408 Request Timeout received on REGISTER request to sip.genesyscloud.com for BYOC trunk trunk_id: 8f3a2b1c-4d5e-6f7g-8h9i-0j1k2l3m4n5o.
We are deploying a multi-tenant Partner App that requires reliable SIP trunk connectivity for several customers in the US1 region. The integration uses a dedicated SIP registrar hosted in AWS us-east-1 with a static public IP. The Genesys Cloud BYOC trunk is configured with this IP in the allowlist and uses TLS 1.2 for signaling.
The issue manifests intermittently, approximately once every 4-6 hours, where the registrar fails to register with Genesys Cloud. The logs show the REGISTER request being sent correctly, but no 200 OK response is received within the 32-second expiry window. Subsequent retransmissions also result in 408 Request Timeout. A manual restart of the SIP registrar service usually resolves the issue temporarily, but the problem recurs.
Environment details:
- Genesys Cloud Region:
US1 - BYOC Trunk Type:
SIPwith TLS - SIP Registrar:
Kamailio 5.6.4onUbuntu 22.04 - Network: Direct connect from AWS
us-east-1to Genesys Cloud endpoints - Firewall: No intermediate proxies, direct UDP/TCP 5061 traffic
We have verified the following:
- The static IP is correctly added to the BYOC trunk allowlist in the Genesys Cloud admin console.
- TLS certificates are valid and not expired.
- Network connectivity tests (telnet to
sip.genesyscloud.com:5061) succeed consistently. - No rate limiting errors (
429 Too Many Requests) are observed in the logs.
The intermittent nature of the 408 timeout suggests a potential issue with the Genesys Cloud SIP registrar load balancer or a transient network path issue. However, given that other customers on the same AWS region do not report similar issues, we suspect a configuration mismatch in our BYOC trunk settings.
Has anyone encountered similar 408 Request Timeout errors with BYOC trunks in the US1 region? Are there specific timeout or retry settings in the Kamailio configuration that should be adjusted for Genesys Cloud BYOC? We are considering increasing the register_expires value, but want to ensure this aligns with Genesys Cloud best practices. Any insights or troubleshooting steps would be appreciated.