What is the difference between a switch and a router?

A switch (English: Switch, meaning "switch") is a network device used for electrical signal forwarding. It provides an exclusive electrical signal path for any two network nodes that access the switch. The most common switch is an Ethernet switch. Other common ones are dian voice switches, fiber switches, and so on.

First, the main features

The main functions of the switch include physical addressing, network topology, error checking, frame sequence, and flow control. At present, the switch also has some new features, such as support for VLAN (Virtual Local Area Network), support for link aggregation, and even some have firewall functions.

Learning: The Ethernet switch knows the MAC address of the device connected to each port, and maps the address to the corresponding port and stores it in the MAC address table in the switch cache.

Forward/Filter: When the destination address of a data frame is mapped in the MAC address table, it is forwarded to the port connected to the destination node instead of all ports (if the data frame is a broadcast/multicast frame, it is forwarded to all ports) .

Eliminate loops: When the switch includes a redundant loop, the Ethernet switch avoids loop generation through the spanning tree protocol, while allowing the existence of a backup path.

In addition to being able to connect to the same type of network, switches can also interconnect between different types of networks, such as Ethernet and Fast Ethernet. Many switches today offer high-speed ports that support Fast Ethernet or FDDI to connect to other switches in the network or to provide additional bandwidth for critical servers with large bandwidth.

In general, each port of the switch is used to connect to a separate network segment, but sometimes in order to provide faster access speed, we can connect some important network computers directly to the ports of the switch. In this way, the network's key servers and important users have faster access speeds and support greater traffic.

Second, the basic functions of the switch:

1. Like a hub, the switch provides a large number of ports for cable connections, allowing for star topology routing.

2. Like repeaters, hubs, and bridges, when it forwards a frame, the switch regenerates an undistorted square electrical signal.

3. Like a bridge, the switch uses the same forwarding or filtering logic on each port.

4. Like a bridge, a switch divides a LAN into multiple collision domains. Each collision domain has an independent broadband, thus greatly increasing the bandwidth of the LAN.

5. In addition to the capabilities of bridges, hubs, and repeaters, the switch provides more advanced features such as virtual local area networks (VLANs) and higher performance.

Third, the basic terminology

The transmission mode of the switch is full-duplex, half-duplex, full-duplex/half-duplex adaptive. The full-duplex of the switch means that the switch can receive data while sending data. The two are synchronized, which seems like our usual time. Just like dian, you can hear each other while you speak. Current switches support full duplex. The advantage of full duplex is that the delay is small and the speed is fast.

When it comes to full-duplex, you can't help but mention another concept that closely corresponds to it, that is, "half-duplex". The so-called half-duplex means that only one action occurs in a time period. For a simple example, a narrow At the same time, only one car can pass at the same time. When there are two cars facing each other, in this case, only one can pass first. When the other car is opened again after the first time, this example shows the half. The principle of duplex. Early walkie-talkies, as well as early hubs, were half-duplex products. With the continuous advancement of technology, half-double unions have gradually withdrawn from the historical arena.

Broadly speaking, there are two types of network switches: WAN switches and LAN switches. WAN switches are mainly used in the telecommunications field to provide a basic platform for communication. LAN switches are used in local area networks to connect terminal devices such as PCs and network printers. From the transmission medium and transmission speed, it can be divided into Ethernet switches, Fast Ethernet switches, Gigabit Ethernet switches, FDDI switches, ATM switches and Token Ring switches. From the scale application, it can be divided into enterprise-level switches, department-level switches and workgroup switches. The scale of each vendor is not completely consistent. Generally speaking, enterprise switches are rack-mounted, and department-level switches can be rack-mounted (small number of slots) or fixed-configuration, and workgroups. The level switch is a fixed configuration (simplified function). On the other hand, from the perspective of the scale of the application, when it is a backbone switch, the switch supporting large enterprise applications with more than 500 information points is an enterprise-class switch, and the switch supporting medium-sized enterprises with 300 information points or less is a department-level switch, and supports 100. The switches within the information points are workgroup switches. The switch described in this document refers to a LAN switch.

Fourth, the principle

Working at the data link layer, the switch has a very high bandwidth back bus and internal switching matrix. All the ports of the switch are connected to the back bus. After the control circuit receives the data packet, the processing port searches the address comparison table in the memory to determine the NIC (network card) of the destination MAC (the hardware address of the network card). On which port, the packet is quickly transmitted to the destination port through the internal switching matrix. If the destination MAC does not exist, it will be broadcast to all ports. After receiving the port response, the switch will "learn" the new address and add it to the internal MAC. In the address table. The switch can also be used to "segment" the network. By comparing the MAC address table, the switch only allows the necessary network traffic to pass through the switch. Through the filtering and forwarding of the switch, the collision domain can be effectively reduced, but it cannot divide the network layer broadcast, that is, the broadcast domain. The switch can perform data transmission between multiple port pairs at the same time. Each port can be regarded as a separate network segment, and the network devices connected to it can enjoy the full bandwidth alone, without competing with other devices. When node A sends data to node D, node B can send data to node C at the same time, and both of these transmissions enjoy the full bandwidth of the network and have their own virtual connections. If a 10 Mbps Ethernet switch is used here, then the total throughput of the switch at this time is equal to

2×10 Mbps=20 Mbps, and when using a 10 Mbps shared HUB, the total throughput of one HUB will not exceed 10 Mbps. In short, a switch is a network device that can perform packet encapsulation and forwarding of data frames based on MAC address identification. The switch can "learn" the MAC address and store it in the internal address table. By establishing a temporary switching path between the originator and the target receiver of the data frame, the data frame can be directly sent from the source address to the destination address.

