dazed&confused
04-22-2003, 12:31 PM
I've always been given some good pointers from these forums, and I hope someone can help me again. Following is a problem that I have been unable to solve. In fact, I'm not even sure I'm on the right track.
Our LAN is much larger than this, but for the sake of simplicity, I'll try to narrow this description down as far as possible: All our servers are running Win2000 Server and all workstations are XP Pro. Server1 is running an Oracle-based application, which the client workstations can run with no problem via a mapped drive. Server2 is a dedicated Terminal Server, which other users log into in order to access that same Oracle application via a mapped drive. Those particular clients connect to the Terminal Server via XP Remote Desktop Connection.
(This 2nd group of clients is located 3 blocks away and connects via a point-to-point T1 line, but I've documented that I can re-create this problem on the local LAN, just the same.)
*The Problem: The Terminal Clients open and run the application just fine. But frequently, they leave the open application idle for several minutes. Then, once they return to the application to continue working, they receive a "Fatal Disk Error." They have the option to click OK or Cancel, but either way the error returns in a millisecond. The only solution is for me to reset their terminal session (or configure it to reset upon disconnect), but that only works as long as their next session remains active.
Incidentally, we have another server that is running Terminal Services, and the problem doesn't seem to occur on it. I've triple-checked all the configuration settings, and they're all identical between the 2 servers. The "Fatal" problem does not occur if you're physically standing at the Terminal Server and using the application. So... as best I can tell, the problem is SOMEWHERE in Terminal Services or in the communication between Terminal Server and Client.
Is there some type of communication that is not forwarded by default through a Terminal Session? Like, some type of communication that would keep the application alive? Oddly enough, there are no entries in the Event Logs that suggest anything wrong. Any suggestions?
Our LAN is much larger than this, but for the sake of simplicity, I'll try to narrow this description down as far as possible: All our servers are running Win2000 Server and all workstations are XP Pro. Server1 is running an Oracle-based application, which the client workstations can run with no problem via a mapped drive. Server2 is a dedicated Terminal Server, which other users log into in order to access that same Oracle application via a mapped drive. Those particular clients connect to the Terminal Server via XP Remote Desktop Connection.
(This 2nd group of clients is located 3 blocks away and connects via a point-to-point T1 line, but I've documented that I can re-create this problem on the local LAN, just the same.)
*The Problem: The Terminal Clients open and run the application just fine. But frequently, they leave the open application idle for several minutes. Then, once they return to the application to continue working, they receive a "Fatal Disk Error." They have the option to click OK or Cancel, but either way the error returns in a millisecond. The only solution is for me to reset their terminal session (or configure it to reset upon disconnect), but that only works as long as their next session remains active.
Incidentally, we have another server that is running Terminal Services, and the problem doesn't seem to occur on it. I've triple-checked all the configuration settings, and they're all identical between the 2 servers. The "Fatal" problem does not occur if you're physically standing at the Terminal Server and using the application. So... as best I can tell, the problem is SOMEWHERE in Terminal Services or in the communication between Terminal Server and Client.
Is there some type of communication that is not forwarded by default through a Terminal Session? Like, some type of communication that would keep the application alive? Oddly enough, there are no entries in the Event Logs that suggest anything wrong. Any suggestions?