Cisco Three-Layer Hierarchical Model

Cisco has its own networking model made up of three layers ”the Core layer, the Distribution layer, and the Access layer. Together, these three layers are called the Cisco Hierarchical model.

Large networks are extremely complicated and use multiple protocols, large configurations, and very diverse technologies. A hierarchical model helps to summarize a complex collection of equipment into an understandable model. When you place equipment properly in a network design, the network will function more efficiently . The Cisco Hierarchical model is used to design, implement, and maintain a scalable, reliable, cost-effective hierarchical internetwork. Each of the layers has specific responsibilities, and the layers are logical and not necessarily physical. Sometimes a single device can be placed in multiple layers, as well.

At the top of the Cisco Hierarchical model is the Core layer; at the bottom of the model is the Access layer; these layers are separated by the Distribution layer. Cisco classifies switches and routers the same way it classifies networks. The better the device functions and the more features the device has, the higher the layer in which it gets placed. Figure 2.3 illustrates the three layers of the Cisco Hierarchical model.

Figure 2.3. The Cisco Hierarchical model.

A hierarchy offers the benefits of helping you position certain routers and switches in identifiable logical places in your network. For example, a very popular router is the Cisco 1700 series, which is used in small offices and can't support more than about 50 or so users. In a large organization you would put that type of router at the Access layer in the Cisco Hierarchical Model. You wouldn't want to place that router as the core router for a company with 40 locations, for example, because the processing required for just a few locations alone would overwhelm that size router.

The Core Layer

The Core layer is the heart of the network. This layer is responsible for transporting the largest amount of data and sometimes voice traffic. The Core layer needs to be reliable, have the lowest latency possible, and work as quickly as possible. A failure or routing problem at the Core layer means that multiple locations and most users are affected. Placing fault tolerant devices at this layer is critical to network operations.

You should never have the devices in the Core layer assigned to anything that might slow down network traffic. You should avoid, for example, configuring a router to use access lists or packet filtering, route between virtual local area networks (VLANs), or allow support for workgroup access at the Core layer. As the network grows larger, you also should avoid expanding by adding additional routers or switches at the Core layer. Instead, upgrade or replace the devices at this layer to accommodate network growth.

The Core layer should facilitate speed and redundancy. It is common to find Gigabit Ethernet and Fast Ethernet (with redundant links) or Asynchronous Transfer Mode (ATM) interfaces at this layer. By using routing protocols with lower convergence times, you can increase the speed of data traveling at this layer as well.

The Distribution Layer

The Distribution layer is commonly referred to as the workgroup layer . This layer is responsible for serving as the communication point between the Access layer and the Core layer. This layer of the network is kind of like the network's post office. Access lists and route maps used with routing protocols act as the postman in the Distribution layer, deciding whether the packets traversing the network are valid or junk mail, and determining the route that packets will take. If the Distribution layer determines that the packet is junk, the packet is not allowed to go to another point in the network block and is returned. You need to remember that this layer provides routing, filtering, packet access to the Core layer, and WAN access to other locations or the Internet. The Distribution layer also is the layer where access lists, packet filtering, and access policies for the network should be enforced. An organized and maintained security plan and firewalls should be placed at this layer. Routing protocol redistribution, VLAN routing, and static routing also should be located in this layer.

The Access Layer

The Access layer is where the local users reside; this layer is sometimes referred to as the desktop layer . The network resources that most users need will be available locally. Any data or voice traffic destined for remote locations are handled by the distribution layer.

Access control policies, such as access lists on the router, can still be implemented here in this layer, as well as the creation of separate collision domains. Frequently, static routing is used instead of dynamic routing protocols at this layer.