Two strange "faults" in Windows2000

Our company is a medical institution. The entire hospital network is composed of an HP server installed with WIN2000 Server version and dozens of terminal computers using WIN98 as the platform. The database system uses Microsoft SQL 7.0 and the development tool is SYBASE PowerBuild 6.0. In the last year's operation, two strange "faults" have occurred, which almost caused system paralysis. Here we will describe the detailed process as follows, in order to provide some inspiration and reference for everyone.
"Fault" 1:After the system was running for a few months, the operation speed of the entire management system became very slow. Both registration and price-based drug delivery were slow, leading to the failure of the whole hospital to work normally. Observe the server and find that hard drive lights often flash off. Turn on the system performance monitoring in the task manager and find that the CPU usage remains high. Click any program icon, and the response of the program is very slow, the screen refresh speed seems to be playing cartoons. The first thing we thought of was virus infection. So we spent nearly yuan to buy the online version of a famous anti-virus software and found nothing after scanning the entire network. So I suspected that it was infected with an unknown virus, and the software company's comrades once again failed to solve the problem. Therefore, it is suspected that it is a server hardware fault. After HP personnel are asked to detect the fault, there will be no problems with CPU and memory. When it is almost exhausted, it may be a problem with WIN2000 configuration. Then, click "start"> "program"> "Management Tools" in turn to check the various management items under this topic, and finally find the problem. Originally, under "Event Viewer", "application logs", "security logs", and "system logs" were all filled with records. Some of the records are as many as thousands. We assume that WIN2000 may read information from multiple log files at any time to ensure "security" during runtime, compare with the current situation and determine the next step after calculation. This will greatly increase the system resource consumption. In addition, once the log contains multiple error records, the situation becomes more serious. After the backup, all the log records are cleared and the server is restarted. The entire network is restored.

"Fault" 2:After the system returns to normal, we will receive the last lesson. We regularly clear the log files and turn on the system performance monitor to check CPU usage at any time. But the strange situation has appeared again. Sometimes, when the entire network is not working, the hard drive lights will flash continuously, and the CPU usage rate reaches 100%. This situation will automatically disappear for about one minute. Unfortunately, if the hospital's outpatient or inpatient department has a task to perform at this time, the response speed of the whole system will be slow, such as snail bait, and the system will not work properly. We follow the previous solution, clearing logs every day cannot solve the problem. After careful observation, we found that this phenomenon occurs on a regular basis every 16 hours, so we suspect that it is a program task similar to WIN98, however, the "scheduled task" management item cannot be found in WIN2000. When there is no way to implement it, go to the log file and see it. It is not intended to be found in the system log, there is a pair of "error" and "warning" records that appear again every 16 hours, so the eyes are bright, open the "error" record, the content is "this machine is the PDC in the forest root domain. Use the net command 'net time/setsntp: ', Configure synchronization from the external time source. "The error code is 62;" warning "the record is:" due to repeated network problems, the Time Service has not found the domain controller to be synchronized with it. To reduce network traffic, the time service will take 960 minutes to try again. During this period, no synchronization will be performed, and the network connection has been restored in a timely manner. A cumulative time error may cause some network operations to fail. To notify that the network connection has been restored, you should re-synchronize it. Please execute "w32tm/s" "from the command line "". The error code is 64. As we can see, isn't 960 minutes equal to 16 hours? Because my company's network is an internal out-of-the-box network and the software system uses the server time, the Network Time Synchronization Service is not required. If the service is stopped, the problem will be solved. After searching, we finally found our hometown. Click Start> program> management tool> service ". Find the Windows Time project, double-click it, modify it, stop the service, and change its startup mode to manual. From then on, the "fault" that triggered intermittent attacks within 16 hours never again.
Experience: Microsoft's products have been upgraded, and their operating systems are getting bigger and bigger, with more and more complicated configurations and even more and more bugs. In this case, when the system is not set up at the time, it is easy to have some false faults, and its forms are like virus infection or hardware damage. This requires us not to ignore small problems in the case of computer faults. Otherwise, it may be exhausting and costly. In addition, we understand that it is best to restart WIN2000 and SQL SERVER every two months, otherwise there will be some inexplicable little trouble.