Looking for advice on why the Genesys Cloud WFM Quality API is returning 429 Too Many Requests errors when attempting to simulate a moderate load of concurrent evaluation submissions. The goal is to validate the system’s capacity for bulk quality scoring uploads before going live with a new QA process.
Running JMeter 5.6.2 from a server in Asia/Singapore. The test plan uses a Thread Group with 50 concurrent threads, each executing a loop of 10 iterations. The target endpoint is /api/v2/wfm/quality/evaluations. Each thread sends a POST request with a JSON payload containing a single evaluation record. The payload size is approximately 2KB per request.
The test starts smoothly, but after about 30 seconds, the error rate spikes significantly. Here are the specific details:
- Endpoint: POST
/api/v2/wfm/quality/evaluations - Tool: JMeter 5.6.2
- Location: Asia/Singapore
- Threads: 50 concurrent
- Ramp-Up: 0 seconds (instant start)
- Loop Count: 10
- Error: HTTP 429 Too Many Requests
The response headers for the 429 errors include a Retry-After field, but the values are inconsistent, sometimes suggesting a wait of 1 second, other times 5 seconds. The standard X-RateLimit-Reset header is also present, but it seems to reset faster than the errors stop occurring, which is confusing.
I have tried adding a constant throughput timer to limit the requests per second, but the 429 errors still appear. The documentation mentions rate limits for WFM APIs, but it is not clear if there are specific limits for quality evaluation submissions versus other WFM resources like schedules or groups.
Is there a recommended pacing strategy for bulk quality uploads? Or are there specific headers or authentication methods that help mitigate rate limiting for this endpoint? Any insights on how to structure the JMeter test to stay within the acceptable throughput limits would be appreciated.