Quick question about the Virtual Agent endpoint behavior under high concurrency.
429 Too Many Requests
Running JMeter 5.6 against /api/v2/conversations/webchat/events with 100 concurrent users in US1. The limit seems much lower than standard API throughput expectations.
Environment details:
- Genesys Cloud US1
- Tool: JMeter 5.6
- Endpoint:
/api/v2/conversations/webchat/events - SDK Version: N/A (Direct HTTP Requests)
- Concurrent Users: 100
- Test Duration: 5 minutes
- Bot ID:
va-12345678-1234-1234-1234-1234567890ab
The test plan simulates a simple QA bot interaction. Each virtual user initiates a webchat session, sends a single text message event, and waits for the bot response. After approximately 15 seconds, the majority of threads receive a 429 Too Many Requests error with the following body:
{
"message": "Too Many Requests",
"code": "too_many_requests",
"status": 429
}
The retry-after header is present but varies between 10 and 30 seconds, which breaks the load test timing. The issue persists even when reducing concurrent users to 50. I have verified that the bot configuration is correct and responsive in the Genesys Cloud UI. The problem appears specific to the API gateway rate limiting for the Virtual Agent service, not the bot logic itself.
Is there a specific rate limit for webchat events per bot or per organization? I need to understand if this is a hard limit for load testing purposes or if there is a way to increase the throughput for QA scenarios. I am trying to validate the system stability under expected peak loads, but the 429 errors prevent accurate measurement of response times and error rates.
Any insights on the rate limiting strategy for Virtual Agent APIs would be appreciated. I have checked the API documentation but found no clear guidance on concurrency limits for webchat events.