Trying to understand why specific Architect flows hang during trunk failover events.
Context:
Managing 15 BYOC trunks in the Asia/Singapore region. The primary carriers handle load without issue. However, when traffic shifts to secondary carriers due to failover logic, the IVR experience degrades significantly. The SIP registration remains stable on the secondary trunks, but the application layer seems to drop connections. Specifically, the flow enters a ‘Play Prompt’ node, but the audio never streams. The trace shows a 200 OK from the media server, yet the client receives no RTP. This correlates with a delay in the SIP 183 Session Progress message. The environment uses Genesys Cloud standard release. No firewall changes were made.
Question:
Is there a known latency threshold in the media path that causes timeouts when switching from primary to secondary BYOC carriers? The secondary carriers have higher latency (approx 150ms RTT) compared to primary (approx 50ms). Should I adjust the SIP timer settings or the media server configuration to accommodate this variance? Any insights on configuring the failover logic to handle higher latency carriers without dropping the media stream?