1. Project content
The main goal of this plan is to meet the needs of most campus network applications in a school campus LAN. On the basis of existing independent information classrooms, the school office building, teaching building and auxiliary auxiliary management department The main buildings form a campus network. The main contents included are:
1) Formulate a reasonable campus network overall construction plan and implementation plan based on the school's needs for information point arrangements and network applications;
2) Implement structured local area network cabling for several major buildings such as school office buildings, teaching buildings and auxiliary auxiliary management departments;
3) Complete the laying of two six-core outdoor backbone optical cables from the network center of the office building to the teaching building and the bungalow where the ladder classroom is located.
4) Realize the connection, installation, configuration and debugging of the school's core switches and secondary switches;
5) Implement connection and network debugging of newly purchased servers and some microcomputer network workstations.
2. Structured wiring design for a school campus LAN construction project
A structured wiring system is a transmission network within a building or group of buildings. It can not only connect data communication equipment, switching equipment, and other information management systems to each other, but also connect these devices to external communication networks, including buildings to the outside. The connection point on the network line, all cables between the data terminals in the work area, and the associated wiring components. A good structured wiring system has a certain degree of independence for the equipment it serves, and can interconnect many different communication devices such as data terminals, PCs and hosts, and public system devices. The general wiring system consists of six subsystems: inter-building subsystem, equipment subsystem, management area subsystem, vertical (backbone) subsystem, horizontal subsystem, and work area subsystem.
The campus network is a campus network, and the subsystems between the buildings are connected by optical cables, which can provide gigabit bandwidth and have sufficient room for expansion. The vertical subsystem is located in the shaft of high-rise buildings, and multi-mode optical cables or large-pair twisted-pair cables can be used. Integrate the management area subsystem into the equipment room subsystem for centralized management.
For multiple buildings, multiple equipment rooms can be used. It is divided into the central equipment room and the building equipment room. The central equipment room is the control center of the entire local area network. There are various network devices (switches, routers, video servers, etc.) for external (Internet) internal communication. The central switch passes The optical cable is connected to the switching equipment between the building equipment to ensure high-speed data transmission. Place wire racks and network equipment between this equipment, terminate the backbone cables from each floor in the building, and terminate the optical fiber connected to the network center.
The wiring in the building includes horizontal wiring and backbone wiring. The horizontal system adopts super five kinds of twisted pairs, the new building adopts the method of concealed installation in the wall, and the old building adopts the method of open installation of PVC wire duct.
This solution is proposed by the network company according to the basic requirements of a school (hereinafter referred to as the school) on the computer network and in accordance with the international commercial building wiring system standard (EIA / TIA 568B, ISO11801) to meet the application of the campus network. The goals are:
Realize 60 information points in the school office building, 136 information points in the teaching building, and 26 information points in the bungalow (a total of 222 information points, structured cabling in a LAN that exceeds Category 5 wiring standards.
According to the design requirements of EIA / TIA 568B (International Commercial Building Wiring System Standard), not only must you select wiring materials and equipment that meet international standards, but also design and construction in accordance with their standards. According to the requirements of the bidding party, we plan to select the products and technologies of the well-known wiring manufacturer that meets international standards-American AMP Company.
2.1 User demand for information point distribution and bandwidth
According to the requirements of the school, there are 222 LAN information points implemented this time. Among them, 60 office buildings are distributed on 3 floors, and 136 teaching buildings are distributed on 5 floors. The 26 information points of the bungalow are used for the ladder classrooms and auxiliary management departments of the school. The buildings are connected by multimode optical cables to form the campus LAN of the school.
The central computer room of the school LAN is located on the west side of the third floor of the office building. The wiring room of the teaching building is located on the north side of the east side of the third floor.
According to school arrangements, the specific distribution of information points is:
Office building Teaching building Bungalow Total
One floor Two floors Three floors One floor Two floors Three floors Four floors Five floors
22 20 18 34 30 30 28 14 26 222
60 136 26 222
In order to meet the school's future needs for various network applications, in addition, with the advancement of technology and technology, considering that the current various wiring materials are relatively close in price, it is proposed to route all information points according to the UTP UTP standard. Each information point can achieve a bandwidth of not less than 100Mb, which not only supports the network transmission bandwidth requirements of general school information management applications, but also fully supports the transmission of multimedia information in MPEG-2 and other formats.
This wiring scheme only considers data information point wiring, and does not involve voice information points and their equipment.
2.2 Integrated wiring design in the building
The characteristics of the school office building and teaching building are: without the elaborate decoration of the wall panels and ceilings. Therefore, it is proposed to lay open trunks in corridors and rooms. For aesthetic reasons, the trunk trunks of the horizontal subsystems can be concealed as much as possible. Only use the open trunks when it is unavoidable.
The physical size of the main part of the office building is about 35 meters long and 14 meters wide, with a total of 3 floors. The central computer room is located on the west side of the third floor, and the furthest information point to the east end of the first floor does not exceed 60 meters.
The teaching building is about 70 meters long and 14 meters wide, and has 5 floors. The secondary switch cabinet is located at the east end of the third floor. The information point farthest from the cabinet is the south room on the west side of the first floor. The longest distance is about 91 meters. The building is a building that has been used for many years. There are both wired closed-circuit feeders and telephone lines in the corridor of the open channel and structured wiring of the old local area network. The newly added trunking must avoid various situations that may cause interference. It is also necessary to maintain the appearance of the wire channel. Therefore, the wiring design and process need to be carefully handled.
According to the requirements of EIA / TIA 568 and ISO11801, the length of the twisted pair between the active device and the active device should be kept within 100 meters. Of course, when using Category 5 twisted pair, with 10Mbps bandwidth, individual information points can reach about 130 meters under the last circumstance (some network product manufacturers promise to reach 200 meters). In this solution, the distance from any desktop information point to the wiring room will not exceed 94 meters (the actual maximum distance does not exceed 91 meters), which fully meets the requirements of the technical specifications.
The main structured wiring materials (distribution frames, UTP cables, information sockets, plugs and other key components) of this project are selected from American AMP products. The product is quite competitive in the market due to the following characteristics:
AMP is the world's oldest connector production plant. Its product life, reliability and electrical performance have been tested for decades. Most computer manufacturers and communication equipment manufacturers use AMP's connectors as the cable connections for their equipment.
AMP's Netconnect product line includes all products that cover structured wiring systems and fully complies with international standards.
The AMP company's panel is square in shape and milky white in color, which is more harmonious with the domestic commonly used power sockets, switches and telephone sockets.
The socket on the panel is provided with a sliding dust cover. After the plug is pulled out, the cover automatically bounces off, which improves the cleanliness of the information socket.
