A server, also called a server, is a device that provides computing services. Because the server needs to respond to service requests and process them, in general, the server should have the ability to undertake services and guarantee services.
The composition of the server includes the processor, hard disk, memory, system bus, etc., and is similar to the general computer architecture, but because it needs to provide highly reliable services, it is Management requirements are high.
In the network environment, according to the different types of services provided by the server, it is divided into file server, database server, application server, WEB server and so on.
A server is a computer that manages resources and provides services to users. It is usually divided into a file server, a database server, and an application server. A computer or computer system running the above software is also called a server. Compared with ordinary PCs, servers require higher stability, security, and performance. Therefore, hardware such as CPU, chipset, memory, disk system, and network are different from ordinary PCs.
You can measure whether the server has achieved its design purpose from these aspects; R: Reliability; A: Availability; S: Scalability; U: Usability; M: Manageability, that is, the server RASUM metrics.
The server  must have a certain "scalability", because the enterprise network cannot be permanent, especially in the current information age. If the server does not have a certain scalability, when the number of users is not enough, a server worth tens of thousands or even hundreds of thousands will be eliminated in a short time, which is unbearable for any enterprise. In order to maintain scalability, it is usually necessary to have a certain amount of expandable space and redundancy on the server (such as disk array bays, PCI and memory slot slots, etc.).
Scalability is embodied in whether the hard disk is expandable, whether the CPU can be upgraded or expanded, and whether the system supports a variety of optional mainstream operating systems such as Windows NT, Linux, or UNIX. Only in this way can the initial investment be fully utilized in the later period.
2. Ease of use
The function of the server is much more complicated than that of a PC, not only its hardware configuration, but also its software system configuration. For a server to achieve so many functions, it is impossible to imagine without comprehensive software support. However, a large number of software systems may cause a decline in the performance of the server, and managers cannot effectively operate it. Therefore, when designing a server, many server vendors must fully consider the availability and stability of the server, as well as the ease of use of the server.
The ease of use of the server is mainly reflected in whether the server is easy to operate, whether the user navigation system is perfect, whether the chassis design is user-friendly, whether there are key recovery functions, whether there is an operating system backup, and whether there is sufficient training support.
For a server, a very important aspect is its "availability", that is, the selected server can meet the requirements of long-term stable work, and cannot often cause problems. In fact, it is equivalent to the reliability proposed by Sun.
Because the server is facing the entire network of users, rather than a single user, in large and medium-sized enterprises, the server is usually required to be uninterrupted. In some special application areas, even if there is no user to use, some servers have to work uninterrupted, because it must continuously provide users with connection services, regardless of whether they are at work or off work, whether it is workdays, rest, holidays . This is the fundamental reason why the server must be extremely stable.
In general, specialized servers work continuously 24x7, especially like some large web servers, such as those used by large companies, web servers, and iqdeWEB servers that provide public services. For these servers, maybe the number of times that they are really working and booting is only once, that is, the time when they just bought back the full installation and configuration and put it into official use. After that, it worked uninterrupted until it was completely scrapped. If something goes wrong at every turn, the network is unlikely to stay up and running. In order to ensure that the server has a high "availability", in addition to requiring the quality of each component to pass.
In addition, necessary technical and configuration measures can be taken, such as hardware redundancy, online diagnosis, and so on.
4.Easy to manage
Among the main features of the server, there is another important feature, which is the "easy management" of the server. Although we say that the server needs to work continuously without interruption, no matter how good the product is, it may fail. Take a sentence that people often say is: no
I do n’t know if it may be bad, but I do n’t know when it ’s bad. Although the server has sufficient guarantees in terms of stability, it should also have necessary measures to avoid errors, detect problems in a timely manner, and can also maintain it in time if it fails. This not only reduces the number of servers
The chance of errors can also greatly improve the efficiency of server maintenance. In fact, it is the serviceability (Serviceability) proposed by Sun.
