What’s the best way to handle WebRTC timeout errors in JMeter for Genesys Cloud US1? running jmeter 5.6 to simulate 100 concurrent agents connecting to the agent desktop sdk v2.4 in the us1 region. after about 3 minutes of load, many websocket connections drop with a 1006 abnormal closure error. the logs show ICE candidate gathering failed and then connection reset. the initial connection works fine but the stability under load is poor. i am using the standard ws://wss-us1.genesis.com endpoint for the websocket handshake. is there a specific header or keepalive mechanism i need to add in the jmeter websocket sampler to prevent this? also, the x-api-key seems valid but the auth token refresh logic might be interfering with the websocket lifecycle. any examples of a jmeter script that maintains a stable webrtc session for more than 10 minutes would be very helpful. currently, the test fails because the agents disconnect and cannot receive calls. i have checked the network logs and there are no obvious firewall blocks. the issue seems to be related to how the sdk handles reconnection attempts during high concurrency. please advise on best practices for load testing webrtc softphones with jmeter.