Trying to understand the root cause of persistent SDP negotiation failures when initiating outbound calls via WebRTC softphones through our AP-Southeast-1 BYOC trunks.
Background
Environment: Genesys Cloud AP-Southeast-1
Trunk Type: BYOC (15 active trunks, currently testing outbound on Singapore primary)
Client: Genesys Cloud WebRTC Softphone (Latest Chrome Build)
SDK Version: Genesys Cloud JavaScript SDK v3.2.1
Issue
Outbound calls fail immediately upon connection attempt with a “Call Failed” status in the UI. The browser console logs indicate an ICE gathering timeout, followed by a failed SDP offer/answer exchange. Specifically, the iceConnectionState transitions to failed within 2 seconds. This behavior is isolated to calls routed through the BYOC trunks; internal calls and PSTN calls via Genesys-managed trunks function normally.
Troubleshooting
- Verified SIP credentials and outbound routing patterns in Admin. All routes are active.
- Captured network traces. The initial SIP INVITE reaches the carrier, but the subsequent RTP stream fails to establish due to NAT traversal issues on the carrier side.
- Confirmed firewall rules allow UDP traffic on ports 50000-59999 for the carrier IP ranges.
- Tested with different WebRTC clients (Zoom Phone, Teams) routed through the same BYOC trunk, resulting in identical SDP failures.
Has anyone encountered carrier-specific SDP quirks with BYOC trunks in this region? Are there specific SDP attributes or ICE candidates that need to be forced or excluded for certain carriers in APAC?