Having some issues getting my configuration to work… I am running a stress test on the Genesys Cloud Performance API to validate our dashboard rendering speed under high concurrency. We are using JMeter version 5.6.2 with a custom script designed to hit the /api/v2/analytics/interactions/details endpoint. The goal is to simulate 500 concurrent agents pulling real-time metrics. However, when the virtual user count exceeds 150, the response rate drops significantly, and we start seeing a flood of 429 Too Many Requests errors. This is happening despite our account having the standard enterprise rate limits enabled.
The environment is set up in the US East region, and we are testing against a production-like tenant with about 10,000 historical interactions. The JMeter configuration uses a Thread Group with 500 threads, a ramp-up time of 60 seconds, and a loop count of 5. We have added the necessary Authorization headers and are using OAuth 2.0 client credentials flow for authentication. The error response body typically contains a Retry-After header suggesting a delay of 1 to 2 seconds, but implementing this delay reduces the test effectiveness as we need to measure peak throughput.
I have checked the API documentation regarding rate limiting, which states that the limit is applied per user context. Since we are using a service account, I assumed the limit would be higher or shared differently. We are not seeing any issues with WebSocket connections for the softphone, only with the REST API calls for analytics data. The latency spikes from an average of 200ms to over 2 seconds before the 429 errors start appearing consistently.
Can someone clarify if there are specific limits for the analytics endpoints during load testing scenarios? We need to understand if this is a hard cap on the service account or if there is a configuration in the Genesys Cloud admin portal that allows us to increase the API throughput for testing purposes. We are trying to ensure our custom dashboards can handle the load without timing out.