AMP's Netconnect products have a good performance-price ratio. Similar products, AMP brands are generally more than 10â„… lower in price in the domestic market than AVAYA products. LAN planning topology diagram is as follows:
2.3 Links between buildings
2.3.1 Comparison of several commonly used transmission media
The connecting cables between the buildings are exposed outdoors, and are affected by the sun and rain forests and the interference of thunder and lightning all the year round. From the perspective of several commonly used wired transmission media such as twisted pairs, thin coaxial cables and thick coaxial cables and optical cables, From the aspects of anti-interference, anti-corrosion, high bandwidth and allowable transmission distance, outdoor armored optical cable is the most suitable transmission medium. Here are some analysis and comparison of these commonly used transmission media:
(1) Twisted pair
At present, the commonly used twisted pair in indoor wiring refers to unshielded twisted pair copper wire (UTP), and its transmission rate is closely related to the distance. Although within 100, five types of unshielded twisted pair can transmit 125MHz (100Mbps), but it The main cables between buildings are not suitable for cooperation. The reasons are:
Copper wire without a shielding layer can easily induce lightning and cause interference outdoors, or even damage the equipment.
UTP is prone to aging and has a short life due to the bad outdoor environment.
According to the PDS wiring rules, UTP cables are not allowed to exceed 100 meters (especially 100Mb Fast Ethernet), so it is not suitable for outdoor backbone cables.
As for shielded twisted-pair copper wire (STP), in theory, anti-interference and transmission distance are better than UTP, but it requires a higher grounding performance, and it is generally difficult to guarantee from the process, even if it is done, its investment Very impressive, otherwise it will be self-defeating. This is also the main reason why STP has not been widely popular.
(2) Coaxial cable
Traditional coaxial cables mainly include thick coaxial cables and thin coaxial cables. The thick coaxial cable has higher tensile strength than the thin coaxial cable, good mechanical properties and longer effective transmission distance (the maximum distance of the former is 500 meters, the latter is 185 meters), but the construction is difficult, and Due to fewer and fewer applications, it is difficult to find sources of goods on the market, and its cost has exceeded that of optical cables. The transmission rate of both the initial and thin cables is 10Mbps, and can only be used as a shared medium. However, due to the convenience of thin cable construction and the low price, if the number of workstations on the shared network segment is not large (within a dozen sites), indoor use can be considered. But reliability and maintainability are slightly worse. With the development of network technology, these two transmission media have been gradually eliminated.
(3) Optical cable
The optical cable has high bandwidth (generally up to hundreds of MBit or even dozens of Gbit level), long transmission distance, strong anti-interference ability and good security. The prices of its related products have also dropped significantly year by year. At present, it is the ideal transmission medium for the industry as the backbone of the network.
For a school campus network, it involves the inter-building LAN interconnection of 3 buildings. Because the network center is located in the office building, according to the star-shaped network topology, the characteristics of the current centralized application mode of client / server (C / S) or browser / server (B / S) are considered, so the other two The buildings should be connected to the network center of the office building.
As far as most campus network applications are concerned, among the transmission information carried on the campus network, the transmission volume of multimedia information will be larger and larger, such as multimedia teaching, electronic reading, video on demand and other applications. Therefore, no matter from the current or future development point of view, the backbone transmission medium must have the ability to carry Gigabit rates. In addition, as an outdoor transmission medium, it also needs to have good anti-interference, reliability, anti-aging and high life characteristics. However, the transmission media with these characteristics can only meet the requirements with optical cables.
According to the mode, the optical cable can be divided into multi-mode and single-mode. They can carry Gigabit transmission rates, but the distance between them is quite different. The former has a maximum transmission distance of only 275 meters (62.5 / 125μm) to 550 meters (50 / 125μm) due to its optical transceiver Semiconductor devices (LED), so the price is relatively cheap; and because single-mode fiber uses a laser device as an optical transceiver, the transmission distance can reach tens of kilometers, and the price is several times higher than the former. For this solution, since the distance of the farthest optical cable between buildings will not exceed 200 meters, it is advisable to use a 62.5 / 125μm multimode optical cable to save investment.
2.3.2 Outdoor backbone link design
Topology solution
According to the needs of the school, it is necessary to interconnect the indoor information points of three buildings such as office buildings, teaching buildings and bungalows into the campus LAN of the school. From the convenience of management and maintenance, it is conducive to increasing the speed of all information points to access the central server, and considering the geographical location as a whole, a better topology solution is: take the network room of the office building as the center, and radiate to other 2 in a star structure Building.
Medium selection for interconnection between buildings
As mentioned in the previous analysis, for the outdoor backbone transmission medium of the current campus network, the optical cable has become the only effective choice. It not only has a good bandwidth and its scalability, but also is very superior in many aspects such as anti-interference, reliability, stability and service life. The disadvantage is that the construction is slightly more difficult and the cost is higher. For the former, as long as the design is well-considered and the construction is in place at once, sufficient forward-looking and sufficient scalability are considered, this problem can be solved well. To this end, it is planned to use multi-mode 6-core optical cables for the outdoor backbone links from the office building network center to the other two building buildings (the teaching building has more information points, you can consider 2 gigabit uplink links, or press LAG The way is bundled into a 2 Gigabit backbone link. Once any link fails, the other link will automatically run as a Gigabit rate link. There are also two cores for backup). This enables high-quality, redundant, reliable, and stable high-quality backbone links.
There are currently at least three methods for the connection of optical cables: welding, grinding and cold compression. Although the grinding method has a certain application market in multi-mode optical cables, experience has shown that the consistency is poor, often different from person to person, and the attenuation is generally large. Splicing is easy to ensure performance, especially in single-mode optical cables. Cold compression bonding is a new process that requires special equipment, and its performance is not as good as welding. We generally use fusion cable connection method.
3. Bandwidth analysis
3.1 Consideration of backbone bandwidth
The backbone network is actually the connection channel between the central switch and the secondary switch (for a network environment that is not too large, that is, the access level) and between the central switch and the server. For switched Ethernet, there are three types of bandwidth: 10M, 100M and 1000M. For this solution, according to the network scale, application characteristics and the will of the school, the backbone bandwidth is planned to be considered as 1000M. Between the central switch and the server, you can choose between 100M or 1000M according to the size and application type of the server, and the 1000M full-duplex mode should be adopted between the central switch and the access switches in office buildings, teaching buildings, and bungalows.
For a local area network that is not very large and the area is not too scattered, it is appropriate to reduce the number of cascades of switches or hubs as much as possible to avoid increasing delay. For this program, there are only two levels.
