<\/span><\/h3>\n<\/li>\n<\/ul>\n(1) Fixed-configuration switch can be understood as an iron box. Generally, fixed-configuration switch has fixed number of ports, fixed power modules, fans, etc.; Therefore, fixed-configuration switch is typically not expandable.<\/p>\n
(2) Fixed-configuration switch supports stacking technology\u00a0to improve scalability, so we can form multiple fixed-configuration switch into one switch.<\/p>\n
![\"Chain](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2023\/06\/Chain-connection-and-ring-connection.png\")
<\/p>\n
Chain connection and ring connection<\/span><\/strong><\/p>\n3) Under normal circumstances, fixed-configuration switch is used in a network access layer or aggregation layer.<\/p>\n
\n- \n
<\/span>Modular switch<\/strong><\/span><\/span><\/h3>\n<\/li>\n<\/ul>\nA modular switch is based on a chassis, and interface boards, switch boards, and power modules can be configured independently according to requirements. The scalability of a frame switch is generally based on the number of slots. Modular switches are generally used in the core position\u00a0of network.<\/p>\n
![\"Networking](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2023\/06\/Networking-diagram-1.png\")
<\/p>\n
Networking diagram<\/span><\/strong><\/p>\nAs shown in the network diagram above: In the data center network, CE5800, CE6800, and CE8800 are box-type devices, which are generally used as the access layer; The CE128 is a frame-type device and is generally used as the core layer.<\/p>\n
Therefore, when selecting a device, you can determine whether to choose fixed-configuration switch or a modular switch based on the actual use level of the switch.<\/p>\n
<\/span>Function<\/b><\/strong><\/span><\/span><\/h2>\nSwitches are classified according to the working protocol layer: they\u00a0can be divided into Layer 2 switches and Layer 3 switches.<\/p>\n
Differences between Layer 2 switches and Layer 3 switches\uff1a<\/p>\n
<\/span>Layer 2 Switch:<\/strong><\/span><\/span><\/h3>\nMain functions of the switches that work on the second layer of the data link layer of the OSI reference model, include physical addressing, error checking, frame sequence and flow control. (As shown in the figure below, the Layer 2 switch works at the data link layer and can process data frames)<\/p>\n
![\"Layer](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2023\/06\/Layer-2-switch.png\")
<\/p>\n
Layer 2 switch<\/span><\/strong><\/p>\n<\/span>Layer 3 Switch:<\/span><\/strong><\/span><\/h3>\nA device with three-layer switching function, is a second-layer switch with third-layer routing function, which is an organic combination of the two, not simply superimposing the hardware and software of the router device on the LAN switch. (As shown in the figure below, the three-layer switch works at the network layer and can process data packets)<\/p>\n
![\"Layer](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2023\/06\/Layer-3-switch-1.png\")
<\/p>\n
Layer 3 switch<\/span><\/strong><\/p>\n<\/span>Number of ports<\/b><\/strong><\/span><\/span><\/h2>\n<\/span>Fixed-configuration switch<\/span><\/strong><\/span><\/h3>\nThe number of ports that a switch can provide is basically fixed for each type of fixed-configuration switch, generally 24 or 48 access ports and 2-4 uplink ports. Take Huawei CE5850-48T4S2Q-EIas an example (as shown below). There are 48 1000M access ports, 4 10G uplink ports, and 2 40G uplink ports;<\/p>\n
![\"Sample](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2023\/06\/Sample-diagram-of-a-fixed-configuration-switch2.png\")
<\/p>\n
Sample diagram of a fixed-configuration switch<\/span><\/strong><\/p>\n<\/span>Modular switch<\/span><\/strong><\/span><\/h3>\nThe number of ports of modular switches are related to the number of boards configured, which is generally the maximum number of ports that each chassis can support when the interface board with the highest density is configured. Take Huawei’s CE12804 as an example, which supports 4 service board LPUs, and the ports are related to the specific board model. For a 36-port 100G board, there are a total of 144 100G ports when the board is fully inserted.<\/p>\n
![\"Sample](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2023\/06\/Sample-diagram-of-a-modular-switch2.png\")
<\/p>\n
Sample diagram of a modular switch<\/span><\/strong><\/p>\n<\/span>How to choose a switch based on the number of ports:<\/strong><\/span><\/span><\/h3>\nWhen selecting a switch, it needs to be based on the current business situation and future scalability. The number of switch ports represents the number of terminals you need to access.<\/p>\n
For a switch with 48 access ports, if one terminal occupies one port, then one switch can connect to 48 terminals. Five such switches are needed in a company with 200 employees.<\/p>\n