It is impossible for a large server to stop and repair like a computer you usually use. In this way, it is likely to cause the entire network to be paralyzed, and the loss can not be measured by money. In order to solve this problem, server manufacturers have provided many new technologies, such as redundancy technology, system backup, online diagnostic technology, fault alarm technology, memory error correction technology, hot-plug technology, and remote diagnostic technology. Most failures can be repaired in a timely manner without downtime.
The easy management of the server is also reflected in whether the server has an intelligent management system, whether it has an automatic alarm function, whether it has a management system independent of the system, and whether it has an LCD monitor. Only in this way can administrators easily manage and work efficiently.
At present, according to the architecture, servers are mainly divided into two categories:
Non-x86 servers: Includes mainframe, minicomputer, and UNIX servers. These are servers that use RISC (Reduced Instruction Set) or EPIC (Parallel Instruction Code) processors, and mainly use UNIX and other specialized operating systems. There are POWER and PowerPC processors from IBM, SPARC processors and EPIC processors developed by Sun and Fujitsu, mainly Intel Itanium processors. This kind of server is expensive, the system is closed, but it has good stability and strong performance. It is mainly used in the core systems of large enterprises such as finance and telecommunications.
x86 server: also known as CISC (Complex Instruction Set) architecture server, which is usually referred to as a PC server. It is a server based on the PC architecture that uses Intel or other processor chips compatible with the x86 instruction set and the Windows operating system. Cheap price, good compatibility, poor stability, unsafe, mainly used in small and medium enterprises and non-critical business.
Dividing by application level is also commonly referred to as "dividing by server level" or "by network scale". It is the most common method of dividing servers. It is mainly based on the level of server application in the network (or server grade). Divided. It should be noted that the server grade referred to here is not divided according to the server CPU frequency, but is based on the comprehensive performance of the entire server, especially some of the server-specific technologies used. According to this division method, servers can be divided into: entry-level servers, workgroup-level servers, department-level servers, enterprise-level servers, and video servers.
This type of server is the most basic type of server and the lowest-level server. With the increasing improvement of PC technology, many entry-level servers are now configured similarly to PCs, so some people currently think that entry-level servers are equivalent to "PC servers". This type of server contains not many server features, and usually only has the following features:
1) There are some basic hardware redundancy, such as hard disks, power supplies, fans, etc., but they are not required;
2) SCSI hard disks are usually used, and now there are SATA serial interfaces;
3) Some components support hot swap, such as hard disk and memory, etc. These are not necessary;
4) usually only one CPU, but not absolute;
5) Memory capacity supports up to 16GB.
This type of server mainly uses Windows or NetWare network operating system, which can fully meet the needs of office-type small and medium-sized network users for file sharing, data processing, Internet access and simple database applications. This server is very similar to a general PC. Many small companies simply use a high-performance brand PC as the server, so this server is the same as a high-performance PC brand in terms of performance and price. Almost the same.
Entry-level servers have a relatively limited number of terminals (usually around 20), and they have poor stability, scalability, and fault-tolerant redundancy. They are only suitable for data exchange without large databases and small daily network traffic. A small business with a long, uninterrupted connection. However, it should be noted that there are currently relatively large server developments and manufacturers. There are several levels of enterprise-level servers that we will talk about later. The lowest-level enterprise-level server is called "entry level." "Enterprise server", the entry level mentioned here does not have the same meaning as the "entry level" mentioned above, but this division is still relatively small. Another point is that this kind of server generally uses Intel's dedicated server CPU chip, which is based on the Intel architecture (commonly known as "IA structure"). Of course, this is not a rigid standard, but because the server's application level requires and Price restrictions.
The workgroup server is a server one level higher than the entry-level server, but still belongs to the low-end server category. It can also be seen from this name that it can only connect as many users as one working group (about 50), the network size is small, and the server stability is not as high as the application environment of the enterprise server we will talk about below. Of course, other performance requirements are correspondingly lower. The workgroup server has the following main features:
1) Usually only supports single or dual CPU application servers (but not absolute, especially SUN's workgroup servers have workgroup servers that can support up to 4 processors, of course, the price of this type of server is also A little different)
2) SM bus that supports large-capacity ECC memory and enhanced server management functions
3) Comprehensive functions, strong manageability, and easy to maintain
4) Using Intel server CPU and Windows / NetWare network operating system, but some are using UNIX series operating system
5) can meet the needs of small and medium network users in data processing, file sharing, Internet access and simple database applications.
