Trying to understand the rate limiting behavior for the messaging API during high-concurrency load tests.
Running JMeter 5.6.2 from Singapore. Targeting /api/v2/messaging/conversations. Simulating 300 concurrent chat session creations. Hitting 429 Too Many Requests immediately after the ramp-up phase starts. The response headers show Retry-After: 1, but adding a pause doesn’t help much when the volume spikes.
Current JMeter config:
- Thread Group: 300 threads, ramp-up 30s
- HTTP Request Defaults: Keep-Alive enabled
- Token: Using a valid OAuth2 Bearer token with
purview:messaging:readandpurview:messaging:write
The goal is to validate system capacity for a new digital channel launch. Is there a specific header or parameter I’m missing to handle the burst? Or is the limit hardcoded per tenant regardless of the plan?
Any insights on how to structure the request to avoid immediate throttling would be appreciated.
Thanks for the help.