Looking for advice on handling WebRTC connection limits in Genesys Cloud US1 during high-concurrency load tests.
Running a JMeter 5.6 script to simulate 200 concurrent agents logging in and initiating outbound calls via the softphone. The goal is to validate WebSocket stability before a major deployment. The test uses the standard WebRTC signaling flow.
Environment details:
Genesys Cloud US1
JMeter 5.6
User Agent: Custom Java HTTP Client
Endpoint: /api/v2/telephony/webRTC/connections
After 50 concurrent threads, the test starts failing with 403 Forbidden errors. The error message indicates “Invalid token” or “Unauthorized access”, even though the OAuth tokens are refreshed correctly every 50 minutes. The initial connections succeed, but subsequent connections from new threads fail immediately.
Is there a specific rate limit for WebRTC connection requests? Or is this a WebSocket session limit issue? The documentation mentions concurrent session limits, but does not specify request rate limits for the signaling endpoint.
Any tips on structuring the JMeter config to handle these limits gracefully? Using a constant throughput timer does not seem to resolve the 403 errors.