LAN Design

When building a LAN that satisfies the needs of a small- or medium-sized business, your plan is more likely to be successful if a hierarchical design model is used. Compared to other network designs, a hierarchical network is easier to manage and expand, and problems are solved more quickly.

Hierarchical network design involves dividing the network into discrete layers. Each layer provides specific functions that define its role within the overall network. By separating the various functions that exist on a network, the network design becomes modular, which facilitates scalability and performance. The typical hierarchical design model is broken up in to three layers: access, distribution, and core

The access layer interfaces with end devices, such as PCs, printers, and IP phones, to provide access to the rest of the network. The access layer can include routers, switches, bridges, hubs, and wireless access points (AP). The main purpose of the access layer is to provide a means of connecting devices to the network and controlling which devices are allowed to communicate on the network.

The distribution layer aggregates the data received from the access layer switches before it is transmitted to the core layer for routing to its final destination. The distribution layer controls the flow of network traffic using policies and delineates broadcast domains by performing routing functions between virtual LANs (VLANs) defined at the access layer. Distribution layer switches are typically high-performance devices that have high availability and redundancy to ensure reliability

The core layer of the hierarchical design is the high-speed backbone of the internetwork. The core layer is critical for interconnectivity between distribution layer devices, so it is important for the core to be highly available and redundant. The core area can also connect to Internet resources. The core aggregates the traffic from all the distribution layer devices, so it must be capable of forwarding large amounts of data quickly. 

Network Diameter

When designing a hierarchical network topology, the first thing to consider is network diameter. Diameter is usually a measure of distance, but in this case, we are using the term to measure the number of devices. Network diameter is the number of devices that a packet has to cross before it reaches its destination. Keeping the network diameter low ensures low and predictable latency between devices.

Bandwidth Aggregation

Each layer in the hierarchical network model is a possible candidate for bandwidth aggregation. Bandwidth aggregation is the practice of considering the specific bandwidth requirements of each part of the hierarchy. After bandwidth requirements of the network are known, links between specific switches can be aggregated, which is called link aggregation. Link aggregation allows multiple switch port links to be combined so as to achieve higher throughput between switches. Cisco has a proprietary link aggregation technology called EtherChannel, which allows multiple Ethernet links to be consolidated

Redundancy

Redundancy is one part of creating a highly available network. Redundancy can be provided in a number of ways. For example, you can double up the network connections between devices, or you can double the devices themselves. Implementing redundant links can be expensive. Imagine if every switch in each layer of the network hierarchy had a connection to every switch at the next layer. It is unlikely that you will be able to implement redundancy at the access layer because of the cost and limited features in the end devices, but you can build redundancy into the distribution and core layers of the network.

Switch Features

When you are selecting a switch, you need to decide between fixed configuration or modular configuration, and stackable or non-stackable. Another consideration is the thickness of the switch expressed in number of rack units. Fixed configuration switches are just as you might expect, fixed in their configuration. What that means is that you cannot add features or options to the switch beyond those that originally came with the switch. The particular model you purchase determines the features and options available. For example, if you purchase a 24-port gigabit fixed switch, you cannot add additional ports when you need them.

Modular Switches

Fixed configuration switches are just as you might expect, fixed in their configuration. What that means is that you cannot add features or options to the switch beyond those that originally came with the switch. The particular model you purchase determines the features and options available. For example, if you purchase a 24-port gigabit fixed switch, you cannot add additional ports when you need them.


Stackable Switches

Stackable switches can be interconnected using a special backplane cable that provides high-bandwidth throughput between the switches. Cisco introduced StackWise technology in one of its switch product lines. StackWise allows you to interconnect up to nine switches using fully redundant backplane connections.

The stacked switches effectively operate as a single larger switch. Stackable switches are desirable where fault tolerance and bandwidth availability are critical and a modular switch is too costly to implement. Using cross-connected connections, the network can recover quickly if a single switch fails

Performance

When selecting a switch for the access, distribution, or core layer, consider the ability of the switch to support the port density, forwarding rates, and bandwidth aggregation requirements of your network.

