Could someone clarify why the Agent Scripting API returns 429 Too Many Requests when JMeter hits 50 concurrent threads? We are testing script playback start/stop endpoints on Genesys Cloud US1. The flow uses standard REST calls to /api/v2/agent-scripting/playbacks.
The error happens instantly. No delay. The response body shows a retry-after value of 0. This is confusing. Usually, rate limits allow some burst. But here, it stops everything.
Here is the JMeter setup.
Thread Group has 50 threads. Ramp-up is 10 seconds. Loop count is 10. Each thread executes a GET request to fetch the script template, then a POST to start playback. The interval between requests is 0ms. We want to simulate a burst of agents starting scripts at the same time.
The environment is Genesys Cloud US1. We are using the latest public API version. Authentication is via OAuth 2.0 client credentials. The token is valid. It does not expire during the test. The issue is purely about request volume.
We checked the API documentation. It mentions rate limits per organization. But it does not specify the exact limit for scripting endpoints. Is there a hard cap? Or is it dynamic based on load?
We need to know the threshold. If we hit 50 threads, we get 429. If we drop to 20 threads, it works fine. The boundary is unclear. This makes capacity planning difficult. We cannot predict how many agents can start scripts simultaneously without hitting errors.
Is there a way to increase the limit? Or do we need to stagger the requests in JMeter? Adding a 100ms delay between threads helps, but it defeats the purpose of load testing. We want to test the system under peak stress.
Please share any insights. Are there specific headers we should add? Or is this a known limitation of the Scripting API? We are stuck on this blocker. Any help is appreciated.