We are facing an issue with clients using port 443 outside the DMZ and internal port 444 being used to redirect to internal IS servers (2). Clients are posting via webservices and see time out errors periodically. This is not happening for all data sends to our systems but it seems to be sporadic.
If there aren’t any resource bottle necks in your system (you will see CPU spikes at the same time when those timeouts occur if that’s the case) you can increase the timeout value or decrease it and add a retry mechanism. You need to find out where that timeout come from first though. Having more threads then you are supposed to have can also cause that. You can use this document for referance. These kind documents can be found in empower.