Is it possible to increase the WebSocket connection limit for a single Webchat deployment in Genesys Cloud US-East? We are running load tests to determine the maximum capacity for our messaging channel before the holiday season. The goal is to sustain 500 concurrent active chat sessions. However, the system starts failing significantly after 150 concurrent sessions.
We are using JMeter 5.6.2 with a custom WebSocket sampler plugin. The test plan connects to the standard webchat endpoint /api/v2/conversations/websockets. The load ramp-up is linear, adding 10 sessions every 5 seconds. At 150 sessions, the error rate spikes to 40%. The specific error returned in the WebSocket close frame is code 1008 (Policy Violation) with the message “Rate limit exceeded”. This happens even though our API key has not hit the standard REST API rate limits. The WebSocket connections themselves seem to be throttled independently.
We have verified the following:
- Checked the admin portal for any explicit concurrency caps on the Webchat deployment. The field shows “Unlimited” or is greyed out, suggesting it defaults to a system limit we cannot see or modify.
- Analyzed the network traffic. The connections are established successfully, but the server closes them immediately after the handshake when the concurrent count hits ~150. No 504 Gateway Timeout errors are present, just the 1008 policy violation.
The latency for successful sessions remains under 200ms, so it is not a general performance degradation. It feels like a hard cap on WebSocket connections per deployment or per IP range. We are testing from a single EC2 instance in us-east-1 to simulate a centralized load balancer.
Is this a known limitation for the free tier or standard deployment? Or is there a configuration in the Architect flow or the Webchat settings that allows us to request a higher limit? We need to know if we must split the load across multiple Webchat deployments to achieve 500 concurrent sessions.