Five, the difference between switches and routers

Traditional switches have evolved from bridges and belong to the second layer of OSI, the data link layer device. It is addressed according to the MAC address, routing is selected through the station table, and the establishment and maintenance of the station table is automatically performed by the switch. The router belongs to the OSI layer 3, the network layer device, which is addressed according to the IP address and generated by the routing table routing protocol. The biggest advantage of the switch is fast. Because the switch only needs to identify the MAC address in the frame, it is simple to generate the selective forwarding port according to the MAC address, which is convenient for the ASIC to implement, so the forwarding speed is extremely high.

But the working mechanism of the switch also brings some problems.

1. Loop: According to the switch address learning and station table establishment algorithm, loops are not allowed between switches. Once a loop exists, the spanning tree algorithm must be started to block the port that generated the loop. The router's routing protocol does not have this problem. There can be multiple paths between routers to balance the load and improve reliability.

2. Load concentration: There can only be one path between switches, so that information is concentrated on one communication link, and dynamic allocation cannot be performed to balance the load. The routing protocol algorithm of the router can avoid this. The OSPF routing protocol algorithm can not only generate multiple routes, but also select different optimal routes for different network applications.

3. Broadcast Control: The switch can only narrow down the collision domain, but not the broadcast domain. The entire switched network is a large broadcast domain, and broadcast messages are scattered throughout the switched network. The router can isolate the broadcast domain, and broadcast packets cannot continue to be broadcast through the router.

4. Subnetting: The switch can only recognize the MAC address. The MAC address is a physical address and has a flat address structure, so the subnet cannot be divided according to the MAC address. The router recognizes the IP address. The IP address is assigned by the network administrator. It is a logical address and the IP address has a hierarchical structure. It is divided into a network number and a host number, which can be conveniently used to divide the subnet. The main function of the router is to use Connect to different networks.

5. Confidentiality: Although the switch can also filter the frame based on the source MAC address, destination MAC address, and other frame contents of the frame, the router reports the source IP address, destination IP address, and TCP port address of the packet. Text filtering is more intuitive and convenient.

Six, the difference between switches and hubs

1. From the perspective of the OSI architecture, the hub belongs to the first layer of physical layer equipment, and the switch belongs to the second layer of data link layer equipment of the OSI. That is to say, the hub only plays a role in synchronizing, amplifying and shaping the data transmission. The short frame=fragment in the data transmission cannot be effectively processed, and the integrity and correctness of the data transmission cannot be guaranteed; Synchronize, amplify, and shape the data, and filter short frames, fragments, and so on.

2. From the working mode, the hub is a broadcast mode, that is to say, when a certain port of the hub works, all other ports can listen to the information, which is easy to generate a broadcast storm. When the network is large, the network performance will be very high. The switch can avoid this phenomenon. When the switch works, only the requesting port and the destination port respond to each other without affecting other ports, so the switch can isolate the collision domain and effectively suppress the broadcast storm. produce.

3. From the perspective of bandwidth, no matter how many ports a hub has, all ports share a bandwidth. Only two ports can transmit data at the same time. Other ports can only wait, and the hub can only work in half-duplex mode. For the switch, each port has an exclusive bandwidth. When the two ports work, it does not affect the work of other ports. At the same time, the switch can work in half-duplex mode and can work in full-duplex mode. .

Seven, the exchange method

The switch is exchanged in the following three ways:

1) Straight-through:

A straight-through Ethernet switch can be understood as a line matrix dian switch that crosses between ports. When it detects a packet on the input port, it checks the packet header of the packet, obtains the destination address of the packet, starts the internal dynamic lookup table and converts it into the corresponding output port, connects it at the intersection of the input and output, and passes the packet to the packet. The corresponding port implements the switching function. Since there is no need for storage, the delay is very small and the exchange is very fast, which is its advantage. Its disadvantage is that because the contents of the packet are not saved by the Ethernet switch, it is impossible to check whether the transmitted packet is incorrect or not, and cannot provide error detection capability. Since there is no buffer, it is not possible to directly connect input/output ports with different rates, and it is easy to drop packets.

2) Store and forward:

The store-and-forward method is the most widely used method in the field of computer networks. It stores the data packet of the input port first, and then performs CRC (Cyclic Redundancy Check) check. After processing the error packet, the destination address of the data packet is taken out, and the lookup table is converted into an output port to send the packet. For this reason, the store-and-forward method has a large delay in data processing, which is its deficiency, but it can detect errors in packets entering the switch and effectively improve network performance. Especially important is that it can support the conversion between ports of different speeds, keeping the high-speed port and low-speed port work together.

3) Fragment isolation:

This is a solution between the first two. It checks if the length of the packet is 64 bytes. If it is less than 64 bytes, it is a fake packet, the packet is discarded; if it is larger than 64 bytes, the packet is sent. This method also does not provide data verification. Its data processing speed is faster than the store-and-forward method, but slower than the straight-through.