A hub (HUB) or switch (Switch) is the most commonly used device in a local area network. Generally, they provide convenience for users to find network faults. For example, by observing whether the indicator light of the port connected to the HUB (or Switch) is lit, you can determine whether the network connection is normal. For 10/100Mb/s adaptive HUB (or Switch), you can also judge whether the connected computer is working in 10Mb/s state or 100Mb/s state by the different colors of the connection port indicators. Therefore, in most applications, the use of HUB (or Switch) is beneficial to network maintenance. However, improper use or damage of the HUB (or Switch) will cause problems to the network connection.

[Fault phenomenon]

After upgrading the network from 10Mb/s to 100Mb/s, the network cannot work normally.

[Fault Analysis and Processing]

In a LAN, when the network connection range is large, the network transmission distance can be expanded through cascading between HUBs. In a 10Mb/s network, up to four levels can be cascaded, making the maximum transmission distance of the network reach 600m. But when the network is upgraded from 10Mb/s to 100Mb/s or a new 100Mb/s LAN is built, if the 100MHUB is connected using ordinary methods, the LAN will not work properly. As we all know, only two 100Mb/s HUBs are allowed to be cascaded in a 100Mb/s network, and the connection distance between two 100Mb/s HUBs cannot be greater than 5m, so the maximum distance when using a HUB in a 100Mb/s LAN is 205m . If the actual connection distance does not meet the above requirements, the network will not be able to connect. This point must be paid enough attention, otherwise it will easily cause serious errors when users plan the network.

[Fault Phenomenon]

A certain unit set up a local area network by itself, using two 16-port 10M*** shared hubs (with a cascade port), and all computers Connected to the HUB in the switchboard room through HUB. The HUBA is connected to the 16th port of the HUBB through the cascade port, the HUBB is connected to the HUB in the main computer room through the cascade port, and the other ports are connected to the workstations respectively. The entire workstation uses static IP addresses, whose values ??are 192.168.0.2, 192.168.0.3...and so on. 192.168.0.1 is assigned to the NT server. Each computer (including the server ) subnet masks are all 255.255.255.0. Before officially connecting to the server, each workstation was set up and tested using the Ping command. The results were all connected, and all workstations connected to HUB A could also be connected to the workstation connected to HUB B using the Ping command. However, when connected to the server, only the workstation connected to HUB B can log in to the server, but the workstation connected to HUBA cannot log in.

[Fault Analysis and Treatment]

By observing the indicator lights of the computer network card and the indicator lights of each port on the two HUBs, it was found that the 16th port of HUBB was not the same as HUB A. In addition to the indicator lights corresponding to the cascade port not lighting up, the indicator lights of all network cards and other ports are evenly lit, indicating that the connection between the computer and the HUB is normal, so the problem is most likely to be between the cascade port of the HUBA and the On the 16th port of HUB B. According to this situation, it is suspected that at least one of the cascade port of HUB A and the 16th port of HUB B is broken. In order to further confirm that the port is bad, the positions of the two HUBs can be exchanged, but the result remains the same. Next, I tried to plug the twisted pair connected to the cascade port of HUB A into an ordinary port elsewhere on HUB B. As a result, the problem was solved. All workstations in the network can connect to the server, and the two HUBs are connected to workstations can all respond to each other. It can be seen from this that the cascade port of some HUBs and the port immediately adjacent to it are not two independent ports, but should belong to the same port (although there are two independent physical ports). Many previous HUBs used a toggle switch to select the cascade port between two ports, but this switch was omitted in subsequent products. However, if one of the ports is used as a cascade port, the other port will One port will be invalid.

[Fault Phenomenon]

A HUB connecting two buildings often burns out, sometimes three or four times in a month.

[Fault Analysis and Treatment]

After testing, the power supply system in Building A has aged. The absolute voltage of the neutral line is 30V and the absolute voltage of the live line is 250V. However, the voltage measured with a multimeter is still 220V; from UB to Building B HUB, the two HUBs have to withstand a potential difference of 30V, which may cause damage. The solution is very simple, just connect a ground wire in the switch room of Building A.

