How to select servers? Which brand is the best? (Less than 10 thousand RMB)

First, we must understand what a PC server is? The so-called PC Server is an Intel-based server. Unlike some large servers, such as mainframe and UNIX-based servers, most of them run Windows or Linux operating systems and are generally used, the latter is mostly for professional purposes, such as banking, large manufacturing, logistics, securities... In other industries, the average person has little chance of access. Generally, if the PC server is out of type, it can be roughly divided into three categories:

(1) Indexing Server (Tower server ):

As shown in:

It can be used for servers that can be placed independently on the desktop or on the ground. Most servers have a large number of expansion slots and hard disk space. No additional equipment is required and can be used after power-on. Therefore, it is the most widely used.

(2) rack-mounted servers.

A server that can be installed with cabinets is mainly used to save space. The height of the server is 1u, And the 1u is about 44mm. Due to limited space and Limited scalability, for example, most 1u servers have only one to two PCI expansion slots. in addition, the heat dissipation performance has become a very important factor. At this time, the power of each manufacturer is shown here. The disadvantage is the need for organic cabinets and other equipment, which is mostly used by enterprises with large server usage.

(3) blade servers:

This is a more space-saving product than rack-mounted servers. the main structure is a large main chassis. Many cards can be inserted inside, and a card is equivalent to a server. of course, the heat dissipation is very important here, and various manufacturers often install large powerful fans for heat dissipation. although this type of server saves a lot of space, it may be free of charge only for the main chassis, and is rarely used except for large enterprises.

After learning about the types of servers, how can I purchase a suitable PC server? We can consider the following aspects:

(1) stable performance:

This is one of the most important factors for servers. however, for many people, "stability" seems to be a very abstract term. It seems that every server vendor is emphasizing that its products are very stable. in fact, "stability" is not completely traceable, or expensive products are stable. I provide some tips for reference: AAA

(2) Overall assembly quality:

Generally, products assembled by large manufacturers have certain quality control and manufacturing processes. Therefore, if you can open the chassis for observation, it is not difficult to find that if the wiring is messy, the chassis materials are thin, components with poor matching or CPU, memory and hard disk without original warranty stickers should never be included in the scope of consideration.

(3) Good heat dissipation design:

Most servers need to operate for a long time, so good heat dissipation performance is very important. the heat dissipation performance can be determined by the manufacturer's data, the heat dissipation wind strength or the actual test. servers with good heat dissipation often have better stable performance. for example, Asus rack-mounted server AP1600R-S5 with no wire design, heat dissipation performance excellence.

(4) Commitment to after-sales services:

Manufacturers who have confidence in their products usually provide better service content. For example, the standard service of Asus servers is a three-year limited warranty service and the on-site service is provided within one business day; there are also a variety of value-added services available for you to choose from.

(5) overall reputation:

Generally, the reputation of server products is very important. It is also a way to choose recommended brands or old brands in the market. however, sometimes there will be new brands or products that are also very good, which depends on the recommendations of some experts or trial tests.

(6) Authoritative evaluation recommendations:
Some authoritative magazines often have some ratings and are also a reference basis. However, the most important thing is to look at some actual performance tests and compare the reports to obtain objective opinions.

(7) actual test:

If possible, it is best to purchase a small number of products for testing, install the software you want to use, and run it for a long time. the server should be able to run for a long time. It is best to download some test software on the website. What is the actual operation to observe its stability?

(8) upgrade and maintenance costs:

Many brand servers may not have a high total price at the time of purchase, but they have considerable upgrade and maintenance costs. for example, for servers of some foreign brands, we often start to have low cost pressure to increase market share, but in some parts of the accessories that will be upgraded in the future, such as CPU and memory, hard Disk, disk array card... And so on. In addition, after the expiration of its original warranty period, the maintenance cost of its renewal is also very expensive, resulting in a dilemma similar to "It is easy to buy a car to raise a car. therefore, this is also a part to consider.

(8) Manufacturer R & D and manufacturing strength:

Many manufacturers have launched PC Server products, some of which are very cheap, but users should really consider the strength of the manufacturers themselves. generally, international brands are more powerful and have rich R & D experience. It is important that the vendor should have experienced R & D teams. however, due to fierce competition in recent years, the price of PC servers has dropped a lot. Therefore, most of the foreign-owned PC servers are OEM by other manufacturers. Therefore, when considering these brands of servers, we should exclude the superstitious knowledge of the Brand about the power and R & D capabilities of the foundry to consider the purchase criteria. After all, what we actually buy is the products of the foundry, just the brand.

