Frequent restarts of Unix servers should not be underestimated.

For administrators, restarting the server is not a joke. For Windows server administrators, it is a common practice to restart Windows devices frequently, but in Unix systems this approach is not effective —— restarting by default does not bring any form of improvement.

Two cases of Unix server restart

The actual situation is: server restart operation should rarely occur — — please note that very few. Here I cite kernel updates and hardware replacement as examples, as they are the two main reasons for the restart in the Unix world. Some people have been advocating that what does not restart the server will bring some serious security risks, which is simply nonsense. If there is a security risk in the service project and the application, then the vulnerability patch can solve the problem, and the patch often does not require the device to be restarted. If the security risk exists in the kernel module, generally only need to uninstall the corresponding module, install the patch, and finally reload the module. Once there is a security risk in the kernel, it is indeed necessary to restart the server operation. But outside of this situation, there is no real reason to restart the Unix server.

Some people think that if you don't restart, other forms of risk will often follow. For example, some key services are not properly enabled at boot time, which will lead to a series of hidden dangers. Of course, this statement is correct in itself, but as long as the management work is in place, this is actually a kind of worry. Only the rookie who just took over the server device will forget to set the startup parameters of the service project correctly. But then again, if your server is in the construction phase, and it does not involve any production aspects, then you may wish to conduct various types of restart testing at will, without any adverse effects. And I think this is the best time to get familiar with the restart mechanism.

But there is another aspect to consider: those who take the restart operation as one of the important steps in troubleshooting are holding the dead pigs not afraid of boiling water, and intend to expose the problem at one time. Just say a set of Unix devices that have problems, some services that are still running can't actually start again, and this will show up after rebooting —— maybe due to segmentation failure or other Strange reason.

Reasons for Unix server restart

If we just take a few minutes to check the device and restart the device, then the real cause of the failure is completely lost in the time —— Perhaps a junior administrator inadvertently deleted some parts of the /boot directory or the /etc, /usr/lib64 directories while running a set of stupid scripts. This is the culprit in causing segmentation failures and equipment instability. However, once we choose to restart the server directly without digging into the problem, then obviously the problem will become more serious. If there is no accident, then everyone should start the recovery image —— this means that we need to face a lot of recovery work —&mdash At the same time, the production server will also be in a down state.

These are just some of the reasons we should try to avoid restarting in the Unix world. Rather than being a troubleshooting method, it's better to think of it as a kind of desperate gamble —— either find the problem, or destroy everything and slowly rebuild it. In short, no one can use the /var partition to restart the device to completely correct the error.

Please do what you need to do before the server restarts.

In most cases, it is extremely important not to restart, because the key content in the system that can help us fix the problem is It must exist before the restart, but it may not be there after the restart. After the restart, the problem will definitely reappear, but once the solution disappears with the restart behavior, the fault itself falls into an infinite loop of no solution. Unless someone decides not to reboot, try to find the source of the problem. Regrettably, there are very few people who can make such wise choices. The actual situation is: a small faulty memory stick can cause great trouble to the normal operation of the system and the startup state of the device. At this time, the right medicine is the best policy, and the restart will only bring additional losses.

Therefore, in the future when everyone is facing problems, if there is a guy who says something & ldquo; oh, it is better to restart it to see if you want to give him two big mouths. Restart is of course one of the options, but before implementing the restart, please make sure that we have taken all the measures we can think of; after all, what we have saved is our own time and energy.