We know that there is only one physical signal transmission channel in a ring network, and all signals are transmitted through a transmission medium. In this way, there is a contradiction between nodes competing for the channel, and the transmission efficiency is low.

After the hub is introduced as a network concentrating device, each station uses its own dedicated transmission medium to connect to the hub. There is no longer only one transmission channel between each node. The signals sent back by each node are concentrated through the hub, and the hub then transmits the signal After shaping and amplification, it is sent to all nodes, so that at least there will be no collision on the upstream channel. However, a hub-based network is still a LAN with a shared medium. The shared medium here is actually the internal bus of the hub, so when the upstream channel and the downstream channel send data at the same time, there will still be signal collisions. When the hub detects a collision from its internal port, it generates a collision enhancement signal (Jam) and transmits it to the target port to which the hub is connected. At this time, all data will not be sent successfully, causing a network traffic jam.

We can use a vivid realistic situation to illustrate the occurrence of this network phenomenon, that is, there are cars approaching from two directions on a single lane at the same time.

We know that vehicles in one direction are usually allowed to pass on a single lane. However, in small towns, where conditions are limited, there is usually no such regulation. It is possible that vehicles in both directions are allowed to pass on a single lane. , but must pass at different times. The same is true in a hub. Although each node has its own independent channel for connecting to the hub, there is only one and the same channel inside the hub. Both uplink and downlink data must be sent and received through this shared channel. Data, this may be like a single lane. When data is sent on the upstream and downstream channels at the same time, traffic jams may occur. That’s easy to understand, right?

Because of this shortcoming of the hub, it cannot be used alone in larger networks (usually with switches and other equipment to share a small part of the network communication load), such as in large city centers It cannot be the same as a single lane, because the larger the network, the greater the chance of network collisions. Because of this, the data transmission efficiency of the hub is relatively low, because it can only transmit data in one direction at the same time, which is the so-called simplex mode. If a hub is to be used as a single hub device in the network, the network size should be within 10 units, and the hub bandwidth should be above 10/100Mbps.

In addition to the disadvantage of hubs sharing bandwidth, there is another aspect that must be considered when choosing a hub, and that is its broadcast method. Because the hub is a pure hardware network bottom device, it basically does not have intelligent memory capabilities, let alone learning capabilities. It also does not have the MAC address table that a switch has, as shown in Figure 2.

This method of broadcasting data has two shortcomings: coupled with the shared bandwidth method introduced above, it is more likely to cause network traffic jams and further reduce network execution efficiency.

After understanding the working principle of the hub, let's learn about the installation and connection method of the hub in the LAN.

The most important thing about an access device is its interface technology. Different interfaces are used in different application environments, and different applications correspond to corresponding interfaces. This is not only true for hubs, but also for switches that will be discussed later. , routers, etc. are the same. The interface of the hub is relatively the simplest. In order for everyone to master the various application connections of the hub, it is necessary for us to understand some of the main interfaces of the hub.

Hubs usually provide three types of ports, namely RJ-45 ports, BNC ports and AUI ports, to suit networks built by connecting different types of cables. Some high-end hubs also offer fiber optic ports and other types of ports.

(1) RJ-45 interface

The RJ-45 interface can be used to connect an RJ-45 connector and is suitable for networks built with twisted pairs. This type of port is the most common , Generally speaking, Ethernet hubs will provide this kind of port. The multi-port hub we usually talk about refers to the number of RJ-45 ports it has. As shown in Figure 3.

The RJ-45 port of the hub can be directly connected to terminal devices such as computers and network printers, and can also be connected to other switches, hubs and other hub devices and routers. It should be noted that the twisted pair cables used have different patching methods when connecting to different devices. For details, please refer to the introduction to the network cable production chapter introduced earlier.