Compared with entry-level servers, workgroup servers have improved performance, enhanced functions, and some scalability, but fault tolerance and redundancy performance are still not perfect, and they cannot meet the application of large database systems, but the price is higher than the former. A lot more expensive, generally equivalent to the total price of 2 to 3 high-performance PC brand machines.
3. Department-level server
This type of server belongs to the mid-range server category, and generally supports symmetric processor structures with more than two CPUs, and has relatively complete hardware configurations, such as disk arrays and storage bays. The biggest feature of department-level servers is that in addition to all the server characteristics of the workgroup server, it also integrates a large number of monitoring and management circuits. It has comprehensive server management capabilities and can monitor status parameters such as temperature, voltage, fans, and chassis. Standard server management software enables managers to keep abreast of server working conditions. At the same time, most department-level servers have excellent system scalability, which can meet the needs of users to upgrade the system online in time when the business volume is rapidly increasing, which fully protects the user's investment. It is a necessary link for various basic-level data collection units in the enterprise network to maintain smooth communication with the highest-level data center. It is generally the first choice for medium-sized enterprises and can also be used in finance, post and telecommunications industries.
Department-level servers generally use CPU chips developed by IBM, SUN, and HP. These chips are generally RISC structures, and the operating systems used are generally UNIX series operating systems. Now Linux is also widely used in department-level servers.
The departmental server can connect about 100 computer users and is suitable for small and medium-sized enterprise networks with higher processing speed and system reliability. Its hardware configuration is relatively high and its reliability is higher than that of the workgroup server. Of course, its The price is also high (usually the sum of about 5 high-performance PCs). Because this type of server requires more components to be installed, the chassis is usually larger, and it is rack-mounted.
Enterprise servers belong to the ranks of high-end servers. Because of this, there are not many enterprises that can produce such servers, but also because there is no industry standard hardware to specify what level of enterprise servers need to reach, so now I have seen many companies that do not have development Companies that produce enterprise-level servers claim to have enterprise-level servers. At the very least, enterprise-class servers use a symmetric processor structure with more than 4 CPUs, some of which are up to dozens.
In addition, it also has an independent dual PCI channel and memory expansion board design. It has high memory bandwidth, large-capacity hot-swap hard disks and hot-swap power supplies, strong data processing capabilities, and cluster performance. The chassis of this enterprise-class server is even larger, generally cabinet-type, and some also consist of several cabinets, like mainframes. In addition to all the server features of departmental servers, the biggest feature of enterprise server products is that it also has high fault tolerance, excellent expansion performance, fault pre-alarm function, online diagnostics, and hot swap performance of RAM, PCI, CPU, etc. . Some enterprise servers also introduce many of the finer features of mainframe computers. The chips used in this type of server are also unique CPU chips developed by several major server developers and developed by the manufacturers themselves. The operating systems used are also generally UNIX (Solaris) or LINUX.
Enterprise servers are suitable for financial, securities, transportation, post and telecommunications, communications, or large enterprises that need to process large amounts of data, high processing speed, and high reliability requirements. Enterprise servers are used in large networks with hundreds or more networked computers that require high processing speed and data security. Enterprise-class servers have the highest hardware configuration and the highest system reliability.
The appearance of the rack server does not look like a computer, but like a switch, there are 1U (1U = 1.75 inches = 4.45CM), 2U, 4U and other specifications. The rack server is installed in a standard 19-inch cabinet. This structure is mostly a functional server.
