- Genesys Cloud US1 (Enterprise)
- JMeter 5.6.2 with SIP plugin 1.5
- Concurrent threads: 150
- SIP trunk: Genesys Cloud SIP Trunking
- Load pattern: Ramp-up 30s, Hold 60s, Ramp-down 30s
Is it possible to configure the SIP stack to handle a higher volume of simultaneous BYE messages during teardown?
We are running a basic load test to validate call capacity. The INVITE phase works fine. 150 calls connect. The hold phase is stable. But when the ramp-down starts, JMeter sends BYE messages to terminate the calls. We are seeing a high rate of 408 Request Timeout errors on the BYE requests. The error happens consistently when more than 20 BYEs are sent within the same second.
The JMeter log shows:
javax.sip.RequestTimeoutException: Request timed out: BYE sip:[email protected];transport=udp
We have checked the network. No packet loss. The latency to the Genesys Cloud SIP endpoints is normal, around 45ms. The issue seems to be on the server side or a limitation in how the SIP stack processes teardown signals under load.
We tried increasing the JMeter SIP request timeout to 10 seconds. The error still occurs. The calls do eventually disconnect, but the JMeter test reports them as failures because of the 408.
This is blocking our performance validation. We need to know if this is a known limit for concurrent BYE processing in Genesys Cloud. Or if there is a configuration setting on the SIP trunk or the Virtual Contact Center that can help. We are using the default settings for the SIP trunk. No custom headers or advanced routing rules.
The goal is to simulate a realistic end-of-shift scenario where many agents hang up at once. We need the teardown to be clean and successful in the test results. Any advice on how to adjust the load pattern or the SIP settings would be appreciated. We are new to this scale of testing.