Does anyone know why our predictive dialing campaigns are experiencing a sudden spike in 480 Temporarily Unavailable responses specifically on our Singapore-based BYOC trunks? We are running Genesys Cloud version 2024-09 with Architect flow version 12.4. The outbound dialing strategy is set to predictive with a 1.5 efficiency ratio, yet the disconnect rate has jumped from 2% to 15% over the last 48 hours. This issue is isolated to the SIP trunks managed by our primary APAC carrier, while the US and EU routes remain stable with sub-1% disconnect rates. The SIP traces show that the INVITE is accepted by the carrier, but the subsequent 183 Session Progress message is delayed by approximately 2 seconds before the 480 response is generated, causing the platform to mark the call as a disconnect rather than a busy or no-answer.
We have verified that the trunk capacity is not exceeded and that the carrier’s failover logic is not triggering prematurely. The SIP credentials and outbound routing rules were last updated three weeks ago and have functioned without incident since then. I suspect there might be a change in the carrier’s SIP signaling handling or a mismatch in the SDP negotiation parameters within the Genesys Cloud outbound configuration. Has anyone encountered similar latency issues with specific carriers in the APAC region, or are there specific diagnostic logs we should request from the carrier to identify the root cause of this 480 response delay?