(2) BNC port

The BNC port is an interface used to connect to a thin coaxial cable. It is usually connected through a BNC T-type connector, as shown in Figure 4 It is a schematic diagram of a hub BNC port connected through a BNC-type connector.

Most 10Mbit/s hubs have a BNC port. When the hub has both BNC and RJ-45 ports, it can connect to the twisted pair network through the RJ-45 port and the thin cable network through the BNC interface. Therefore, twisted pair and thin coaxial cables can be realized A connection between two networks using different communication transmission media.

This dual interface feature can be used to be compatible with the original thin coaxial cable network (10Base-2), and can realize the gradual transition to the mainstream twisted pair network (10Base-T). Of course, it can also realize remote coaxial cable communication. The connection between the axis cable network (less than 185 meters), Figure 5 below is a schematic diagram of a network connection between two different transmission media networks through the BNC interface and twisted pair RJ-45 interface of the hub. .

Similarly, if the distance between the two networks is greater than 100 meters and the connection between the two networks cannot be achieved using twisted pair, you can also use a thin coaxial cable through the BNC port of the hub. Transmission connects two transmission networks, and both networks can still use twisted pair, a cheap and common transmission medium, as shown in Figure 6. However, it should be noted that the distance between the two networks still cannot be greater than 185m.

(3) AUI port

The AUI port can be used to connect the AUI connector of the thick coaxial cable, so this interface is used to connect to the thick coaxial cable network, its schematic diagram As shown in Figure 7, there are relatively few hubs with this interface, mainly found in some backbone-level hubs.

Due to the high cost of networks using thick coaxial cables as transmission media and the difficulty of wiring, it is very rare to actually use thick coaxial cables for wiring in practice. However, since a single section of thick coaxial cable (10Base-5) supports a transmission distance of up to 500 meters, thick coaxial cable can be used as a communication cable to connect longer-distance networks. Therefore, it can also be used as a cheap remote connection solution. The connection diagram is shown in Figure 6 above. The difference is that the connection medium between networks used here is thick coaxial cable.

With the help of a transceiver, the AUI port can also be connected to an RJ-45 interface, a BNC interface, or even an optical fiber interface. Figure 8 below shows from left to right: AUI to RJ-45 transceiver (used to realize the connection between the AUI port and the RJ-45 interface), AUI to BNC transceiver (used to realize the connection between the AUI port and the BNC interface) connection), AUI to ST transceiver (used to realize the connection between the AUI port and the optical fiber interface). Of course, there are many types of such transceivers, such as RJ-45 to RS-232, RJ-45 to BNC, etc. But don’t underestimate this little thing. Just guess the price. I think in most cases you will underestimate it. I couldn’t understand the price when I bought it for the first time, but I asked After visiting many stores (it was really hard to find one), I realized that I had to accept the fact of the price. Generally speaking, the genuine product of this kind costs about 130 yuan. Is it beyond your expectation? The main brands of this kind of interface transceiver are: D-Link, HP, etc. This product plays the role of an interface type conversion (of course it is not as simple as cable connection, it needs to be completed through a certain circuit), so it is usually called an adapter.

(4) Hub stacking port

Of course, this port is only available on stackable hubs. Its function is just like its name, it is used to connect two Stackable hubs. Generally speaking, a stackable hub has two similar-looking ports at the same time: one is marked UP and the other is marked DOWN. When connecting, a cable is used to connect from the UP port of one hub to the DOWN of another stackable hub. The ports are all female, so the cable ends must be male. However, this kind of cable is provided by the manufacturer when you purchase the stackable hub. If it is damaged or lost, you can also buy it directly from the computer store. To make one, just explain the purpose to the merchant. The port diagram is shown in Figure 9 below.

The installation of the hub is relatively simple, especially the fool hub, as long as it is fixed in the wiring cabinet and the power cord is plugged in. Which twisted pair needs to be connected, just plug the RJ-45 end of the twisted pair into the hub port. Although smart hubs can be used once they are fixed, if you want to achieve remote management, you must perform necessary configurations and specify IP address information for the hub. In addition, rack-mounted hubs are generally used in some large networks, which involves rack installation of the hub.

