How do I correctly to handle Architect flow timeouts during high concurrency testing?
Running JMeter 5.6 against Genesys Cloud US1. We are simulating 300 concurrent callers entering a specific IVR flow via the REST API simulation. The flow contains a simple ‘Set Variable’ and ‘Queue’ action. After about 50 concurrent sessions, we start seeing failures in the JMeter logs.
The error message looks like this:
Error 504: Gateway Timeout. The Architect flow execution did not complete within the expected timeframe. Flow ID: abc-123-def.
We have checked the WFM capacity and it seems sufficient for the load. The issue appears specific to the Architect engine processing the concurrent requests. Is there a known limit for simultaneous flow executions per tenant?
We are using the default timeout settings in JMeter (30 seconds). Should we increase this? Or is there a configuration in Genesys Cloud Architect to optimize for higher throughput? We need to validate our capacity planning for peak hours.