3.2 Client bandwidth analysis
The client's bandwidth is generally shared 10Mb, exchanged 10Mb and exchanged 10 / 100Mb. Although the former has low cost, the network performance is poor and it is difficult to meet the application requirements. Most of the network cards currently available on the market are 10 / 100Mb adaptive. The client applications for teaching buildings and office buildings may be mostly school management, but there is no shortage of multimedia information transmission such as low-speed surgery and CT images. Moreover, the price of the 10Mb network card and the 100Mb network card on the market is almost the same. Therefore, it is planned to configure all the clients of these two buildings to be adaptively switched to the desktop by 10 / 100Mb (provided that all inter-building connections are Module cable).
Back to Contents
4. 10 / 100Mb switch product selection
For the selection of the central switch, it is important to consider its switching capacity, scalability, and has good second-layer performance and third-layer functions. However, in selecting the grade and scale of the core switch, we must combine the school's network scale and application characteristics, and leave room for expansion appropriately. However, buying network products now is not tailor-made after all, but chooses the grade and model as close to the required as possible from the existing products of various manufacturers in the market according to their own needs. If the selected configuration is too high, it will cause a waste of resources, and the configuration is too low, not only does not have room for expansion, but also affects network performance. However, when designing the scale and performance of various grades of products, different network manufacturers will produce large differences according to their own technology and technological level. In addition, for products with similar performance and configuration, the prices of different countries, brands, and manufacturers vary greatly. Generally speaking, the products of well-known foreign brands are much higher than those of domestic ones. Even domestic network products have several grades. Therefore, there are trade-offs to consider when choosing a network product.
At present, the main brands of foreign LAN products that are more popular in China are: CISCO, AVAYA, 3COM and so on. These products are characterized by mature technology, stable performance, reliability, and rich functions. But the price is relatively high. Such products are widely used in financial, securities and large enterprises with high requirements. It is worth noting that in recent years, many emerging network product manufacturers have rapidly emerged in China, such as Digital China, Huawei, Gangwan, TCL, and Shida. The products of these manufacturers are comparable in function and performance to similar foreign products, but they are more flexible in terms of configuration and port combination, especially in terms of price. The core switch with high port density reflecting the advanced technology not only cannot be used in this solution, but also will cause a greater degree of port waste. Therefore, in view of the needs and environment of the second normal school, we believe that the domestic network products of China Digital are more suitable. The modular core switch port density is not very high, but the switching capacity is higher and the performance-price ratio is higher.
In this scheme, the core switch selects DCRS-6512. The switch is suitable for small and medium-sized campus networks and campus networks, and provides excellent performance and reliability based on leading technologies. The DCRS-6512 switch is designed to take advantage of the potentially huge switching capabilities of Gigabit Ethernet, and its non-blocking structure can ensure that each port easily has full line-speed switching capability, ensuring line speed under huge traffic and network load The second layer and the third layer are exchanged. Can be used as an ideal network core switch.
The DCRS-6512 chassis itself can provide 12 user slots for I / O switching, and can provide up to 24 Gigabit ports and 96 100M ports. The switching capacity of the whole machine is 48Gbps, which can achieve 36Mpps line-speed full-layer packet forwarding rate. The configuration and combination of its modules are very flexible. In this scheme, an 8-port Fast Ethernet copper cable module is selected for connecting to a 100M rate PC server; at the same time, two 2-port Gigabit Ethernet copper cable modules are also provided, providing 4 gigabits. The copper cable ports can be connected to the two access switches of the office building in a dual-link aggregation mode, which can both expand the bandwidth and realize automatic link backup; if necessary, it can also be used in the future to connect a specific server Gigabit network card. In addition, two 2-port Gigabit Ethernet multimode fiber optic cable modules were selected, which are mainly used to connect the secondary switches of the teaching building and the bungalow with multimode fiber. Its functions and performance fully meet the various network applications of the school.
The secondary switch selects DCS3426 and DCS3628S. They are all Gigabit network-connectable secondary switches with 10 / 100Mb rate access. The former is used to connect office buildings and bungalows to core switches independently. The latter is a stackable switch, mainly used in teaching buildings. Due to the dense port of the teaching building, in order to save optical fiber links and Gigabit uplink ports, the switches are now stacked together with a 4Gb bandwidth (maximum 6 stackable), and then two Gigabit links are aggregated Connect to the core switch. The following table lists the main functions and performance of core switches and secondary switches:
Switch type DCRS-6512 DCS-3426 DCS-3628S
Backplane speed 48 Gbps 18 Gbps 18 Gbps
Forwarding bandwidth 36Mpps 148800 0pps / port 1488000 pps / port
Number of MAC addresses supported 32k 12k 12k
VLAN support 802.1Q 802.1Q 802.1Q
Number of VLANs / Dynamic VLAN 256/4096 256/4096 256/4096
STP 802.1d 802.1d 802.1d
QoS 802.1p, 4 queues 802.1p, 2 queues 802.1p, 2 queues
Stacking support No No Yes / 4Gb
FEC (LAG) support Yes, 8 ports / bar, 12 groups Yes, 8 ports / bar Yes, 8 ports / bar
Maximum 10 / 100Base-TX 96 pcs 24 pcs 24 pcs
Maximum 1000Base-SX 24 pcs 2 pcs 2 pcs
Maximum 1000Base-T 24 pcs 2 pcs 2 pcs
Expansion slot supports 12 2 3
SNMP MIB II support Yes Yes Yes
5. Completion and acceptance
1. The information socket and power socket are installed in place, neat and beautiful.
2. The main wiring system is installed in place, neat and beautiful.
3. Layout of wire channel and cable is neat and beautiful.
4. The system connection is neat and beautiful.
5. The quantity and model of equipment materials are consistent with the design.
6. Provide 100% information point test and performance test report.
7. The installation of the entire network system reaches an arbitrary point of information extraction, which can achieve network interconnection (such as WIN98 / WIN2000 interconnection)
Network equipment selection
The prices of different network equipment vary greatly. Let us first briefly understand the network equipment.
HUB: This is the so-called hub. It can be divided into several types, including common HUB, stacking HUB, port switching HUB and so on.
The bandwidth of 100M HUB is shared, that is to say, the 24 ports of the 24-port HUB share 100M bandwidth. If 24 ports transmit data at the same time, then the bandwidth of each port is only about 10M. The stacking HUB is a stackable HUB, that is to say, if we need to network 48 machines, we can use 2 stacking HUBs to stack up as a 48-port HUB.
Switch: It can be considered as a high-performance HUB. Its 100M bandwidth is independent, or it allows several ports to transfer data at a speed of 100M at the same time. The switch usually also has a routing function.
