Windows 2000 weird faults in two cases

Our unit is a medical institution, the entire hospital network is a WIN2000 server version of the HP server and dozens of WIN98 platform for the terminal computer composition. The database system uses Microsoft's SQL 7.0 version, the development tool for Sybase Company's Powerbuild 6.0. In the past year's operation, there have been two odd cases of "fault", which almost led to the paralysis of the system, in this detailed process described below, with a view to the Enlightenment and reference.

"Fault" one: The system runs a few months later, the entire management system running speed becomes very slow, whether it is registered or pricing hair drugs are unresponsive, resulting in the whole hospital can not work properly. Observe the server, found that the hard drive lights often flash more than, open Task Manager in the system performance monitoring, found high CPU occupancy rate, click on any program icon, the program's response is very slow, the screen refresh speed seems to be in the cartoon. We think of the first thing is a virus infection, so spend nearly million to buy a famous anti-virus software network version, scan the entire network after nothing. So suspicion is infected with some kind of unknown virus, and the software company comrades to try again still can not solve the problem. Therefore, it is suspected that the server hardware failure, please HP company personnel to detect, CPU, memory, and so on, in the near end of the tether, suddenly think that may be WIN2000 configuration problems. So, click on "Start" → "Program" → "Management tools", the column of the various management items in turn check, and finally found the problem. Originally, under Event Viewer, records are densely filled in logs such as application log, security log, and system log. Some of these records as many as nearly thousands, we suspect that WIN2000 at runtime, in order to "ensure" security, you may want to read the information in multiple log files at any time, and compare with the current situation, after the calculation to decide what to do next, this will greatly increase the system's resource consumption. Also, once there are multiple error records in the log, the situation is even worse. So, after we make a backup, empty all the log records, restart the server, the entire network finally regained vitality.

Failure two: After the system runs back to normal, accept the last lesson, we periodically empty the log file, and open the system Performance Monitor, at any time to see the CPU occupancy. But the strange situation has arisen again. Sometimes, in the absence of any work on the entire network, the hard drive lights flash again, CPU occupancy rate of 100%, this situation lasts a minute or so automatically disappear. Unfortunately, if at this time the hospital outpatient or inpatient department has a task to do, the whole system response speed will be slow as the snail, simply can not work, we follow the last solution, every day to clear the log, or solve the problem. After careful observation, we find that this phenomenon occurs on a regular basis every 16 hours, so suspicion is similar to WIN98 planning tasks in the wrong, but in the WIN2000 can not find "scheduled task" This management item, when there is no way, and to the log file to see, inadvertently found in the system log, a pair of "error "and" warning "records are repeated every 16 hours, so the immediate light, open the" error "record to see, the content is" this machine is in the forest root domain PDC. " Use the net command ' net TIME/SETSNTP: ' To configure synchronization from an external time source. "Its error code is 62; The warning record reads:" Due to repeated network problems, the time service has not found the domain controller with which it is synchronizing. To reduce network traffic, the time service has to wait 960 minutes to try again. No synchronization will be made during this time, and the network connection is restored. Accumulated time errors can cause some network operations to fail. To tell the network connection to be restored, you should resynchronize and perform "w32tm/s" from the command line. The error code is 64. We look, 960 minutes is not exactly 16 hours? Because the network of my unit is internal outside domain network, and the software system takes the server time, the network time synchronization service is not needed, this service stops the question will be solved. After looking, we finally found its home. Click "Start" → "Program" → "Management Tools" → "service". Locate the Windows Time item, double-click to modify it, stop the service, and change its startup mode to manual. Since then, the 16-hour intermittent outbreak of the "Fault" has never appeared.

Experience: Microsoft's products have been upgraded, its operating system is becoming larger and more complex, and even more bugs. In this case, the system settings are not at that time, it is easy to have some false faults, and its manifestation is very much like a virus infection or hardware damage and other major problems. This requires that when we encounter a computer failure, do not overlook the small problem. Otherwise, it may be exhausting, costly and futile. In addition, we realize that WIN2000 and SQL Server are best restarted every two months, otherwise there will be some inexplicable little trouble.