How do I maintain a server, and what does server maintenance mean?

When we talk about web server security, the best advice I can give you is not to be a layman. When hackers start attacking your network, they will first check for general security vulnerabilities before considering more difficult ways to breach your security system. So, for example, when all the data on your server is in a FAT partition, even installing all the security software in the world won't help you much.

For this reason, you need to start with the basics. You need to convert all disk partitions on the server that contain sensitive data to NTFS format. Again, you'll need to keep all your anti-virus software up to date.

I recommend that you run anti-virus software on both the server and the desktop terminal. The software should also be configured to automatically update the virus database files on a daily basis. You should also know that you can install anti-virus software for Exchange Server.

The software scans all incoming emails for infected attachments, and when it finds a virus, it automatically quarantines the infected email before it reaches the user.

Another great way to protect your network is to limit the amount of time users spend in the company to access the network. A temporary employee who normally works during the day should not be allowed to access the network at 3 a.m. unless the employee's supervisor tells you that it is for a special project.

Maintenance is a telecom thing. Maintenance includes: constant temperature, constant humidity, and a level 8 shockproof place is where the server is placed (generally a telecommunications service point).

Maintenance is generally carried out by shutting down the server at night, mainly shutting down the server for cleaning, and monitoring the operation of the server. Generally, 30 minutes is the maximum time. NetEase is two hours on Tuesday afternoon; Sina portal is night.

After shutting down, you can't open all the ** on this server.

In general, there is a problem in the maintenance of server data, and there is a problem with the Fengyou or **program. When there is a problem in the company, it will generally close to find the problem and deal with the problem. After the processing is completed, ** access will be opened.

Avoid panic and talk from the outside world saying that the server is being repaired or upgraded. It is not excluded that there will be real data upgrades, but this kind of normal, the maintenance date set in advance will be announced in advance. There will be no sudden upgrades.

There is no notification. Generally, server maintenance is mainly divided into two parts: hardware maintenance and software maintenance. 1. Hardware maintenance: Hardware maintenance refers to the detection, replacement, and upgrading of server hardware, most commonly upgrading CPU, memory, and adding disks.

Disk expansion is the most common type of sales, because the longer the service runs, the larger the amount of data generated, and more disks need to be used for storage. 2. Software maintenance: There are many software maintenance parts, mainly including:

1.Operating system upgrades. There is no essential difference between a server and a PC, and they also run a basic operating system, so various patches need to be upgraded to protect the system security and avoid possible system errors.

2.Software Maintenance. Software refers to all kinds of software running on the server, because the services provided by the server are often increasing, so the software that provides the service also needs to be upgraded, bug fixed, or even rewritten.

3.Data Maintenance. Data is the core of the service, so data backup is a regular job.

4.Security maintenance. There are often security vulnerabilities in the system that have not yet been discovered, and the attack behavior can be identified and prevented by checking the operation logs of the database and system.

Hardware maintenance requires the server to be shut down, while software maintenance often requires a temporary shutdown of the service, which is absolutely necessary, although it will cause inconvenience to the service user. And it's not impossible to deal with the situation that frequent services can't be done - multiple servers and multiple databases, but the cost increases.

The maintenance of the server is to troubleshoot the hardware and software of the server and avoid the interference of external forces to the greatest extent, so as to ensure its normal and stable operation. The first thing to do is to change the password permission regularly, and the operating system and login background provided by the general independent server provider are the passwords that have been set, usually the default password, and then handed over to the enterprise users. After reinstalling the system, the user should change the password as soon as possible, and then also need to change the password regularly, which not only protects the security of the data, but also effectively ensures that the independent server is not invaded.

The second point is that when the firewall is not easily changed, when the general provider provides an independent server, the firewall will also be set up accordingly, and most of this setting will be adjusted according to the business and scale of the enterprise. When you make changes to the firewall, do not modify the port at will to avoid abnormal use of the independent server. Ruixun Network has a professional technical team to escort your server and be on call 24*7 hours.

Ruixun Network summarized 7 points to dig the chain: 1. Do a good job in the initial security protection of the server, 2. Modify the remote port of the server. 3. Set a complex password. 4. Patch known vulnerabilities 5. Multi-server protection is not scattered.

6. Firewall parallel game technology. 7. Back up the data regularly.

We know that servers provide round-the-clock services for all kinds of platform systems, so basically servers are running 7x24 hours a day. In such a long period of operation, the server needs to be maintained regularly in order to detect potential problems in time, just like humans, and regular medical check-ups are required.

1. System upgrade and patch repair

The most used server system is windowsserver and linux, no matter what kind of system it is, there are certain defects, so when a vulnerability is exposed, we must fix it as soon as possible to prevent it from being exploited by hackers.

Overall, Linux is more stable than Windows Server.

2. Server running status check:

We log in to the server at least once a week to check its performance, for example:

Check whether the server is under heavy load, and when the load is large, it will directly affect the stability of the server.

Disk usage: When some programs are abnormal, a large number of logs will be written, resulting in full server disks;

whether the memory consumption is normal;

Whether the bandwidth usage is normal, and the bandwidth usage will cause others to load too slowly when requesting services;

In addition, it is necessary to focus on checking various logs to determine whether there are malicious attacks and other behaviors.

Many people buy servers and throw them there, rarely manage them, and they may be hacked after a long time.

3. Application service check:

Whether the service process is normal and whether the process is abnormal and killed by the system;

Whether the directory and file permissions are normal, and whether malicious files have been uploaded;

4. Hard-to-take the spine parts for inspection and maintenance

The hardware of the server will slowly age under the long-term operation of the jujube. And among all the parts, hard drive damage is the most common. Of course, the hardware inspection is not carried out by us, but the computer room management will assist us to do the inspection, because the servers are hosted in the remote computer room, and we cannot check directly.

It should be noted here that the power stability factor should be considered, and the computer room we hosted before had several power outages, which particularly damaged the hardware life in this case.