Deployment Differences Between Two-Layer and Three-Layer Network Architectures
The tree topology is recommended as the physical architecture of a campus network. This topology facilitates network deployment and management, and has good scalability. In most cases, a campus network using the tree topology has a hierarchical architecture that consists of the terminal layer, access layer, aggregation layer, and core layer. In actual deployment, you can flexibly select a two-layer or three-layer network architecture based on the network scale and service requirements.
Two-Layer Network Architecture
Figure 2-1 shows a two-layer network architecture that consists of a core layer and an access layer.
To ensure device-level and link-level reliability on the network, it is recommended that CSS be configured at the core layer, stacking be configured at the access layer, and core and access devices be connected through Eth-Trunk interfaces. If standalone access devices can provide sufficient access capacity for downstream terminals, you do not need to configure stacking at the access layer.
The networking where CSS, stacking, and Eth-Trunk are used is loop-free. The configuration is simple because complex ring network protocols (such as RSTP, MSTP, and RRPP) and reliability protocols do not need to be configured. The networking ensures device-level and link-level reliability, simplifies the network topology, and reduces the deployment and maintenance workload.
Three-Layer Network Architecture
Figure 2-2 shows a three-layer network architecture that consists of a core layer, an aggregation layer, and an access layer.
To ensure device-level and link-level reliability on the network, it is recommended that CSS be configured at the core layer, stacking be configured at the aggregation and access layers, and core, aggregation, and access devices be connected through Eth-Trunk interfaces. If standalone access devices can provide sufficient access capacity for downstream terminals, you do not need to configure stacking at the access layer.
Deployment Differences
The difference between the two network architectures is that the three-layer network architecture has the aggregation layer, whereas the two-layer network architecture does not have the layer. The aggregation layer is between the core and access layers and connects to both layers. Aggregation switches aggregate traffic from access switches, process the traffic, and provide uplinks to the core layer.
- Network scale: For example, the number of NEs is related to the investment cost.
- Network complexity: The network maintenance cost and fault locating complexity vary depending on the network complexity. The more complex the network is and the more failure points are, the more difficult fault locating is and the higher the maintenance cost is.
- Transmission distance: A network using the three-layer architecture is larger than a network using the two-layer architecture when the differences between transmission media are not considered.
In general, the two-layer network architecture is applicable to small-scale campuses because it is simple and contains a small number of NEs, and a network constructed using this architecture has fewer failure points. The three-layer network architecture is applicable to large-scale campuses because it is complex and contains a large number of NEs, and a network constructed using this architecture has more failure points.
The two-layer network architecture is usually used in actual deployment. If the transmission distance is short and access devices can be directly connected to core devices that provide enough interfaces, the aggregation layer can be omitted, which is a common practice. This reduces the total cost and maintenance workload, and facilitates network status monitoring.