The network center is the core of the company's network. In order to avoid possible collisions on the network, our core equipment is a 350T switch from Bay. It adapts to the 10 / 100M network with routing functions and has good overall performance. All HUBs are directly connected to the switch, and important servers are also directly connected to the switch, so that the advantages of fast switch speed and high bandwidth can be fully utilized.
The HUB we chose is Intel ’s Express 100TX-BASE stack HUB (Intel is not willing to only do CPU, it also makes network cards, graphics cards, maybe it will be a chassis someday. :-), it has 24 ports , Can be stacked. Made using Bay's technology, the price is good.
Note: There is a small button on the left of the first port of the HUB. Press it to make 1, 2, and 3, 6 of the first port cross. This port is a port dedicated to the connection between the HUB and the HUB or switch. In fact, this is how our company's HUB is connected to the switch.
Network card: Every computer needs a network card. We initially selected 3Com's 3C905, 10/100 adaptive network card. We also bought a few Intel's 82557. After a period of use, one of the Intel's network cards broke. In comparison, 3Com's network card quality is better than Intel's network card. Later, we used the DFE-500TX network card from D-Link, and the price / performance ratio was quite good.
Network wiring system: Choose five types of wiring system of AMP company. When making the network cable, pay attention to that it is not simply to connect the 8 wires of RJ-45 one by one. You must ensure 1, 2 twisted pairs, 3, 6 twisted pairs, 4, 5 twisted pairs, 7 and 8 pairs If the twisting is only for one-to-one connection instead of guaranteeing 1, 2 and 3, 6 twisting, it may cause the site with a long network cable to become unstable or even unable to work normally.
Network configuration, construction server settings: There are a total of 2 servers on the LAN, one of which is used as an internal file server. The other is used as an Internet server. The Internet server runs Windows NT + IIS + Exchange Server and provides WWW, FTP, and Email services.
Construction: When calculating the length of the network cable, it is necessary to reserve a 10% margin to avoid the necessary bypass and other unpredictable situations in case of the structural reasons of the building.
An integrated wiring system is not so much a computer project as a construction project. The actual performance has a lot to do with the installation process. During construction, pay attention to that the network cable cannot withstand excessive curvature and avoid close to strong interference sources and building subsystems ( That is, the network cable connecting the two buildings must be strengthened. We use steel pipes for this part of the network cable. The advantages of this are: high strength and strong anti-interference ability.
IP address allocation: According to the relevant provisions of RFC1597, in order to facilitate future connection to the Internet and to consider the development of the campus network, it was decided to use a Class B network inside the campus, with a network number of 172.16 and a corresponding subnet mask of 255.255.0.0.
Computer name naming rules: department code + serial number, the mantissa of the IP address is consistent with the mantissa of the computer name. For example, 172.16.1.1 ==> Technical Department rd1.
Understand the IP address and subnet mask. I ca n’t help thinking about the subnet mask here:
We know that an IP address is a dotted decimal number, and each IP address is composed of two parts: a network number and a host number. The network number marks a physical network. All hosts on the same network need the same network number, and the network number is uniquely determined on the Internet. The host number determines a TCP / IP host such as a workstation, server, router, etc. in the network. For the same network, the host number is unique. Through the network number + host number, we can determine the location of a host on the Internet.
Since the network number + host number can determine a host, what is the use of subnet mask?
In order to adapt to networks of different sizes, the Internet defines 5 types of IP addresses:
Class A addresses: the highest digit is 0, the next 7 digits represent the network number, and the remaining 24 digits represent the host number. A total of 126 networks are allowed, with approximately 17 million hosts per network.
Class B address: the highest 2 digits are 10, the next 14 digits are the network number, and the remaining 16 digits are the host number. It allows 16384 networks, and each network has about 65000 hosts.
Class C address: the highest 3 digits are 110, the next 21 digits are the network number, and the remaining 8 digits are the host number. It allows 2 million networks, and each network has about 254 hosts.
Class D address: the upper 4 bits are 1110, used for multicast.
Class E address: High 4 is 1111, only for experiment, reserved for future application.
If you are an administrator of a class A network, you will definitely have a headache for managing a large number of hosts. In order to facilitate management, you need to split it into many small subnets according to the actual situation. How to split it? This requires the use of subnet masks.
The subnet mask is a 32-bit address used to distinguish the network number from the host number, so that TCP / IP can determine whether an IP address is a local network or a remote network.
Each host on the TCP / IP network needs a subnet mask. If the network has not been subnetted, the default subnet mask should be used. When the network is divided into subnets, a custom subnet should be used shield.
When TCP / IP is initialized, the IP of the host is ANDed with the subnet mask to get a number M. When data needs to be sent, the TCP / IP protocol uses the subnet mask to “AND†with the destination IP to obtain a number D. When M and D are equal, the TCP / IP protocol considers the data packet to belong to the local network, otherwise, if not equal, the data packet is sent to the IP router.
For example: the IP of a host is 192.0.2.1, the subnet mask is: 255.255.255.0, then M = 192.0.2.0, if it sends a data packet to 192.0.2.114, then D = 192.0.2.0, M = D, TCP / IP knows that 192.0.2.114 is on the local network. If sending data to 193.0.2.1, D = 193.0.2.0, M and D are not equal, then the data packet is sent to the router.
Default subnet mask: all bits of the corresponding network number are set to 1, and the host number is set to 0. Such as:
* Class A network default subnet mask: 255.0.0.0
* Class B network default subnet mask: 255.255.0.0
* Class C network default subnet mask: 255.255.255.0
Custom subnet mask: divide a network into several subnets, each segment needs to use a different network number or subnet number, in fact, we can think of the host number is divided into two parts: subnet number, subnet Host number.
By dividing subnets, you can mix multiple technologies to overcome the current technical limitations. The most important thing is to reduce broadcast transmission and reduce network congestion.
How to define the subnet mask?
Before starting to divide, analyze your current needs and future demand plans, it is important to consider from the following aspects:
1. The number of physical segments in the network 2. Number of hosts per physical segment Step 1: Determine the number of physical network segments and convert them to binary numbers.
Step 2: Calculate the binary digits of the physical network. For example: you need 6 subnets, the binary value of 6 is 110, a total of 3 bits.
Step 3: Convert the required digits to decimal in high order. If you need 6 subnets, the binary value of 6 is 110, a total of 3 digits, so the first three digits of the host number will be used as the subnet number. The value of 11100000 is 224, the subnet mask is 255.224.0.0 for Class A networks, 255.255.224.0 for Class B networks, and 255.255.255.224 for Class C networks.