(9) ability to solve problems:

Although many foreign companies are not doing well in R & D technology, most R & D departments are outside China. When a customer encounters a problem, the local sales base should respond to the foreign headquarters, in this way, the feedback speed will inevitably be affected, and the R & D organization levels of domestic manufacturers are not so complex, which can be prioritized.

Looking at the above points, the most important thing to buy a good PC server is to meet your own needs, as well as stability and after-sales service assurance, of course, the most effective way, undoubtedly, the architecture is a simulated environment. It is the safest way to test and operate on the ground. consumers should also stand on a more rational standpoint and evaluate them based on various key points. Do not blindly trust the brand, but purchase a very expensive product, the quality is not much higher than the average Brand's server

 

 

 

 

 

 

Advantages and disadvantages of three types of server Structures

Generally, servers are divided by three structures: rack, tower, and blade. It is divided by the shape of the server, because it is applied to different environments. The following describes the differences among the three servers.

　　I. Tower server

Advantage: the shape and structure of the Tower server are similar to those of the vertical PC we usually use. Because of the strong scalability of the server motherboard and a pile of more slots, therefore, the head is larger than the general motherboard. Therefore, the host chassis of the Tower server is larger than the standard ATX chassis. Generally, sufficient internal space is reserved for redundant expansion of hard disks and power supplies in the future.

Because the Tower server has a large chassis, the server configuration can be very high, and redundant expansion can be more complete, it has a wide range of applications, it should be said that one of the most frequently used servers is the Tower server.

Disadvantages: at present, common entry-level and working-level servers basically use this server structure type. However, because there is only one host, there is a limit even for upgrading and expanding, therefore, in some enterprises with high application requirements, single-host servers cannot meet the requirements and require multi-host collaboration. The tower servers are too large and independent, collaborative work is inconvenient in space occupation and system management, which is also the limitation of tower servers. However, in general, the functions and performance of such servers can basically meet the requirements of most enterprise users, and the cost is usually relatively low. Therefore, such servers still have very extensive application support.

　　Ii. Rack-mounted servers

Advantage: as a server model designed for the Internet, rack servers are servers designed according to uniform standards and used together with cabinets. It can be said that the rack-mounted Server is a tower server with an optimized structure. Its design purpose is to minimize the occupation of server space, the direct advantage of space reduction is that the price will be much lower when the data center is hosted.

Many specialized network devices use a rack-mounted structure (mostly flat, like a drawer), such as switches, routers, and hardware firewalls. The Rack server is 19 inch in width and the height is U (1U = 1.75 inch = 44.45mm). There are usually 1U, 2U, 3U, 4U, 5U, 7U) Standard servers. The cabinet size also adopts general industrial standards, usually ranging from 22U to 42U. The cabinet has a detachable sliding drag frame based on the height of U. You can flexibly adjust the height according to the height of your server, to store network devices such as servers, hubs, and disk cabinets. After the server is placed, all its I/O lines are drawn from the rear of the Cabinet (all the interfaces of the Rack server are also located at the rear), which are uniformly placed in the cable slot of the Cabinet, it is usually labeled to facilitate management.

Disadvantages: rack-mounted servers are much smaller than tower servers, so these servers are limited in scalability and heat dissipation, and accessories must be filtered, generally, the expansion of devices is not complete. Therefore, the performance of a single machine is limited and the application scope is limited. You can only focus on some applications, such as remote storage and Web service provision.

Iii. blade servers

Advantage: The Blade Server is a low-cost server platform with High Availability and High Density. It is specially designed for special application industries and High-Density computer environments. Applicable to digital media, medicine, aerospace, military, communication, and other fields. Each "Blade" is actually a system motherboard. They can start their own operating systems through a local hard disk, such as Windows NT/2000, Linux, Solaris, and so on, similar to independent servers.

In this mode, each motherboard runs its own system and serves different user groups and is not associated with each other. However, you can use system software to assemble these boards into a cluster server. In cluster mode, all the mainboards can be connected to provide a high-speed network environment, and resources can be shared to serve the same user group. Insert a new "Blade" into the cluster to improve overall performance. Because each "Blade" is hot-swappable, the system can be easily replaced and the maintenance time can be minimized. It is worth mentioning that the system configuration can be achieved through a set of smart KVM and 9 or 10 CPU boards with hard disks. The CPU can be configured as different subsystems. Servers in a rack can share a set of optical drives, soft drives, keyboards, displays, and mouse with the new Smart KVM Conversion Board to access multiple servers, this allows you to easily upgrade, maintain, and access files on the server.

