SIP 408 Timeouts on Secondary BYOC Trunks During Heavy Analytics Queries

Need some help troubleshooting intermittent SIP 408 registration drops on our secondary BYOC trunks in the APAC edge. This occurs precisely when high-volume analytics queries run against the 15-trunk setup. The primary trunks remain stable, but the failover logic fails to re-register the secondary endpoints within the expected window. Is this a known resource contention issue?

SIP/2.0 408 Request Timeout