The main goal of this plan is to meet the needs of most campus network applications in a school campus LAN. On the basis of existing independent information classrooms, the school office building, teaching building and auxiliary auxiliary management department The main buildings form a campus network. The main contents included are:
1) Formulate a reasonable campus network overall construction plan and implementation plan based on the school's needs for information point arrangements and network applications;
2) Implement structured local area network cabling for several major buildings such as school office buildings, teaching buildings and auxiliary auxiliary management departments;
3) Complete the laying of two six-core outdoor backbone optical cables from the network center of the office building to the teaching building and the bungalow where the ladder classroom is located.
4) Realize the connection, installation, configuration and debugging of the school's core switches and secondary switches;
5) Implement connection and network debugging of newly purchased servers and some microcomputer network workstations.
2. Structured wiring design for a school campus LAN construction project
A structured wiring system is a transmission network within a building or group of buildings. It can not only connect data communication equipment, switching equipment, and other information management systems to each other, but also connect these devices to external communication networks, including buildings to the outside. The connection point on the network line, all cables between the data terminals in the work area, and the associated wiring components. A good structured wiring system has a certain degree of independence for the equipment it serves, and can interconnect many different communication devices such as data terminals, PCs and hosts, and public system devices. The general wiring system consists of six subsystems: inter-building subsystem, equipment subsystem, management area subsystem, vertical (backbone) subsystem, horizontal subsystem, and work area subsystem.
The campus network is a campus network, and the subsystems between the buildings are connected by optical cables, which can provide gigabit bandwidth and have sufficient room for expansion. The vertical subsystem is located in the shaft of high-rise buildings, and multi-mode optical cables or large-pair twisted-pair cables can be used. Integrate the management area subsystem into the equipment room subsystem for centralized management.
For multiple buildings, multiple equipment rooms can be used. It is divided into the central equipment room and the building equipment room. The central equipment room is the control center of the entire local area network. There are various network devices (switches, routers, video servers, etc.) for external (Internet) internal communication. The central switch passes The optical cable is connected to the switching equipment between the building equipment to ensure high-speed data transmission. Place wire racks and network equipment between this equipment, terminate the backbone cables from each floor in the building, and terminate the optical fiber connected to the network center.
The wiring in the building includes horizontal wiring and backbone wiring. The horizontal system adopts super five kinds of twisted pairs, the new building adopts the method of concealed installation in the wall, and the old building adopts the method of open installation of PVC wire duct.
This solution is proposed by the network company according to the basic requirements of a school (hereinafter referred to as the school) on the computer network and in accordance with the international commercial building wiring system standard (EIA / TIA 568B, ISO11801) to meet the application of the campus network. The goals are:
Realize 60 information points in the school office building, 136 information points in the teaching building, and 26 information points in the bungalow (a total of 222 information points, structured cabling in a LAN that exceeds Category 5 wiring standards.
According to the design requirements of EIA / TIA 568B (International Commercial Building Wiring System Standard), not only must you select wiring materials and equipment that meet international standards, but also design and construction in accordance with their standards. According to the requirements of the bidding party, we plan to select the products and technologies of the well-known wiring manufacturer that meets international standards-American AMP Company.
2.1 User demand for information point distribution and bandwidth
According to the requirements of the school, there are 222 LAN information points implemented this time. Among them, 60 office buildings are distributed on 3 floors, and 136 teaching buildings are distributed on 5 floors. The 26 information points of the bungalow are used for the ladder classrooms and auxiliary management departments of the school. The buildings are connected by multimode optical cables to form the campus LAN of the school.
The central computer room of the school LAN is located on the west side of the third floor of the office building. The wiring room of the teaching building is located on the north side of the east side of the third floor.
According to school arrangements, the specific distribution of information points is:
Office building Teaching building Bungalow Total
One floor Two floors Three floors One floor Two floors Three floors Four floors Five floors
22 20 18 34 30 30 28 14 26 222
60 136 26 222
In order to meet the school's future needs for various network applications, in addition, with the advancement of technology and technology, considering that the current various wiring materials are relatively close in price, it is proposed to route all information points according to the UTP UTP standard. Each information point can achieve a bandwidth of not less than 100Mb, which not only supports the network transmission bandwidth requirements of general school information management applications, but also fully supports the transmission of multimedia information in MPEG-2 and other formats.
This wiring scheme only considers data information point wiring, and does not involve voice information points and their equipment.
2.2 Integrated wiring design in the building
The characteristics of the school office building and teaching building are: without the elaborate decoration of the wall panels and ceilings. Therefore, it is proposed to lay open trunks in corridors and rooms. For aesthetic reasons, the trunk trunks of the horizontal subsystems can be concealed as much as possible. Only use the open trunks when it is unavoidable.
The physical size of the main part of the office building is about 35 meters long and 14 meters wide, with a total of 3 floors. The central computer room is located on the west side of the third floor, and the furthest information point to the east end of the first floor does not exceed 60 meters.
The teaching building is about 70 meters long and 14 meters wide, and has 5 floors. The secondary switch cabinet is located at the east end of the third floor. The information point farthest from the cabinet is the south room on the west side of the first floor. The longest distance is about 91 meters. The building is a building that has been used for many years. There are both wired closed-circuit feeders and telephone lines in the corridor of the open channel and structured wiring of the old local area network. The newly added trunking must avoid various situations that may cause interference. It is also necessary to maintain the appearance of the wire channel. Therefore, the wiring design and process need to be carefully handled.
According to the requirements of EIA / TIA 568 and ISO11801, the length of the twisted pair between the active device and the active device should be kept within 100 meters. Of course, when using Category 5 twisted pair, with 10Mbps bandwidth, individual information points can reach about 130 meters under the last circumstance (some network product manufacturers promise to reach 200 meters). In this solution, the distance from any desktop information point to the wiring room will not exceed 94 meters (the actual maximum distance does not exceed 91 meters), which fully meets the requirements of the technical specifications.
The main structured wiring materials (distribution frames, UTP cables, information sockets, plugs and other key components) of this project are selected from American AMP products. The product is quite competitive in the market due to the following characteristics:
AMP is the world's oldest connector production plant. Its product life, reliability and electrical performance have been tested for decades. Most computer manufacturers and communication equipment manufacturers use AMP's connectors as the cable connections for their equipment.
AMP's Netconnect product line includes all products that cover structured wiring systems and fully complies with international standards.
The AMP company's panel is square in shape and milky white in color, which is more harmonious with the domestic commonly used power sockets, switches and telephone sockets.
The socket on the panel is provided with a sliding dust cover. After the plug is pulled out, the cover automatically bounces off, which improves the cleanliness of the information socket.
