Observing a persistent issue with predictive outbound campaigns on the AP-1 region BYOC trunk. When concurrent call volume exceeds 400 DSR, the carrier gateway begins returning SIP 486 Busy Here responses for approximately 40% of attempted connections. The trunk registration remains stable, and latency metrics to the carrier edge are within acceptable limits (sub-50ms).
The Architect flow logs indicate that the dial-out action succeeds, but the subsequent SIP INVITE fails at the carrier level. This behavior suggests a potential mismatch in how the Genesys Cloud platform handles early media or SDP negotiation under high load compared to lower volume scenarios.
Has anyone encountered similar SIP 486 spikes on AP-1 BYOC trunks during peak predictive routing periods? Are there specific trunk configuration parameters, such as max-concurrent-calls or SIP timeout settings, that need adjustment to accommodate higher burst capacities without triggering carrier-side rejection? The carrier support team claims their gateway is functioning correctly and points to potential formatting issues in the initial SIP request.