Tutorial on maintenance and disassembly of server hardware system

In response to the phrase "it is more difficult to win the world to keep the world", server maintenance is like "keep the world "! The network is paralyzed by a slight random gain. Maintaining servers requires a thorough understanding of the hardware, operating systems, and application software systems of the servers. It is also a system engineering, hardware System maintenance is the most basic and important part of server maintenance and an endless task.

Disassembly

The simplest maintenance of the server hardware system is to modify the server and add or upgrade some components. To this end, this article begins with the server disassembly.

As the price of server accessories gradually decreases, it is not very cost-effective to work with only the standard configuration. At the same time, the server itself also provides users with a wide range of upgrade space, in this way, more and more users choose to upgrade server configurations to achieve better operation efficiency. For example, you can add a CPU, memory, hard disk capacity, a gigabit optical network card, a scsicard, a RAID card, or a RAID card to improve data fault tolerance for a multi-core motherboard. In addition, some server components, such as power supplies and fans, may also experience aging or failure due to long periods of uninterrupted operation, which also requires us to promptly replace them.

Precautions

Power off properly

For a configured server, it is easy to disable and enable it. However, if the operation method is not proper, it may cause unnecessary damage to the device. Therefore, it is necessary to give a detailed introduction to server startup and shutdown.

Shutdown operation:

1. Disable the operating system.

2. Power off the host.

3. Turn off the peripheral power (such as disk arrays and tape libraries ).

4. Power off other devices and cabinets.

5. Turn off the total power supply.

Boot operation:

1. Turn on the total power supply.

2. Power on the computer cabinet.

3. Power on external devices (such as disk arrays and tape libraries ).

4. After the external device self-check is complete, the host is powered on.

Comprehensive Prevention of static electricity

Although the computer is powered by electricity, static electricity can cause fatal damage to the computer, especially the permanent damage to the motherboard chip. There are many causes of static electricity: various power supply devices may produce static electricity, computer CRT monitors may generate static electricity, and mobile phones and walkie talkie at work will also generate a large amount of static electricity in the surrounding space, this requires us to be especially careful with the interference of static electricity and prevent it comprehensively.

1. Try to make the server and its surrounding equipment properly grounded.

2. if the device is not grounded, when we want to contact the board inside, we should first discharge it with grounding wires or other discharge facilities.

3. When operating the device, you should wear anti-static gloves. If not, you can also touch the metal outside the chassis by hand to release the static electricity carried by yourself.

Ensure proper layout in the chassis

In general, the layout and space design of the components inside the server chassis are scientifically arranged, and the combination is compact but not messy. I believe that users who have purchased professional servers should have experience. We must maintain this layout when adding and detaching server devices.

Note: there should be a considerable amount of reserved space inside the chassis to form a reasonable "air duct", because the CPU is not just the only heat source, high-speed hard drives, optical drives, various chips, and boards all generate heat in this small space. Therefore, when the fan turns, the air in the chassis must flow fully.

Operation Focus

Be careful.

We must be extremely careful when installing and detaching components for computer devices that are highly integrated and precise to servers.

1. the bottom of the CPU is called the "bridge"), which is the core component of a dense integrated circuit. These pins are prone to extrusion deformation and if one of these pins is bent, the CPU cannot work normally, so be careful.

2. the CPU slot of the motherboard is exactly the same as that of the CPU with a small needle. Therefore, if the direction is correct during insertion, you can insert the CPU to the motherboard with a single click. Do not use a lot of effort. If the CPU cannot be plugged in, there is a problem with the direction of insertion.

3. If the direction of the CPU fan is incorrect, it cannot be installed. Therefore, like inserting a CPU, do not use too much effort. Excessive power usage will cause excessive pressure between the motherboard and the CPU and cause damage.

4. the device connected to or to be connected through the cable shall not be powered on, that is, the power supply of the device should be switched off first, and then the cable connection should be plugged in and out) operation. Otherwise, unexpected damage may occur to the device if a cable is connected with a live connection.

5. When adding a server board, such as a memory or PCI device, when inserting slots, be sure to select the desired angle to ensure even power, so as to avoid damage to the Board or slot.

6. Be careful when operating the motherboard. Ensure that the installation is stable and Prevent Deformation of the motherboard to avoid damage to the electronic line on the motherboard.

Be patient

When you detach a device to a server, you must be calm. Before you start, carefully check and prepare the required tools and equipment components. Be sure to be aware of the specific operation process and precautions and be able to implement them effectively. There should be no major problems. Of course, we should not be too nervous. After all, it is just the insertion and removal of several board components. "strategically, we should despise" and enhance self-confidence. "We should pay attention to tactics, reasonable response.

1. If you plan to build a RAID, you 'd better select a hard disk of the same brand and model, which can make full use of the hard disk space and reduce the CPU usage. Once a user uses different hard disks and their respective indicators are different, CPU and other components will inevitably consume a lot of efficiency and time to process data. In addition, do not connect the two hard disks to a hard disk online in the Master-slave mode. Connect the hard disks to the RAID card IDE channel as the Master.

2. Many users have this experience when installing a SCSI device. After the device is installed, the operating system cannot identify the device, which is often caused by the lack of a SCSI device ID.

Note: ididen) is the unique identifier of a SCSI device on the SCSI bus. It must not be repeated. The optional range is from 0 to 15. That is to say, we can use a total of 15 IDs on the device.

3. When removing the device, pay special attention to the hard disk sequence in the server disk array. Many people think that the hard disk sequence in the disk array is not important, which is wrong. If you use 10 hard disks as an array, during initialization, these 10 hard disks should be placed in order in the disk array, divided into the first and second disks... To the tenth hard disk. If the disk array is ordered, the disk must be inserted back to the disk array according to the original placement order. Otherwise, the data may not match the original hard disk order, data is lost due to unrecognized data.

The above describes the problems that need to be paid attention to when adding and detaching servers. We hope to help our network administrator solve some conceptual and operational problems. Of course, if you lack experience in operations, I suggest you consult the after-sales Server service engineers. If you are involved in some difficult technical problems, choosing a professional service provider may be the best choice.