 

 

 

 

 

 

The best server isIBMWith excellent quality and powerful R & D and technical strength.
In additionHPThe server is also characteristic and personalized, And the quality is quite good.
DELLServer to be cheaper and cost-effective as the main selling point.

Otherwise
Inspur, sugon, and Lenovo are the main server brands.
Inspur and sugon are mainly known for their R & D and technical strengths. inspur focuses on Intel and sugon on AMD.
Lenovo is still a PC sales team. Forget it.

 

 

 

Why do I host a server after I buy it ???  

I don't understand. Why do I have to contact a drag-and-drop manager after I buy a server? Isn't the server the same as a computer? Can I leave it at home? Be sure to look after others. What is the reason? Please give answers.

Question added:

What is a fixed IP address? I don't understand it at all. I hope it won't be looked down. What is the requirement of the server on the network speed. Please give a detailed answer with one hundred points.

 

Best Answer:

 

1. The IP address is equivalent to the server's ID card on the network. It is unique. You can use this ID card to identify your server and access the content on your server. Most of the households are dynamic IP addresses, which are not fixed and therefore not suitable for servers. 2. Most of the bandwidth used by home users is ADSL bandwidth, which is characterized by slow downloading, unstable uploading, and frequent disconnection. for servers, there are many interactive accesses, therefore, the upstream and downstream bandwidths must be fast and cannot be dropped. 3. the household's power supply is unstable, and may result in power outages, TRIPS, and other phenomena, affecting the stability of the server or even causing software and hardware damage. The data center has professional power supply equipment and relay equipment and will not be powered off, will not trip. 4. the home computer is usually shut down after it is used up, and the server is usually not shut down once a few months or even a year, which requires a high ambient temperature and humidity and is hot in summer, if the server is not cooled by an air conditioner, it may cause overheating or damage. Hosted machine rooms are equipped with professional high-power air conditioners to ensure constant temperature and humidity, and ensure machine life. 5. If you host your website at home, you have to do everything by yourself. If you do not understand it, you need to check the information everywhere. If you make a mistake, you may face data loss or hardware faults, the data center is equipped with professional technical engineers to handle faults very efficiently. 6. Professional data centers have professional qualifications. Looking for formal data center hosting will benefit a lot. If you find an informal data center, you can leave it alone.

========================================================== ==================================

 

Have you ever seen the data center? If you do this at home, it's a waste. Humidity, dryness, ventilation, heat dissipation (the data center has a dedicated data center freezer), cleanliness, network speed, and you are providing services. Your stable performance is not only the stability of the computer, is the network stable? Do you want to buy a regular route in your house? Enough for you to buy several computers. One person looks at your computer, and two people look at your computer. What about 100 people? I will not blow up your home route. Is your network speed enough? The size of a person's traffic is several KB. It is estimated that the person cannot even log on. Who else can serve the person. As for a fixed IP address, as the name implies, it is a fixed IP address. Because China does not have one IP address per person, dynamic IP addresses are used. That is to say, you will give you an IP address when you access the internet. And then give it to you.

 

========================================================== ==================================

 

The bandwidth stability at home is not good. Of course, if the home has UPS and fiber-optic leased line access, the data center can have no other infrastructure, that is, the air-conditioning fiber backbone network can provide uninterrupted power supply and power generation equipment, and the home can pull a small optical fiber. UPS can also

 

========================================================== ==================================

 

Of course not at home: 1. Home broadband does not have a fixed IP address. Not suitable for websites. IP address replacement requires domain name resetting 2. The home broadband speed is too low. In theory, 10 m lan is used. Cross-region access is also quite slow. Website creation is critical. A fixed IP address is a fixed IP address. It will not be changed due to reboot and power failure. The server is highly reliant on broadband. A simple example is. You need to visit a website. It takes dozens of seconds to open a webpage. If not necessary. I believe you will choose to directly close this website. It is estimated that no access will be made in the future.

 

========================================================== ==================================

 

The landlord is well-founded. vendors provide a lot of hardware environments to ensure the stable operation of your servers, such as: Power (uninterrupted) temperature (constant temperature and humidity) bandwidth (backbone network bandwidth high-speed network) scalability and so on, so they are all placed in the IDC Room 2, of course, there are also servers in the company or their own homes, of course, there will be loss and convenience of maintenance, but stability will be somewhat discounted 3 fixed IP addresses are designed to allow others to access you, broadband also provides fixed IP addresses upstairs said wrong! However, it is very expensive to provide broadband with fixed IP addresses.

 

========================================================== ==================================

 

 

From: http://zhidao.baidu.com/question/140590026.html