Technological advances have led to a rapid decline in the price of switches, the average user also has the ability to buy such equipment, but with the early network equipment (mainly with the hub HUB comparison) compared to what advantages it has? This is also a concern for most users. Here is a comparison of two devices that achieve the same result, but work on different mechanisms: switch (Switch) and hub (HUB).

From the working phenomenon, they are connected to the Ethernet through multiple ports of the device, you can connect multiple users through the network in a star structure, **** enjoy resources or exchange of data. But the breakdown of their working state, but completely different.

The working mechanism of the hub is broadcasting (broadcast), no matter what type of packet received from which port, are in the form of broadcasting the packet to all the remaining ports, connected to these ports on the NIC (NIC) to determine the processing of this information, in line with the left to deal with, or else discarded, which is very easy to produce broadcast storms, when the network is larger Network performance will be greatly affected. From its working state, the HUB performs less efficiently (sending the packet to all ports) and has poor security (all NICs receive it, only the non-destination NICs discard the packet). Also, only one packet can be processed at a time, and collisions occur when packets appear on multiple ports at the same time; packets are processed serially, making them unsuitable for use in larger network backbones.

Switches work quite differently, and the low-end switches are now Layer 2 switches, which switch based on MAC addresses. It analyzes the header information of an Ethernet packet (which contains the original MAC address, destination MAC address, message length, etc.), obtains the destination MAC address, looks up the address cross-reference table stored in the switch (the port corresponding to the MAC address), confirms which port the NIC with the MAC address is connected to, and then sends the packet only to the corresponding port, effectively and efficiently suppressing the broadcasting This effectively suppresses the generation of broadcast storms. This is the biggest difference between Switch and HUB. The backplane bandwidth for forwarding packets within the Switch is much larger than the port bandwidth, so packets are in parallel, which is more efficient, and meets the requirements of large-scale network environments for parallel processing of large amounts of data.

Classification of LAN Switches

●From the point of view of transmission medium and transmission speed, LAN switches can be classified into Ethernet switches, Fast Ethernet switches, Gigabit Ethernet switches, FDDI switches, ATM switches, and token ring switches, etc., which are suitable for Ethernet, Fast Ethernet, FDDI, ATM, and token ring network environments. environments.

●According to the widest common categorization, LAN switches can be divided into three categories: Desktop Switch, Workgroup Switch and Campus Switch.

1. Desktop Switch is the most common type of switch, the most widely used, especially in the general office, small server room and business acceptance of a more centralized business departments, multimedia production centers, website management centers and other departments. In terms of transmission speed, most modern desktop switches provide multiple ports with 10/100Mbps adaptive capability.

2. Group switch is a workgroup switch, which is often used as an expansion device. When the desktop switch cannot meet the demand, most of the group switch is considered directly. Although the group switch has only a small number of ports, it supports more MAC addresses and has good expansion capability, and the transmission speed of the port is basically 100Mbps.

3. Campus switch, this switch is relatively less used, only applied to large networks, and generally used as the backbone of the network, and has fast data exchange capability and full-duplex capability, and provides intelligent features such as fault tolerance and other features. Provide fault tolerance and other intelligent features, but also supports expansion options and Layer 3 switching in the virtual local area network (VLAN) and many other features.

●Based on the architectural features, people also categorize LAN switches into three types of products: rack-mounted, fixed-configuration with expansion slots, and fixed-configuration without expansion slots.

1. Rack-mounted switch This is a slot-type switch, which is more scalable and can support different network types, such as Ethernet, Fast Ethernet, Gigabit Ethernet, ATM, Token Ring, and FDDI, etc., but it is more expensive, and a lot of high-end switches use rack-mounted structure.

2. Fixed Configuration Switch with Expansion Slot It is a switch with a fixed number of ports and a small number of expansion slots. This switch can support other types of networks by expanding other network type modules on top of supporting fixed port type networks. The price of this type of switch is in the middle.

3. Fixed Configuration Switch without Expansion Slots This type of switch supports only one type of network (usually Ethernet), and can be used in small business or office environments for local area networks (LANs), which are the least expensive and most widely used.

Common technical specifications for LAN switches

There are more basic technical specifications for LAN switches, which fully reflect the technical performance and functions of the switches, and are an important source of data for users to refer to when purchasing products. The more important ones are as follows.

1. The number of rack slots: the maximum number of modules that can be inserted in a rack-mounted switch.

2. Expansion slots: The maximum number of modules that can be inserted in a fixed configuration switch with expansion slots.

3. Maximum Stackability: The maximum number of switches that can be stacked in the stacking unit of a stackable switch. Obviously, this parameter also indicates the maximum port density and message point connectivity that can be provided in a stacking unit.

4. Supported Network Types: In general, fixed-configuration switches without expansion slots support only one type of network, while rackmount switches and fixed-configuration switches with expansion slots can support more than one type of network, such as Ethernet, Fast Ethernet, Gigabit Ethernet, ATM, Token Ring, and FDDI. The more network types a switch supports, the greater its availability and scalability will be.

5. Maximum SONET ports: SONET (Synchronous Optical Network) is a high-speed synchronous transmission network specification, the maximum rate of up to 2.5 Gbps. The maximum number of SONET ports of a switch is the maximum number of downlinked SONET interfaces of the switch.

6. Backplane Throughput: Backplane throughput, also known as backplane bandwidth, is the number of packets per second (pps) that pass through the switch, and represents the maximum amount of data that can be throughput between the switch's interface processor or interface card and the data bus. The higher the backplane bandwidth of a switch, the more data it can handle, but also the higher the cost will be.

