Problem
Trying to update an external HTTP data action timeout through the API. The Architect UI caps it at 3 seconds, but our backend service consistently takes around 5 seconds to process the JSON payload. We’ve got a queue backing up because of it. The retry logic isn’t catching the dropped connections properly. Need to push the timeout higher without hitting the platform default.
Code
Used the Python SDK to patch the data action definition. Here is the exact payload we are sending:
{
“timeout”: 8000,
“url”: “https://api.internal.service/lookup”,
“method”: “POST”,
“requestTemplate”: “{{body}}”
}
The request goes through client.data_actions_api.update_data_action(org_id, action_id, body=payload). Response comes back with a 200 OK.
Error
Checking the UI after the call shows the timeout field reverted to 3000. Testing the action in Architect triggers a 408 Request Timeout after exactly 3 seconds. The backend logs show the request is still processing at 4.5 seconds. Seems like the platform is silently clamping the value. We’ve tried setting it directly in the REST proxy configuration too, but the same thing happens. The network trace shows the TCP handshake completes fine. Just the application layer drops it.
Question
Is there a separate endpoint or a specific JSON property to override the platform default? Documentation doesn’t mention a hard limit for enterprise orgs. Logs show the connection just drops at the 3 second mark.