<\/span>Port Speed<\/b><\/strong><\/span><\/span><\/h2>\n<\/span>The port speeds that switch supports:<\/strong><\/span><\/span><\/h3>\nThe port speeds provided by the current switch include100Mbps\/1000Mbps\/10Gbps\/25Gbps and so on.<\/p>\n
<\/span>\u00a0Port speed unit of the switch:<\/strong><\/span><\/span><\/h3>\nThe port speed unit of the switch is bps (bit per second).<\/strong><\/p>\n![\"Switch](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2023\/06\/Switch-port.png\")
<\/p>\n
Switch port<\/span><\/strong><\/p>\n<\/span>Exchange Capacity<\/b><\/strong><\/span><\/span><\/h2>\n<\/span>Switching capacity of\u00a0<\/strong>switch<\/b><\/span><\/span><\/h3>\nSwitching capacity of\u00a0switch is also known as backplane bandwidth or switching bandwidth.<\/p>\n
Switching capacity is the maximum amount of data that can be handled between the switch interface processor (or interface card) and the data bus.<\/p>\n
The backplane bandwidth marks the total data exchange capacity of the switch , and the unit is Gbit\/s.\u00a0The higher the switching capacity of a switch, the stronger the ability to process data, but at the same time the higher the design cost. Twice the capacity of all ports multiplied the number of ports should be less than the switching capacity, so as to realize full-duplex non-blocking switching.<\/p>\n
<\/span>\u00a0The switching capacity is related to the standard of the\u00a0<\/strong><\/span>switch.<\/span><\/span><\/h3>\nFor bus switches, switching capacity refers to the bandwidth of the backplane bus.<\/p>\n
![\"bus](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2023\/06\/bus-switch.png\")
<\/p>\n
Bus switch<\/span><\/strong><\/p>\nFor a switch with switch matrix, the switching capacity refers to the total interface bandwidth of the switch matrix.<\/p>\n
![\"a](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2023\/06\/a-switch-with-switch-matrix.png\")
<\/p>\n
A switch with switch matrix<\/span><\/strong><\/p>\nThis switching capacity is a theoretical calculation value, but it represents the maximum possible switching capacity of the switch. The current switch design ensures that this parameter will not become the bottleneck of the entire switch.<\/p>\n
<\/span>Packet forwarding rate<\/b><\/strong><\/span><\/span><\/h2>\n<\/span>Packet forwarding rate of switch:<\/strong><\/span><\/span><\/h3>\nPacket forwarding rate, also known as interface throughput, refers to the data packet forwarding capability on an interface of a communication device, and the unit is usually pps (packet per second). The packet forwarding rate of the switch is generally the result of actual measurement, which represents the actual forwarding performance of the switch.<\/p>\n
<\/span>\u00a0Calculation method of packet forwarding rate:<\/span><\/strong><\/span><\/h3>\nThe measurement standard of the packet forwarding rate is based on the number of 64-byte data packets (minimum packets) sent per unit time.<\/p>\n
When calculating the packet forwarding rate, the fixed overhead of preamble and frame gap needs to be considered.<\/p>\n
By default, the interframe gap is a maximum of 12 bytes, and users are advised to use the default configuration. If the user modifies the frame gap of the interface to a smaller value, the receiving end may not have enough time to receive the next frame after receiving a data frame, resulting in the inability to process forwarded packets in time and packet loss.<\/p>\n
![\"The](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2023\/06\/The-interframe-gap-by-default.png\")
<\/p>\n
The interframe gap by default<\/span><\/strong><\/p>\nWe know that the length of the Ethernet frame is variable, but the processing power used by the switch to process each Ethernet frame has nothing to do with the length of the Ethernet frame. Therefore, when the interface bandwidth of the switch is constant, the shorter the Ethernet frame length, the more frames the switch needs to process, and the more processing power it needs to consume.<\/p>\n
<\/span>Supplementary points<\/span><\/strong><\/span><\/h2>\n<\/span>When to Use Multiple Routing Protocols?<\/b><\/strong><\/span><\/span><\/h3>\nMultiple routing protocols are used when two different routing protocols need to exchange routing information. Of course, route redistribution can also exchange routing information. The following situations do not need to use multi-routing protocols:<\/p>\n
\n- Upgrade from the old version of the Interior Gateway Protocol (IGP) to the new version of the IGP.<\/li>\n
- You want to use another routing protocol but must keep the original one.<\/li>\n
- You want to terminate internal routes, so they don’t get interfered by other routers that don’t have strict filtering policing.<\/li>\n
- You are in an environment composed of routers from multiple manufacturers.<\/li>\n<\/ol>\n
<\/span>What is a Distance Vector Routing Protocol?<\/b><\/strong><\/span><\/span><\/h3>\n