7. MAC address table size: Each port or device connected to the LAN requires a MAC address, which is used by other devices to locate specific ports and update routing tables and data structures. The size of a device's MAC address table reflects the maximum number of nodes connected to that device that can be supported.

8. Supported protocols and standards: The content of the protocols and standards supported by the LAN switch directly determines the switch's ability to adapt to the network. These protocols and standards generally refer to the networking specifications and equipment standards developed by the International Organization for Standardization. As the switch works on the second or third layer, the work involves all kinds of protocols below the third layer, generally speaking, according to the open interconnection network model can be categorized as follows.

(1) Layer 1 (Physical Layer) protocols Including EIA/TIA-232, EIA/TIA-449, X.21, and EIA530/EIA530A interface definitions, etc. These definitions basically determine the types and roles of the various physical interfaces on the switch.

(2) Layer 2 (Link Layer) protocols Including 802.1d/SPT, 802.1Q, 802.1p and 802.3x.

(3) Layer 3 (Network Layer) protocols Includes IP, IPX, RIP1/2, OSPF, BGP4, VRRP, and multicast protocols, among others.

LAN Switch Selection Factors

Users should mainly consider the following factors when selecting a LAN switch.

1, the choice of external size

If the network is large, or has completed the building level of integrated cabling, engineering requirements for network equipment on the rack centralized management, should be selected rack-mounted group-type switches or campus network switches. If there is no such need, desktop-type switches have a higher performance-price ratio.

2, scalability

The scalability of LAN switches is an important issue in choosing LAN switches. Good scalability is not just about a product having a large number of ports. This is because one of the most important things for a switch application is to determine under what conditions its ports will become congested. So users need to consider the following two aspects.

(1) Internal scalability What is the maximum scalability between 2 stacked switches? Bandwidth growth How many ports can be increased from 10Mbps to 100Mbps when the switch is not overloaded?

(2) External scalability Relates to the maximum rate at which the switch can be uplinked. For example, if there is 1 stackable LAN switch with 24 user ports, and assuming that all 24 ports are capable of transmitting traffic at 10Mbps and that the switch is uplinked at 1Gbps, then if 8 of the ports are increased to 100Mbps, this will cause the uplink to be saturated. This is because with 8 ports at 100 Mbps, the total traffic is 800 Mbps. and the remaining 16 ports, each at a rate of 10 Mbps, are only 160 Mbps total **** This gives a total of 960 Mbps of traffic for the 24 ports. indicating that this switch can no longer handle Fast Ethernet connections without congestion. If the switch is uplinked at 2Gbps, it can only handle a maximum of 19 Fast Ethernet ports or congestion will occur. Therefore, the scalability of the switch directly determines the ability to upgrade the transmission rate of each information point on the LAN.

3. Manageability

For LAN switches, the price paid in operation and management is, again, far more than the purchase cost. Based on this consideration, manageability has begun to become another key factor in evaluating switches.

Generally speaking, the switch itself has a certain degree of manageability, as for the stackable switch also has the advantage of several stacked switches can be managed as a switch, without the need to manage and monitor each LAN switch separately. It is important to note that among the manageable elements are the quality of service (QoS) for handling prioritized traffic, the ability to enhance policy management, the ability to manage virtual LAN traffic, and the ease of configuration and operation. Of these, QoS performance is primarily concerned with reserving the bandwidth needed to support the demands of different service levels. Manageability also involves the switch's support for policies, which are sets of rules that control switch operation. Network administrators use policies to allocate bandwidth and assign priorities to each application traffic and control network access. The focus is on bandwidth management policies and they must meet service level agreement SLAs. distributed policies are an important element of stacking into a stacked switch, and stackable switches should be checked to see if they support directory management features, such as Lightweight Directory Access Protocol (LDAP), to improve the manageability of the switch.

4. Port bandwidth and type

What type of LAN switch to choose, users should first decide based on their own networking bandwidth needs, and then consider from the switch port bandwidth design. From the configuration of port bandwidth, there are three main types on the market today.

The first configuration: n × 10M + m × 100M low-fast port dedicated

The general transmission rate of the backbone network for the 100Mbps full-duplex, the branch rate of 10Mbps. from a technical point of view, this type of configuration of the LAN switch strictly limits the upgrade of the network, the user is unable to realize the high-speed multimedia network, and therefore domestic and foreign manufacturers have basically stopped the production of this product.

Second Configuration: n×10/100Mbps Port Adaptive

Currently, these switches are the dominant products in the market because they have Auto Negotiation, which detects whether the bandwidth of the downlinked device is 100M or 10M, and whether it's full-duplex or half-duplex. When the NIC is connected to a switch, if the NIC supports full duplex, this link can send and receive 100M each, realizing 200M bandwidth, and the same may occur in a switch-to-switch connection, with a very relaxed application environment.

The third configuration: n × 1000M + m × 100M high-speed ports dedicated

With the first type of switch configuration is similar, the difference is that not only the bandwidth to a few orders of magnitude more, but also completely different port types. Switches with this configuration scheme are important equipment in current high-speed network and fiber-optic network access schemes, which can completely solve the bottleneck problem between network servers. For example, 3Com's 3C39024 (1×1000SX + 24×10/100BaseTX) and 3C39036 (1×1000SX + 36×10/100BaseTX) Gigabit uplinks to servers solves the server-to-server bottleneck problem. However, the cost is much higher than the previous two categories.