AMP's Netconnect products have a good performance-price ratio. Similar products, AMP brands are generally more than 10â„… lower in price in the domestic market than AVAYA products. LAN planning topology diagram is as follows:
2.3 Links between buildings
2.3.1 Comparison of several commonly used transmission media
The connecting cables between the buildings are exposed outdoors, and are affected by the sun and rain forests and the interference of thunder and lightning all the year round. From the perspective of several commonly used wired transmission media such as twisted pairs, thin coaxial cables and thick coaxial cables and optical cables, From the aspects of anti-interference, anti-corrosion, high bandwidth and allowable transmission distance, outdoor armored optical cable is the most suitable transmission medium. Here are some analysis and comparison of these commonly used transmission media:
(1) Twisted pair
At present, the commonly used twisted pair in indoor wiring refers to unshielded twisted pair copper wire (UTP), and its transmission rate is closely related to the distance. Although within 100, five types of unshielded twisted pair can transmit 125MHz (100Mbps), but it The main cables between buildings are not suitable for cooperation. The reasons are:
Copper wire without a shielding layer can easily induce lightning and cause interference outdoors, or even damage the equipment.
UTP is prone to aging and has a short life due to the bad outdoor environment.
According to the PDS wiring rules, UTP cables are not allowed to exceed 100 meters (especially 100Mb Fast Ethernet), so it is not suitable for outdoor backbone cables.
As for shielded twisted-pair copper wire (STP), in theory, anti-interference and transmission distance are better than UTP, but it requires a higher grounding performance, and it is generally difficult to guarantee from the process, even if it is done, its investment Very impressive, otherwise it will be self-defeating. This is also the main reason why STP has not been widely popular.
(2) Coaxial cable
Traditional coaxial cables mainly include thick coaxial cables and thin coaxial cables. The thick coaxial cable has higher tensile strength than the thin coaxial cable, good mechanical properties and longer effective transmission distance (the maximum distance of the former is 500 meters, the latter is 185 meters), but the construction is difficult, and Due to fewer and fewer applications, it is difficult to find sources of goods on the market, and its cost has exceeded that of optical cables. The transmission rate of both the initial and thin cables is 10Mbps, and can only be used as a shared medium. However, due to the convenience of thin cable construction and the low price, if the number of workstations on the shared network segment is not large (within a dozen sites), indoor use can be considered. But reliability and maintainability are slightly worse. With the development of network technology, these two transmission media have been gradually eliminated.
(3) Optical cable
The optical cable has high bandwidth (generally up to hundreds of MBit or even dozens of Gbit level), long transmission distance, strong anti-interference ability and good security. The prices of its related products have also dropped significantly year by year. At present, it is the ideal transmission medium for the industry as the backbone of the network.
For a school campus network, it involves the inter-building LAN interconnection of 3 buildings. Because the network center is located in the office building, according to the star-shaped network topology, the characteristics of the current centralized application mode of client / server (C / S) or browser / server (B / S) are considered, so the other two The buildings should be connected to the network center of the office building.
As far as most campus network applications are concerned, among the transmission information carried on the campus network, the transmission volume of multimedia information will be larger and larger, such as multimedia teaching, electronic reading, video on demand and other applications. Therefore, no matter from the current or future development point of view, the backbone transmission medium must have the ability to carry Gigabit rates. In addition, as an outdoor transmission medium, it also needs to have good anti-interference, reliability, anti-aging and high life characteristics. However, the transmission media with these characteristics can only meet the requirements with optical cables.
According to the mode, the optical cable can be divided into multi-mode and single-mode. They can carry Gigabit transmission rates, but the distance between them is quite different. The former has a maximum transmission distance of only 275 meters (62.5 / 125μm) to 550 meters (50 / 125μm) due to its optical transceiver Semiconductor devices (LED), so the price is relatively cheap; and because single-mode fiber uses a laser device as an optical transceiver, the transmission distance can reach tens of kilometers, and the price is several times higher than the former. For this solution, since the distance of the farthest optical cable between buildings will not exceed 200 meters, it is advisable to use a 62.5 / 125μm multimode optical cable to save investment.
2.3.2 Outdoor backbone link design
Topology solution
According to the needs of the school, it is necessary to interconnect the indoor information points of three buildings such as office buildings, teaching buildings and bungalows into the campus LAN of the school. From the convenience of management and maintenance, it is conducive to increasing the speed of all information points to access the central server, and considering the geographical location as a whole, a better topology solution is: take the network room of the office building as the center, and radiate to other 2 in a star structure Building.
Medium selection for interconnection between buildings
As mentioned in the previous analysis, for the outdoor backbone transmission medium of the current campus network, the optical cable has become the only effective choice. It not only has a good bandwidth and its scalability, but also is very superior in many aspects such as anti-interference, reliability, stability and service life. The disadvantage is that the construction is slightly more difficult and the cost is higher. For the former, as long as the design is well-considered and the construction is in place at once, sufficient forward-looking and sufficient scalability are considered, this problem can be solved well. To this end, it is planned to use multi-mode 6-core optical cables for the outdoor backbone links from the office building network center to the other two building buildings (the teaching building has more information points, you can consider 2 gigabit uplink links, or press LAG The way is bundled into a 2 Gigabit backbone link. Once any link fails, the other link will automatically run as a Gigabit rate link. There are also two cores for backup). This enables high-quality, redundant, reliable, and stable high-quality backbone links.
There are currently at least three methods for the connection of optical cables: welding, grinding and cold compression. Although the grinding method has a certain application market in multi-mode optical cables, experience has shown that the consistency is poor, often different from person to person, and the attenuation is generally large. Splicing is easy to ensure performance, especially in single-mode optical cables. Cold compression bonding is a new process that requires special equipment, and its performance is not as good as welding. We generally use fusion cable connection method.
3. Bandwidth analysis
3.1 Consideration of backbone bandwidth
The backbone network is actually the connection channel between the central switch and the secondary switch (for a network environment that is not too large, that is, the access level) and between the central switch and the server. For switched Ethernet, there are three types of bandwidth: 10M, 100M and 1000M. For this solution, according to the network scale, application characteristics and the will of the school, the backbone bandwidth is planned to be considered as 1000M. Between the central switch and the server, you can choose between 100M or 1000M according to the size and application type of the server, and the 1000M full-duplex mode should be adopted between the central switch and the access switches in office buildings, teaching buildings, and bungalows.
For a local area network that is not very large and the area is not too scattered, it is appropriate to reduce the number of cascades of switches or hubs as much as possible to avoid increasing delay. For this program, there are only two levels.
