Just noticed that our multi-org Premium App is experiencing intermittent failures when invoking a custom HTTP request via Architect. The flow is designed to synchronize user attributes with our external CRM during the initial contact routing phase.
HTTP 504 Gateway Timeout
This occurs specifically when the external endpoint exceeds the default 30-second timeout threshold. We are utilizing the latest Genesys Cloud API SDK (v2.14) for the integration logic, but the timeout seems to be enforced at the platform level rather than the application layer. The request is a simple GET to fetch user preferences, which should be lightweight. However, under high concurrency, the response time from the CRM spikes, triggering the timeout.
We have verified that the external service is healthy and responds within acceptable limits for most requests. The issue appears to be tied to the Architect flow’s synchronous execution model. Is there a way to configure a longer timeout for specific HTTP requests within the flow, or should we be redesigning this to use an asynchronous data action? The current setup is causing dropped calls during peak hours, which is unacceptable for our enterprise clients. Any insights on best practices for handling external API latency in Architect flows would be appreciated.