Structurally speaking, hubs can be divided into two types: rack-type and desktop-type. The hubs used by general departments are desktop-type; enterprise computer rooms usually use rack-type hubs. Rack-mounted hubs are easy to fix in a fixed place. They are usually installed in a cabinet with other hubs, switches, and sometimes servers. This facilitates network connection and management, and also saves equipment. space occupied. If you choose a rack-mounted hub when purchasing, you can choose a hub rack (usually provided by the manufacturer). Let's take a look at the installation of a rack-mounted hub.

Rack-mounted hubs are generally installed in cabinets together with other equipment. Of course, these cabinets have corresponding structural standards in the industry, especially in terms of size (such as width, 1U (unit) height, etc.), so that all devices can be installed together conveniently and beautifully. This is why the hub is empty inside, but must be made the same size. Of course, there are other benefits to a large chassis. That means better heat dissipation.

International standard cabinets can be roughly divided into three categories in terms of width: 19 inches, 23 inches and 24 inches, which are mainly based on the requirements of server cabinets. Depending on the number of installed devices, cabinets of different heights can also be selected. The height of the cabinet is usually measured in units of U. U actually means Unit, which in Chinese means one unit, 1U=1.75 inches. The installation of this kind of cabinet is usually carried out mainly according to the following steps.

Step 1: Fix the mounting bracket

Before installing the hub to the cabinet, you should first install the fixing bracket at the specified position of the hub (refer to the operation manual). This is Prepare for later installation of the hub in the rack. Different hubs have different brackets, but the installation principles are basically the same. Figure 10 shows a mounting bracket provided with the Cisco hub.

The size of Cisco's network equipment is mostly 19 inches (because 19 inches is the most popular international cabinet standard). When installing 19-inch network equipment into a 19-inch cabinet, how to fix the mounting bracket As shown in Figure 11. When the size of the cabinet is 23 or 24 inches, the network equipment needs to be installed in the 23 or 24-inch cabinet. The mounting bracket should be fixed as shown in Figure 12.

Step 2: Fix the device

After the mounting bracket is fixed, the next thing to do is to put the hub device with the bracket installed into the corresponding position of the cabinet and fix it in the cabinet. . In fact, this installation method is very easy. In fact, it only requires fixing a few screws. The installation method is shown in Figure 13.

Step 3: Fix the cable guide

After installing the hub into the cabinet, it is necessary to connect the network cable. Generally speaking, there are several network devices in a cabinet. , so there are many network cables concentrated in this cabinet. If these network cables are not sorted out clearly, it will bring great inconvenience to network management. For this reason, we need to bundle, install and organize the network cables. At this time, it is generally necessary to install wire guides for the network cables, so that the bundled network cables become neat, beautiful, and easy to manage. The installation method of the wire guide is shown in Figure 14.

Above we introduced the method of installing a rack-mounted hub in a cabinet. This is generally suitable for larger networks. For small offices, there are usually no cabinets, and the hub can only be installed on the desktop or wall.

To install the hub on the desktop, you can first fix the mounting bracket on the desktop. This installation method should be noted that there are two different installation directions: one is the horizontal installation method where the hub is placed horizontally, Figure 15 shows the horizontal fixing method; the other is to place the hub vertically, and its bracket fixing method is shown in Figure 16.

There are also two ways to install the hub on the wall: one is to fix the hub horizontally on the wall, and the installation method shown in Figure 17 can be used; the other is to fix the hub horizontally on the wall. When installing vertically on the wall, install it as shown in Figure 18.

Although the connection of the hub is simple and basically requires no configuration, by understanding its connection principle, the hub can be better utilized to meet the needs of medium and small network applications. Before formally introducing the connection method of the hub, let's first understand the signal forwarding principle of the hub.