Context:
Running a baseline load test for concurrent WebSocket connections using JMeter 5.6.2. The setup targets Genesys Cloud BYOC Edge 24.1.0 in Asia/Singapore. The goal is to validate API throughput for token generation under stress. The test script initiates 50 concurrent threads, each requesting an OAuth2 token via POST /oauth/token every 60 seconds to simulate active agent sessions. The environment is configured with standard security settings, no custom rate limit overrides applied. The load pattern is steady-state, not a spike. The issue arises when the token refresh rate exceeds 5 requests per second. The platform returns HTTP 429 Too Many Responses immediately, halting the test threads. The response headers indicate the limit is per-client-id, but the documentation suggests higher limits for production tiers. The error occurs consistently at the same threshold regardless of thread distribution. JMeter logs show the requests are spaced correctly, but the aggregate rate triggers the block. The environment uses standard OAuth2 client credentials flow. No IP whitelisting issues detected. The latency to the endpoint is under 100ms, so network is not the bottleneck. The error payload is standard: {“code”:“too_many_requests”,“message”:“You have exceeded the rate limit.”}.
Question:
Why does this setting block token refreshes at such a low RPS threshold during a controlled load test? Is there a specific configuration for OAuth2 rate limits that needs adjustment for BYOC Edge environments, or is this a hard cap for the current tier?