Having some issues getting my configuration to work…
I am running a load test on our SIP trunk configuration in the US1 environment. The goal is to validate the maximum concurrent call capacity before our peak season. However, the test fails consistently after reaching approximately 150 concurrent sessions.
Environment Details:
- Region: US1
- JMeter Version: 5.6
- Genesys Cloud SDK: 2.3.0
- SIP Trunk: Standard SIP Trunk with 500 capacity licensed
- Architect Flow: Simple IVR → Transfer to Queue
Test Scenario:
We are simulating inbound calls using JMeter with the SIP plugin. The script sends SIP INVITE requests to the Genesys Cloud SBC endpoint. We ramp up the thread count to 200 users over 60 seconds. Each call is held for 30 seconds before sending a BYE.
The Problem:
Around the 150th concurrent call, we start seeing a high rate of 408 Request Timeout responses from the SBC. The calls do not ring in the queue; they simply fail at the connection stage. Interestingly, the Genesys Cloud admin dashboard shows the trunk utilization at only 30%, so it is not hitting the license limit.
I checked the SIP Trunk Configuration Guide and ensured that the timeout settings are standard. I also verified that the firewall allows UDP traffic on port 5060 and the media ports.
Observations:
- The 408 errors occur almost immediately after the INVITE is sent.
- Reducing the concurrency to 100 threads results in 0% failure rate.
- The latency between the load generator and the SBC is consistently under 50ms.
Is there a hidden rate limit on SIP INVITE processing per second for standard trunks? Or could this be related to how the SDK handles the WebSocket connection for signaling during high load? I need to understand if this is a configuration error on our side or a platform limitation.
Any insights into troubleshooting SIP timeout errors under load would be appreciated.