3.2 Client bandwidth analysis
The client's bandwidth is generally shared 10Mb, exchanged 10Mb and exchanged 10 / 100Mb. Although the former has low cost, the network performance is poor and it is difficult to meet the application requirements. Most of the network cards currently available on the market are 10 / 100Mb adaptive. The client applications for teaching buildings and office buildings may be mostly school management, but there is no shortage of multimedia information transmission such as low-speed surgery and CT images. Moreover, the price of the 10Mb network card and the 100Mb network card on the market is almost the same. Therefore, it is planned to configure all the clients of these two buildings to be adaptively switched to the desktop by 10 / 100Mb (provided that all inter-building connections are Module cable).
Back to Contents
4. 10 / 100Mb switch product selection
For the selection of the central switch, it is important to consider its switching capacity, scalability, and has good second-layer performance and third-layer functions. However, in selecting the grade and scale of the core switch, we must combine the school's network scale and application characteristics, and leave room for expansion appropriately. However, buying network products now is not tailor-made after all, but chooses the grade and model as close to the required as possible from the existing products of various manufacturers in the market according to their own needs. If the selected configuration is too high, it will cause a waste of resources, and the configuration is too low, not only does not have room for expansion, but also affects network performance. However, when designing the scale and performance of various grades of products, different network manufacturers will produce large differences according to their own technology and technological level. In addition, for products with similar performance and configuration, the prices of different countries, brands, and manufacturers vary greatly. Generally speaking, the products of well-known foreign brands are much higher than those of domestic ones. Even domestic network products have several grades. Therefore, there are trade-offs to consider when choosing a network product.
At present, the main brands of foreign LAN products that are more popular in China are: CISCO, AVAYA, 3COM and so on. These products are characterized by mature technology, stable performance, reliability, and rich functions. But the price is relatively high. Such products are widely used in financial, securities and large enterprises with high requirements. It is worth noting that in recent years, many emerging network product manufacturers have rapidly emerged in China, such as Digital China, Huawei, Gangwan, TCL, and Shida. The products of these manufacturers are comparable in function and performance to similar foreign products, but they are more flexible in terms of configuration and port combination, especially in terms of price. The core switch with high port density reflecting the advanced technology not only cannot be used in this solution, but also will cause a greater degree of port waste. Therefore, in view of the needs and environment of the second normal school, we believe that the domestic network products of China Digital are more suitable. The modular core switch port density is not very high, but the switching capacity is higher and the performance-price ratio is higher.
In this scheme, the core switch selects DCRS-6512. The switch is suitable for small and medium-sized campus networks and campus networks, and provides excellent performance and reliability based on leading technologies. The DCRS-6512 switch is designed to take advantage of the potentially huge switching capabilities of Gigabit Ethernet, and its non-blocking structure can ensure that each port easily has full line-speed switching capability, ensuring line speed under huge traffic and network load The second layer and the third layer are exchanged. Can be used as an ideal network core switch.
The DCRS-6512 chassis itself can provide 12 user slots for I / O switching, and can provide up to 24 Gigabit ports and 96 100M ports. The switching capacity of the whole machine is 48Gbps, which can achieve 36Mpps line-speed full-layer packet forwarding rate. The configuration and combination of its modules are very flexible. In this scheme, an 8-port Fast Ethernet copper cable module is selected for connecting to a 100M rate PC server; at the same time, two 2-port Gigabit Ethernet copper cable modules are also provided, providing 4 gigabits. The copper cable ports can be connected to the two access switches of the office building in a dual-link aggregation mode, which can both expand the bandwidth and realize automatic link backup; if necessary, it can also be used in the future to connect a specific server Gigabit network card. In addition, two 2-port Gigabit Ethernet multimode fiber optic cable modules were selected, which are mainly used to connect the secondary switches of the teaching building and the bungalow with multimode fiber. Its functions and performance fully meet the various network applications of the school.
The secondary switch selects DCS3426 and DCS3628S. They are all Gigabit network-connectable secondary switches with 10 / 100Mb rate access. The former is used to connect office buildings and bungalows to core switches independently. The latter is a stackable switch, mainly used in teaching buildings. Due to the dense port of the teaching building, in order to save optical fiber links and Gigabit uplink ports, the switches are now stacked together with a 4Gb bandwidth (maximum 6 stackable), and then two Gigabit links are aggregated Connect to the core switch. The following table lists the main functions and performance of core switches and secondary switches:
Switch type DCRS-6512 DCS-3426 DCS-3628S
Backplane speed 48 Gbps 18 Gbps 18 Gbps
Forwarding bandwidth 36Mpps 148800 0pps / port 1488000 pps / port
Number of MAC addresses supported 32k 12k 12k
VLAN support 802.1Q 802.1Q 802.1Q
Number of VLANs / Dynamic VLAN 256/4096 256/4096 256/4096
STP 802.1d 802.1d 802.1d
QoS 802.1p, 4 queues 802.1p, 2 queues 802.1p, 2 queues
Stacking support No No Yes / 4Gb
FEC (LAG) support Yes, 8 ports / bar, 12 groups Yes, 8 ports / bar Yes, 8 ports / bar
Maximum 10 / 100Base-TX 96 pcs 24 pcs 24 pcs
Maximum 1000Base-SX 24 pcs 2 pcs 2 pcs
Maximum 1000Base-T 24 pcs 2 pcs 2 pcs
Expansion slot supports 12 2 3
SNMP MIB II support Yes Yes Yes
5. Completion and acceptance
1. The information socket and power socket are installed in place, neat and beautiful.
2. The main wiring system is installed in place, neat and beautiful.
3. Layout of wire channel and cable is neat and beautiful.
4. The system connection is neat and beautiful.
5. The quantity and model of equipment materials are consistent with the design.
6. Provide 100% information point test and performance test report.
7. The installation of the entire network system reaches an arbitrary point of information extraction, which can achieve network interconnection (such as WIN98 / WIN2000 interconnection)
Network equipment selection
The prices of different network equipment vary greatly. Let us first briefly understand the network equipment.
HUB: This is the so-called hub. It can be divided into several types, including common HUB, stacking HUB, port switching HUB and so on.
The bandwidth of 100M HUB is shared, that is to say, the 24 ports of the 24-port HUB share 100M bandwidth. If 24 ports transmit data at the same time, then the bandwidth of each port is only about 10M. The stacking HUB is a stackable HUB, that is to say, if we need to network 48 machines, we can use 2 stacking HUBs to stack up as a 48-port HUB.
Switch: It can be considered as a high-performance HUB. Its 100M bandwidth is independent, or it allows several ports to transfer data at a speed of 100M at the same time. The switch usually also has a routing function.
