Context:
running jmeter 5.6 to simulate heavy load on reporting endpoints in us1. the goal is to test analytics retrieval capacity during peak hours. using the /api/v2/analytics/users/summary endpoint with date range filters. the test plan triggers 100 concurrent requests per second to fetch agent performance metrics.
Question:
Is it possible to export high-volume analytics without hitting rate limits? getting 503 service unavailable errors after about 200 requests. the response headers show retry-after but the service does not recover quickly under sustained load. checking the api documentation for rate limiting on analytics endpoints, but the exact thresholds for summary reports are unclear. is there a batching mechanism or pagination strategy that reduces the load on the reporting service? currently using standard oauth client credentials grant for authentication. the architect flows are not involved here, just direct api calls to the analytics service. looking for best practices on polling intervals or request throttling to avoid these timeouts. any insights on the maximum sustainable throughput for these endpoints?