looking for advice on handling timeout errors when executing data actions under heavy load. we are running a jmeter script from asia/singapore targeting the genesys cloud platform api. the goal is to validate throughput for a custom integration that fetches customer context via a data action in architect.
the setup involves a simple http data action calling an external rest endpoint. under low concurrency (10-20 threads), everything works fine. response times are around 200ms. however, when we ramp up to 500 concurrent sessions simulating real-world peak volumes, the data action starts failing with a generic timeout error after 30 seconds. the external service logs show that the requests are actually arriving and processing within 500ms, but genesys cloud seems to drop the connection or mark it as failed before the response is fully handled.
we have checked the platform api rate limits and are well within the allowed thresholds. the external endpoint is also healthy and can handle the load independently. the issue seems specific to how genesys cloud manages the websocket or http connection pool for data actions during bursts.
has anyone experienced similar behavior with data actions under high concurrency? are there specific configuration settings in architect or the platform api that need adjustment to handle higher throughput for synchronous calls? we tried increasing the timeout setting in the data action configuration to 60 seconds, but it did not help. the error still occurs at the 30-second mark.
also, wondering if the singapore region has any specific latency or connection handling quirks that might affect this. any insights on tuning the data action for better performance at scale would be appreciated. currently, we are stuck at this bottleneck and need to validate the integration for our client’s peak hour requirements.