No idea why this is happening, the SIP trunk connection drops when we hit 500 concurrent calls.
Our load test setup uses JMeter 5.4.1 with a ramp-up period of 60 seconds. The target environment is Genesys Cloud Edge (US-East-1). We are seeing a massive spike in HTTP 408 Request Timeout errors specifically on the /api/v2/telephony/providers/edge/connections endpoint. The error logs show SIP 408: Request Timeout before the call is even established. We have verified the network latency is under 50ms. The WebSocket connections for media seem stable, but the signaling fails. Is there a specific rate limit on the connection establishment API that we are hitting? We are not seeing 429 errors, just 408. Here is the relevant JMeter thread group configuration:
- Thread Count: 500
- Ramp-Up Period: 60 seconds
- Loop Count: 1
The issue happens consistently at around 450 concurrent users. The logs show the SIP INVITE is sent but no 100 Trying response is received within the timeout window. We have checked the Genesys Cloud telephony settings and the SIP trunk configuration looks standard. No custom headers are being added. The edge location is correct. We are using the default timeout settings in the Genesys Cloud admin console.
Is it possible that the WebSocket connection pool is exhausted? We are using the default pool size. The error message in the Genesys Cloud logs is quite generic. It just says Connection timed out. We have tried increasing the timeout in JMeter to 30 seconds, but the issue persists. The calls fail before they even ring the agent.
Any ideas on what could cause a 408 timeout on the connection API during a load test? We have checked the firewall rules and they are open. The SIP trunk provider has confirmed they are receiving the INVITEs. The issue seems to be on the Genesys Cloud side. We are using the latest version of the Genesys Cloud APIs.
Thanks for any help.