Context:
Trying to understand the root cause of a persistent 502 Bad Gateway error occurring when a Genesys Cloud Architect bot node invokes a Data Action to create a case record in ServiceNow. The integration relies on a custom Data Action configured with a POST request to /api/now/table/x_custom_case. The payload includes standard fields like short_description, caller_id, and a sys_id reference pulled from the conversation context. The ServiceNow instance is running the Tokyo release, and the Genesys Cloud tenant is on the latest quarterly update. Network connectivity tests from the Genesys Cloud environment to the ServiceNow instance return 200 OK when using simple HTTP GET requests on public endpoints, indicating the firewall rules are correctly configured. However, when the bot flow reaches the Data Action node, the execution log shows the request failing immediately with a 502 status, despite the JSON payload being valid according to the ServiceNow schema.
Question:
Is there a known limitation or specific header requirement for Data Actions originating from bot flows that differs from standard webhook configurations? The error suggests a proxy or load balancer issue within the ServiceNow environment, but the same API endpoint works flawlessly when called directly from Postman or a standalone script. Could the timeout settings within the Architect Data Action node be triggering an early termination that ServiceNow interprets as a malformed request, resulting in the 502? Any insights into debugging the exact payload received by ServiceNow from the Genesys Cloud side would be appreciated.