Facing intermittent SIP 408 Request Timeout errors on our primary AP-1 BYOC trunk when outbound call volume spikes above 500 DSR. The trunks remain registered, and latency to the carrier gateway stays under 80ms, so network jitter does not appear to be the culprit.
The issue manifests specifically within the Architect outbound flow during the ‘Connect to Customer’ block. Internal Genesys logs indicate the INVITE is sent successfully, but the carrier never responds with a 100 Trying or final response within the 15-second timeout window. Lowering the DSR to 200 resolves the issue, which suggests a stateful inspection limit or a rate-limiting policy on the carrier side rather than a Genesys platform bug.
Has anyone encountered similar behavior with specific APAC carriers where high concurrency triggers silent drops? We have verified the SIP headers are compliant with RFC 3261. Are there specific BYOC trunk settings or keep-alive intervals that can be adjusted to mitigate this carrier-side throttling? The current setup uses standard SIP-T relay configuration without any custom headers.