01.09.2026
10:03
Hi,
I have a case that I cannot fully explain and I would like to clarify how WEBCON BPS handles timeouts for automations triggered automatically.
The scenario is:
* An automation generates a document.
* As part of the process, it retrieves data for approximately **2,000 people in a loop**.
* When I run the same path **manually**, it works correctly. After a few minutes, the process completes and all 2,000 people are retrieved.
* However, when the same path is triggered by a **Cyclic Action / Timer**, it fails after approximately **1 minute** with the following error:
`Error 12002 calling WINHTTP_CALLBACK_STATUS_REQUEST_ERROR, 'The operation timed out.'`
The timeout on the service/automation level has been set to **9400 seconds**, so this does not seem to be caused by an intentionally configured short timeout.
This raises a few questions:
1. Is a **Cyclic Action / Timer** executed in a different context than a manually triggered process, and are different timeout limits applied?
2. Can the **service account** under which the automation is executed affect the timeout or the way the request is handled?
3. Is there any **higher-level timeout** configured at the WEBCON BPS service, IIS, WinHTTP, or another component level that could terminate the operation after ~60 seconds regardless of the automation timeout?
4. How do the following timeout settings relate to each other:
* WEBCON BPS service timeout,
* automation timeout,
* WinHTTP timeout,
* IIS timeout?
5. Could **IIS** be involved here, even though the exact same path works when started manually?
6. Is there a specific WEBCON BPS setting or component responsible for terminating the execution after approximately one minute?
The most interesting part is the difference:
**Manual execution → several minutes → OK**
**Cyclic Action / Timer → ~1 minute → WinHTTP 12002**
Could someone from WEBCON clarify the timeout hierarchy for automations triggered by Cyclic Actions/Timers and, in particular, which component can generate this specific `WINHTTP_CALLBACK_STATUS_REQUEST_ERROR`?
Any insight into where the ~60-second limit comes from would be greatly appreciated.
