We are facing an issue with our application in case it is executed in citrix environments. The application is a client/server application developed in Delphi and .NET. The database in the backend is Oracle. Many customers have the application running in a citrix/terminal server environment. In general that works. Now the problem:
Some of them get a high amount of windows "Not responding" messages in combination of a freezing white screen. To my understanding this is done automatically by windows in case the application does not react to the message loop any more (in a given time). That happens usually in case there is a long running database operation or some UI operation takes longer. The default value here is 5s (hungapptimeout).
To avoid the "not responding" and freezing we already call a PeekMessage(Msg, 0, 0, 0, PM_NOREMOVE) on the Windows API. When running locally that works quite well to avoid the "Not responding" behaviour for long running operations.
On the terminal server (or at least on some configurations) it looks like this is not considered at all.
We also tried to increase/modify the hungapptimeout without any success on the terminal servers (see: https://blogs.msdn.microsoft.com/meason/2010/02/04/hungapptimeout).
- Is there anything that we should take into account? --> For sure it might be the issue that there are simply to many users on that server and the resources might be limited already but we have no sign that indicates something like that.
- Are there special settings on citrix to configure that behaviour.
Hope that someone can help or at least can give a hint.