VNC Remote Desktop solves three server failures

        The author has been engaged in operation and maintenance for many years, and the operation and maintenance work is usually not very stressful, but there are only a few days every month - the server is exhausted! The server that our company installs Windows system rarely runs continuously for more than three months, and regular restart is almost a must. Otherwise, various problems such as crashes, blue screens, slow response, etc. are coming.

Adding a timed restart server command to a scheduled task is really helpful in reducing server failures, but it just covers up the problem rather than the final solution. The price is "to pump" to death"--the server crashes and can't even boot the operating system. At this time, it will be more miserable. In the middle of the night, I have to rush to the engine room to fight the fire. If you catch up with the holiday, travel, and return to your hometown, it is even more miserable. For this reason, I even tried to change my mind many times! A friend recommended that the author use hardware VNC to solve the suddenness. The author used a period of use and experience accumulation. I feel that the pressure on operation and maintenance has been reduced a lot. Here, I would like to introduce you to the operation and maintenance peers to help you get rid of the bitter sea as soon as possible.

When it comes to VNC, most of the operators are familiar with it. It is a generic name for a long-established remote desktop tool. It is known for supporting multiple platforms and has become the de facto standard for remote desktops, including the famous RealVNC and TightVNC. and many more. The so-called hardware VNC means that it is hardware that integrates the functions of the VNC server. It is installed independently of the managed server. It does not need to install any VNC server on the managed server. It only needs to connect the hardware VNC to the managed server. Then you can connect to the managed server through various VNC clients. Compared with various software VNCs, the biggest advantage of hardware VNC is that it does not depend on the server. The hardware VNC can remotely configure the BIOS, view blue screen errors, connect to the server of the giant card, and even connect the server if it crashes. Here is a brief introduction to the three examples of the use of hardware VNC to solve the Windows 2003 server failure, including some tips summarized by the author, I hope to be helpful and inspiring to the operation and maintenance of the same channel:

Example 1. Server CPU or memory usage is too high, slow response

Solution: Find the abnormal process that takes up CPU or content too high and end it as soon as possible. If you can't end, you can consider restarting the server.

Specific operation: It can be imagined that the server response has been very slow at this time, and the VNC or remote desktop service running on the server is often unable to connect. At this point, you should connect to the server through hardware VNC. Considering the slow response of the server, it is difficult to operate. You should reduce the mouse operation and try to complete the operation through shortcut keys. The quickest order of operation that the author summarizes is (1) investigating the task manager by pressing Ctrl+Shift+ESC; (2) switching to the “process” by Shift+Tab; (3) finding the server CPU and After the memory process usage occupies an abnormal process, it ends with the Alt+E key. According to this process, the abnormal process K can be dropped at least 2 or 3 times, and then the specific cause of the abnormality or the corresponding process can be restarted. There is also a small trick, which is to connect the hardware VNC through TightVNC, then press the “Ctrl tool item” in TightVNC and then press Shift+ESC to bring up the task manager. Try it and you will understand. No, no explanation.

If you can't find the problem process, you can only choose to restart the server to recover the fault quickly. The quickest operation I use is: (1) Press TightVNC and press the R button after pressing the Windows flag button. (2) Enter the command ‘ shutdown -r -f -t 00’ in the Run dialog box and press Enter. The system will restart immediately.

Instance 2: Server poisoning or attack causes network bandwidth to be occupied

Solution: Disable network interface, find out the specific cause Re-enable

Specific operation: Under normal circumstances (such as the server is affected by SYN FLOOD attack, the maximum number of connections is full), the network connection of the hardware VNC is not affected, at this time, the controlled host can be directly controlled remotely. “On the network neighborhood" icon, click the right mouse button to select “properties", this will open the “network connection" window, select the corresponding network connection and click the right mouse button to select “disable” In some special cases (such as a large number of packets sent to the LAN after the server is poisoned), the network connection bandwidth of the hardware VNC is also affected. At this time, the remote desktop screen refresh of the hardware VNC will be slower. It is recommended to operate through the shortcut keys. The author's approach is: (1) press the "Windows flag button" under TightVNC, then click the R button, type "ldd; cmd + enter" & rdquo;, this will bring up the Dos command prompt window; (2) in Dos In the window, type the command ‘netsh interface set interface name="eth0" admin=DISABLED ’ to disable the network connection with the network name eth0. The part of the command needs to be replaced with the corresponding network connection name. It is recommended to prepare accordingly in daily maintenance, and replace the Chinese network connection name of “local connection*” with “English” short form of “eth*” format, which is ready for use at key moments. In addition, you can use the ‘netsh interface show interface’ command to view the network connection name and current status. You can re-enable the network connection eth0 by using the command ‘netsh interface set interface name="eth0" admin=ENABLED ’.

Example 3, server blue screen

Solution: Record blue screen information, try to restart the server, if it can not boot, enter safe mode

Specific operation: In the case of blue screen, the hardware VNC can see the actual information of the screen. Sometimes this part of the information can indicate the specific faulty hardware driver or DLL program. In this case, the screen can be displayed through the VNC client. Copy it and leave it for later analysis (of course, in many cases, the blue screen prompt information of Windows is of no value). Then we can contact the computer room to help us restart. In general, the restart can recover more than 80% of the faults. If you enter the operating system and restart the blue screen again, then we can only use the computer room to say "sorry, trouble you, can you restart again? …", and then pegged to the VNC client, once you see the server The boot screen starts to press F8 madly, choose to enter safe mode, and then slowly repair the specific fault. By the way, a small experience, in this way can also know exactly whether the computer room network management to restart the server, this point is not much to explain, understand.

The above are some of the small experiences and techniques that I have summarized to solve the server failure. If you have a better solution, or in practice If you have any questions or difficulties in the operation and maintenance work, please feel free to contact and communicate with the author and learn from each other!