Configuration is broken for some reason… I am trying to simulate a high-volume outbound dialing scenario using the Genesys Cloud platform_api. The goal is to push 500 concurrent calls per second to test the dialer capacity. I am running JMeter 5.6.2 from an AWS instance in ap-southeast-1 to mimic our Singapore office load. The test hits the POST /api/v2/outbound/campaigns endpoint to create campaign entities, followed by rapid-fire requests to /api/v2/outbound/calls to initiate dials. Everything works fine at 50 threads, but the moment I ramp up to 200 threads, the error rate spikes to 60%.
The specific error returned is a 429 Too Many Requests with the body: {“code”: “rate_limit_exceeded”, “message”: “You have exceeded the rate limit for this API endpoint”}. This happens even though I am using a dedicated org with no other users logged in. I have verified the API key permissions, and they include outbound:campaign:write and outbound:calls:write. The latency for successful calls is around 200ms, which seems healthy, but the 429s appear consistently after the 15th request per second from a single client thread.
I am aware that Genesys Cloud has rate limits, but the documentation suggests a limit of 100 requests per second per API key for outbound operations. My JMeter config uses a CSV Data Set Config to feed unique contact IDs and a Throughput Shaping Timer to control the rate. I have tried adding a 100ms delay between requests, which reduces the 429 errors but fails to meet the target throughput. I also tried rotating API keys across threads, but the 429s still persist, suggesting the limit might be tied to the org or IP rather than the key.
Has anyone successfully load-tested the Outbound Dialing API with similar concurrency? I am looking for best practices on handling rate limits in JMeter for Genesys Cloud. Should I be using a different endpoint or batching my requests? Any insights on the actual rate limits for outbound calls would be helpful. I am stuck on this and need to complete the capacity planning report by Friday.