Analytics API 429 errors during JMeter load test

General Discussion
1

Trying to understand why GET /api/v2/analytics/conversations/details/query returns 429 Too Many Requests when JMeter hits 20 TPS.

Background

Running a basic load test against Genesys Cloud Analytics endpoints using JMeter 5.6.2. The environment is a standard US-East-1 instance. No custom rate limit configurations applied.

Issue

Request frequency is 20 requests per second. Each request filters by a specific date range. After 5 seconds, the API starts returning 429 errors with Retry-After: 10 header.

Troubleshooting

  • Verified API key permissions are correct for analytics.
  • Checked developer console, no global rate limit warnings visible.
  • Reduced TPS to 5, errors stop immediately.
  • Documentation says 100 req/min for this endpoint, but 20 TPS exceeds that quickly. Is there a per-user or per-tenant burst limit I am missing?
2

I typically get around this by implementing exponential backoff in JMeter. The default retry logic spikes the rate. Use the Custom JMeter Functions plugin to add a random delay between 1000ms and 5000ms on 429 responses. This smooths the request curve. The API throttles per IP, not just per org. Check the Retry-After header in the response too.