Is it possible to configure higher throughput limits on a Genesys Cloud BYOC Edge instance when running automated load tests?
We are validating our softphone capacity in the Asia/Pacific Southeast region (Singapore). The environment is running on a BYOC Edge setup, version 24.2. We are using JMeter 5.6.2 to simulate concurrent WebSocket connections for media streaming. The test plan is straightforward: establish peer connections, hold for 60 seconds, then terminate.
The issue arises when we push the concurrent user count past 200. At this point, we start seeing a significant increase in connection failures. The logs show the following error:
WebSocket connection failed: 1006 (Abnormal Closure)
This happens almost immediately after the 200th connection is attempted. The JMeter sampler reports a timeout, and the Genesys Cloud audit logs do not show any explicit 429 Too Many Requests errors, which suggests the issue might be at the edge infrastructure level rather than the API gateway level.
We have verified that the bearer tokens are valid and have sufficient scope. We also checked the network connectivity from the Singapore test server to the Genesys Cloud endpoints, and latency is stable. The problem seems to be related to the edge’s ability to handle the sudden spike in concurrent WebSocket handshakes.
We are not sure if this is a hard limit imposed by the BYOC Edge license or a configuration setting we are missing. We have looked through the documentation but could not find specific guidance on tuning the edge for high-concurrency WebSocket scenarios.
Has anyone successfully run load tests against a BYOC Edge with more than 200 concurrent users? Are there any specific configurations or limits we should be aware of? We are trying to determine if this is a capacity issue or a configuration error on our end. Any insights would be greatly appreciated.