For information service companies (such as ISP / ICP / ISV / IDC), when selecting a server, physical parameters such as the volume, power consumption, and heat generation of the server must be considered first, because information service companies usually use large dedicated computer rooms to uniformly deploy and manage large numbers Server room, the computer room is usually equipped with strict security measures, good cooling system, multiple backup power supply system, the cost of the computer room is quite expensive. How to deploy more servers in a limited space is directly related to the service cost of an enterprise. Generally, a rack server with a mechanical size that conforms to the 19-inch industry standard is used. There are also many specifications of rack-mounted servers, such as 1U (4.45cm high), 2U, 4U, 6U, 8U, and so on. 1U rack servers are usually the most space-saving, but the performance and scalability are poor, which is suitable for some relatively fixed fields of business. Products above 4U have higher performance and good scalability, and generally support more than 4 high-performance processors and a large number of standard hot-swappable components. Management is also very convenient. Vendors usually provide corresponding management and monitoring tools, which are suitable for key applications with a large number of visits, but they are relatively large and space utilization is not high.
The so-called blade server (to be precise, it should be called a blade server) refers to that a plurality of card-type server units can be inserted in a standard height rack cabinet to achieve high availability and high density. Each "blade" is actually a system board. They can start their own operating system through "on-board" hard disks, such as Windows NT / 2000, Linux, etc., similar to individual servers. In this mode, each motherboard runs its own system and serves the specified Different user groups are not related to each other. However, administrators can use system software to group these motherboards into a server cluster. In the cluster mode, all motherboards can be connected to provide a high-speed network environment, and simultaneously share resources to serve the same user group. Inserting new "blades" into the cluster can improve overall performance. And because each "blade" is hot-swappable, the system can be easily replaced and maintenance time is minimized.
In some high-end enterprise servers, due to the complicated internal structure and many internal devices, some also have many different equipment units or several servers are placed in one cabinet. This type of server is a cabinet server.
For securities, banks, post and telecommunications and other important enterprises, a system with complete fault self-repair capabilities should be adopted, and key components should be redundantly adopted. For servers used for critical services, dual-machine hot backup high availability systems or high Performance computer, such system availability can be well guaranteed.
Basic protection is only a simple four words, but there are a lot of things to pay attention to, such as the conversion of disk partition format on the server (such as NTFS format), the first time for anti-virus software (including servers and workstations) Update settings, restrict the time period for employees' computers to access the network, arrange for employees to set their own passwords for individual items on the personal computer, and more.
Comprehensive protection. Windows servers, Linux servers, Exchange Servers, Lotus Domino, Windows workstations, and Linux workstations cannot all fall down. They must provide all-round protection, and a gap may cause the entire corporate network to collapse.
Backup protection. Under normal circumstances, many network administrators can back up network server data every day, and the protection of this data can be divided into two types: physical protection and technical protection. Physical protection means that storage records are best placed in a device with certain security barriers to prevent accidental disasters. Technical protection is protected by a password. If the backup program supports encryption, this data should be encrypted. Of course, some network administrators will find too many passwords difficult to remember, and can choose some auxiliary software and tools to help.
Remote access protection. A server offering Free Remote Access (RAS) is no different to a hacker than a tourist attraction with a "welcome" sign. In view of this, the administrator can take two measures of protection. If the remote user connects to the host from a fixed place, the callback function can be used to allow the remote user to disconnect after logging in. Another approach is to restrict all remote users to a single server. The user's access is then restricted to one server, not the entire network. This way, even if there is a hack, it will be isolated on a single machine.
Step out of regular protection. Don't stick to the rules in network deployment. If you use the methods and deployment methods that all novices and newcomers are accustomed to, the chance of being successfully attacked by hackers is naturally much greater. For example, RAS is the most commonly used TCP / IP protocol, but in fact IPX / SP protocol can also be used. If network administrators choose this unconventional protocol, it can indeed increase the difficulty for hackers to invade corporate data.
Yishang has 23 years experience in manufacturing and customizing metal server rack and server cabinet. Welcome to contact our email firstname.lastname@example.org if you need our help.
Hello, please leave your name and email here before chat online so that we won't miss your message and contact you smoothly.