tested with different agent workspaces; issue persists. Reviewed carrier-specific quirks for the secondary provider; no known restrictions on media streams. The latency spike observed in previous posts correlates with this metadata failure, but the direct cause remains unclear. Is there a known limitation with screen recording media URL generation during BYOC failover scenarios in Asia/Singapore? Or is this a configuration gap in the secondary trunk setup?