Overview of Stacks

Definition

A cluster switch system (CSS) can have two CSS-capable switches connected through cables to work as a single logical switch for data forwarding, as shown in Figure 1-1. A CSS is also called a stack. The term stack is used throughout this document.

Figure 1-1 Stack topology

Scenarios

• Improving reliability

  Stack members work in redundancy mode and use inter-device Eth-Trunks to implement inter-device link redundancy.

• Increasing bandwidth and implementing inter-chassis link redundancy

  In Figure 1-2, when the network scale expands and higher uplink bandwidth is required, new switches can be added to set up a stack with the original switches. Multiple physical links of stack members can be configured as an aggregation group to improve the uplink bandwidth of the switches.

  In addition, the downstream switches of the stack can be connected to the stack through inter-device link aggregation to implement device backup and inter-device link redundancy, which improves reliability.

  Figure 1-2 Increasing bandwidth and implementing inter-chassis link redundancy
  
• Simplifying networking
  In Figure 1-3, two switches on the network set up a stack to form a single logical switch. The simplified networking does not require protocols such as MSTP and VRRP, making network configuration much simpler. The use of inter-device link aggregation achieves fast convergence and improves reliability.

  Figure 1-3 Simplified networking
  
• Providing long-distance stacking
  Long-distance stacking allows devices that are far away from each other to set up a stack. In Figure 1-4, the aggregation switches on the two floors are connected over a long distance to set up a stack. In this way, the two switches externally act as a single aggregation switch. This simplifies the network structure and reduces the costs of device management and maintenance. Each floor has two links to the core network, which greatly enhances reliability.

  Figure 1-4 Long-distance stacking