The network center is the core of the company's network. In order to avoid possible collisions on the network, our core equipment is a 350T switch from Bay. It adapts to the 10 / 100M network with routing functions and has good overall performance. All HUBs are directly connected to the switch, and important servers are also directly connected to the switch, so that the advantages of fast switch speed and high bandwidth can be fully utilized.
The HUB we chose is Intel ’s Express 100TX-BASE stack HUB (Intel is not willing to only do CPU, it also makes network cards, graphics cards, maybe it will be a chassis someday. :-), it has 24 ports , Can be stacked. Made using Bay's technology, the price is good.
Note: There is a small button on the left of the first port of the HUB. Press it to make 1, 2, and 3, 6 of the first port cross. This port is a port dedicated to the connection between the HUB and the HUB or switch. In fact, this is how our company's HUB is connected to the switch.
Network card: Every computer needs a network card. We initially selected 3Com's 3C905, 10/100 adaptive network card. We also bought a few Intel's 82557. After a period of use, one of the Intel's network cards broke. In comparison, 3Com's network card quality is better than Intel's network card. Later, we used the DFE-500TX network card from D-Link, and the price / performance ratio was quite good.
Network wiring system: Choose five types of wiring system of AMP company. When making the network cable, pay attention to that it is not simply to connect the 8 wires of RJ-45 one by one. You must ensure 1, 2 twisted pairs, 3, 6 twisted pairs, 4, 5 twisted pairs, 7 and 8 pairs If the twisting is only for one-to-one connection instead of guaranteeing 1, 2 and 3, 6 twisting, it may cause the site with a long network cable to become unstable or even unable to work normally.
Network configuration, construction server settings: There are a total of 2 servers on the LAN, one of which is used as an internal file server. The other is used as an Internet server. The Internet server runs Windows NT + IIS + Exchange Server and provides WWW, FTP, and Email services.
Construction: When calculating the length of the network cable, it is necessary to reserve a 10% margin to avoid the necessary bypass and other unpredictable situations in case of the structural reasons of the building.
An integrated wiring system is not so much a computer project as a construction project. The actual performance has a lot to do with the installation process. During construction, pay attention to that the network cable cannot withstand excessive curvature and avoid close to strong interference sources and building subsystems ( That is, the network cable connecting the two buildings must be strengthened. We use steel pipes for this part of the network cable. The advantages of this are: high strength and strong anti-interference ability.
IP address allocation: According to the relevant provisions of RFC1597, in order to facilitate future connection to the Internet and to consider the development of the campus network, it was decided to use a Class B network inside the campus, with a network number of 172.16 and a corresponding subnet mask of 255.255.0.0.
Computer name naming rules: department code + serial number, the mantissa of the IP address is consistent with the mantissa of the computer name. For example, 172.16.1.1 ==> Technical Department rd1.
Understand the IP address and subnet mask. I ca n’t help thinking about the subnet mask here:
We know that an IP address is a dotted decimal number, and each IP address is composed of two parts: a network number and a host number. The network number marks a physical network. All hosts on the same network need the same network number, and the network number is uniquely determined on the Internet. The host number determines a TCP / IP host such as a workstation, server, router, etc. in the network. For the same network, the host number is unique. Through the network number + host number, we can determine the location of a host on the Internet.
Since the network number + host number can determine a host, what is the use of subnet mask?
In order to adapt to networks of different sizes, the Internet defines 5 types of IP addresses:
Class A addresses: the highest digit is 0, the next 7 digits represent the network number, and the remaining 24 digits represent the host number. A total of 126 networks are allowed, with approximately 17 million hosts per network.
Class B address: the highest 2 digits are 10, the next 14 digits are the network number, and the remaining 16 digits are the host number. It allows 16384 networks, and each network has about 65000 hosts.
Class C address: the highest 3 digits are 110, the next 21 digits are the network number, and the remaining 8 digits are the host number. It allows 2 million networks, and each network has about 254 hosts.
Class D address: the upper 4 bits are 1110, used for multicast.
Class E address: High 4 is 1111, only for experiment, reserved for future application.
If you are an administrator of a class A network, you will definitely have a headache for managing a large number of hosts. In order to facilitate management, you need to split it into many small subnets according to the actual situation. How to split it? This requires the use of subnet masks.
The subnet mask is a 32-bit address used to distinguish the network number from the host number, so that TCP / IP can determine whether an IP address is a local network or a remote network.
Each host on the TCP / IP network needs a subnet mask. If the network has not been subnetted, the default subnet mask should be used. When the network is divided into subnets, a custom subnet should be used shield.
When TCP / IP is initialized, the IP of the host is ANDed with the subnet mask to get a number M. When data needs to be sent, the TCP / IP protocol uses the subnet mask to “AND†with the destination IP to obtain a number D. When M and D are equal, the TCP / IP protocol considers the data packet to belong to the local network, otherwise, if not equal, the data packet is sent to the IP router.
For example: the IP of a host is 192.0.2.1, the subnet mask is: 255.255.255.0, then M = 192.0.2.0, if it sends a data packet to 192.0.2.114, then D = 192.0.2.0, M = D, TCP / IP knows that 192.0.2.114 is on the local network. If sending data to 193.0.2.1, D = 193.0.2.0, M and D are not equal, then the data packet is sent to the router.
Default subnet mask: all bits of the corresponding network number are set to 1, and the host number is set to 0. Such as:
* Class A network default subnet mask: 255.0.0.0
* Class B network default subnet mask: 255.255.0.0
* Class C network default subnet mask: 255.255.255.0
Custom subnet mask: divide a network into several subnets, each segment needs to use a different network number or subnet number, in fact, we can think of the host number is divided into two parts: subnet number, subnet Host number.
By dividing subnets, you can mix multiple technologies to overcome the current technical limitations. The most important thing is to reduce broadcast transmission and reduce network congestion.
How to define the subnet mask?
Before starting to divide, analyze your current needs and future demand plans, it is important to consider from the following aspects:
1. The number of physical segments in the network 2. Number of hosts per physical segment Step 1: Determine the number of physical network segments and convert them to binary numbers.
Step 2: Calculate the binary digits of the physical network. For example: you need 6 subnets, the binary value of 6 is 110, a total of 3 bits.
Step 3: Convert the required digits to decimal in high order. If you need 6 subnets, the binary value of 6 is 110, a total of 3 digits, so the first three digits of the host number will be used as the subnet number. The value of 11100000 is 224, the subnet mask is 255.224.0.0 for Class A networks, 255.255.224.0 for Class B networks, and 255.255.255.224 for Class C networks.
Acetate-Metal Glasses,Acetate Frames,Acetate Material Glasses,Square Acetate Sunglasses
Danyang Hengshi Optical Glasses Co., Ltd. , https://www.hengshi-optical.com