After U2000 was installed on new server with Windows2003 Enterprise Edition and several days of normal works users complain that time for U2000 client login time is about 20~25 min and after login some functions for RTN equipment are unavailable even no in menu (Trials, NE explorer, Alarms and so on).
We decided to simulate remote connection to test server thru RDP. We created 2 RDP sessions (Windows allow only 2 RDP sessions by default) but Windows allow 3rd RDP connection its console connection. Purely by chance we run “mstsc /console” command and connected to test server. When we did “Logoff” from console RDP session this problem happened in our test server.
HQ suggested reinstalling U2000. But we didn’t yield to persuasion. For that were several reasons:
In System monitor all process was in running state;
In System monitor all database was in normal state;
If restart U2000 “SDH nemngr” process all function will restored;
Server working in normal mode for several days and only then this problem appeared;
The same problem we met on test server but then all was blamed on poor server performances.
HQ couldn’t analyze this problem because server stopped write log at 12.30. In the same in system even we found that “administrator” did “Logoff”. We collected several dump files (these files were created in U2000 exactly in login moment). After HQ analyzes them they still couldn’t find a reason.
We even restored scripts file on Huawei office for check and observe and after two days this server continue works normally but only one client connecting to this server and all NE’s was unmanaged.
Reason is that the “mstsc /console” is like a local login and Java after local logoff will stop all own processes.
If use only “mstsc” and login/logoff as “administrator” no such problem.
Better create additional “power user” for RDP and never use “mstsc /console” for RDP connection if U2000 server running.