Talking about the cascade and stack of switches

The cascaded expansion mode is the most conventional and direct way of expansion. Some older networks use hubs (HUBs) as cascaded devices. Because hubs are already quite expensive at the time, most enterprises cannot choose switches as tie-up devices. That is because most of the workgroup user access requirements are generally connected to the hub from a port level on the hub. In this way, the access capability has been greatly improved, but due to some interference and human factors, the overall performance is very low, only to meet the needs of multi-port, simply do not consider the forwarding and switching capabilities. The current cascaded extension mode comprehensively considers the forwarding performance and port attributes of different switches, and can easily achieve multi-user access through a certain topology structure design.

The cascaded mode is the most ideal way to build a large-scale LAN. It can comprehensively use various topology design technologies and redundancy technologies to implement a hierarchical network structure, such as designing redundancy through a dual-homing topology, and implementing redundancy through Link Aggregation technology. With the bandwidth expansion of Up LINK, these technologies are now very mature and widely used in various LANs and MANs.

The cascaded mode uses common Ethernet ports for inter-layer interconnections such as 100M FE ports, GE ports, and emerging 10GE ports.

The cascaded mode is the mainstream technology in Ethernet expansion port applications. It uses a unified network management platform to achieve unified management of the entire network equipment, such as topology management and fault management. The cascaded mode also faces challenges. When the number of cascaded layers is large and there is a large convergence ratio between layers, the edge nodes will experience some delays due to more exchanges and buffers. . The solution is to converge the uplink ports to reduce the convergence ratio, improve the performance of the upper-end device, or reduce the level of cascading. In the cascade mode, in order to ensure the efficiency of the network, it is generally recommended not to exceed four levels. If the network edge node has a port that is extended through a broadcast Ethernet device such as HUB, since it is a direct-through operation mode, there is no exchange and it is not included in the hierarchy. However, it should be noted that the CSMA/CD mechanism of the HUB works because The network performance impact caused by the loopback caused by the conflict will be far greater than the impact of the switch-level connection.

The cascaded mode is an inevitable choice for the formation of a structured network. The use of universal cables (optical fiber) is used in series, and each component can be placed in any position, which is very conducive to integrated wiring.

Stacking technology expansion

Stacking technology is another type of technology that is currently used in Ethernet switches and is a non-standardized technology. Mixing and stacking are not supported between vendors. The stacking mode is defined by each vendor and does not support the topology. Currently there are two main stacking modes: daisy chain mode and star mode. The biggest advantage of stacking technology is that it provides simplified local management and manages a group of switches as one object.

Daisy chained stacking

Daisy-chained stacking is a stacking technology based on cascaded architecture. There is no special requirement on the hardware of the switch. Through the relatively high-speed port serial connection and software support, a multi-switch stacking structure is finally implemented through loops. Redundancy can be achieved to some extent. However, in terms of exchange efficiency, the same-level model is at the same level. Daisy-chain stacking usually uses a high-speed port and two high-speed ports. The structure of the two is shown in Figure 2. When using a high-speed port (GE) mode, sending and receiving are performed on the same port, upstream and downstream, respectively, and a ring structure is finally formed. Data exchange between any two member switches needs to be circulated once through the switch ports of all the switches. The efficiency is low, especially when the number of stack layers is large, the stack ports become a serious system bottleneck. Using two high-speed ports for daisy-chaining, you can choose to implement ring redundancy because you use more high-speed ports. Compared with cascaded mode, daisy-chained stacking mode does not have topology management. Generally, it cannot be deployed in a distributed manner. It is applicable to single-node organizations with high-density port requirements and can be used at the edge of the network.