Port density is the number of ports available on a single switch. Fixed configuration switches typically support up to 48 ports on a single device, with options for up to four additional ports for small form-factor pluggable (SFP) devices. Modular switches can support very high port densities through the addition of multiple switch port line cards

Forwarding rates define the processing capabilities of a switch by rating how much data the switch can process per second. Switch product lines are classified by forwarding rates. Entry-layer switches have lower forwarding rates than enterprise-layer switches. Forwarding rates are important to consider when selecting a switch. If the switch forwarding rate is too low, it cannot accommodate full wire-speed communication across all of its switch ports. Wire speed is the data rate that each port on the switch is capable of attaining, either 100 Mb/s Fast Ethernet or 1000 Mb/s Gigabit Ethernet

As part of bandwidth aggregation, you should determine if there are enough ports on a switch to aggregate to support the required bandwidth. Link aggregation helps to reduce bottlenecks of traffic by allowing up to eight switch ports to be bound together for data communications. Cisco uses the term EtherChannel when describing aggregated switch ports.

Two other characteristics you want to consider when selecting a switch are Power over Ethernet (PoE) and Layer 3 functionality.


Power over Ethernet

Power over Ethernet (PoE) allows the switch to deliver power to a device over the existing Ethernet cabling. This feature can be used by IP phones and some wireless access points. PoE allows you more flexibility when installing wireless access points and IP phones because you can install them anywhere you can run an Ethernet cable. You do not need to consider how to run ordinary power to the device. You should only select a switch that supports PoE if you are actually going to take advantage of the feature, because it adds considerable cost to the switch.

Layer 3 switches offer advanced functionality. Layer 3 switches are also known as multilayer switches.

Switch Features in a Hierarchical Network

Now that you know which factors to consider when choosing a switch, let us examine which features are required at each layer in a hierarchical network. Access layer switches facilitate the connection of end node devices to the network. For this reason, they need to support features such as port security, VLANs, Fast Ethernet/Gigabit Ethernet, PoE, and link aggregation. Port security is applied at the access layer. Consequently, it is an important first line of defense for a network.

Port speed is also a characteristic you need to consider for your access layer switches.

Another feature requirement for some access layer switches is PoE. PoE dramatically increases the overall price of the switch across all Cisco Catalyst switch product lines, so it should only be considered when voice convergence is required or wireless access points are being implemented, and power is difficult or expensive to run to the desired location.

Link aggregation is another feature that is common to most access layer switches. Link aggregation allows the switch to use multiple links simultaneously. Access layer switches take advantage of link aggregation when aggregating bandwidth up to distribution layer switches.

In a converged network supporting voice, video and data network traffic, access layer switches need to support QoS to maintain the prioritization of traffic. Cisco IP phones are types of equipment that are found at the access layer.

Distribution layer switches have a very important role on the network. They collect the data from all the access layer switches and forward it to the core layer switches.

Traffic that is generated at Layer 2 on a switched network needs to be managed, or segmented into VLANs, so it does not needlessly consume bandwidth throughout the network. Distribution layer switches provides the inter-VLAN routing functions so that one VLAN can communicate with another on the network. This routing typically takes place at the distribution layer because distribution layer switches have higher processing capabilities than the access layer switches. Because inter-VLAN routing is performed at the distribution layer, the switches at this layer need to support Layer 3 functions.

Another reason why Layer 3 functionality is required for distribution layer switches is because of the advanced security policies that can be applied to network traffic. Access lists are used to control how traffic flows through the network. An Access Control List (ACL) allows the switch to prevent certain types of traffic and permit others. ACLs also allow you to control which network devices can communicate on the network. Using ACLs is processing-intensive because the switch needs to inspect every packet and see if it matches one of the ACL rules defined on the switch. This inspection is performed at the distribution layer, because the switches at this layer typically have the processing capability to handle the additional load, and it also simplifies the use of ACLs

The distribution layer switches also need to support QoS to maintain the prioritization of traffic coming from the access layer switches that have implemented QoS.  It is important that distribution switches support redundancy for adequate availability. Loss of a distribution layer switch could have significant impact on the rest of the network because all access layer traffic passes through the distribution layer switches. Distribution layer switches are typically implemented in pairs to ensure availability. It is also recommended that distribution layer switches support multiple, hot swappable power supplies.

Finally, distribution layer switches need to support link aggregation.

The core layer of a hierarchical topology is the high-speed backbone of the network and requires switches that can handle very high forwarding rates.

The core layer also needs to support link aggregation to ensure adequate bandwidth coming into the core from the distribution layer switches. Core layer switches should have support for aggregated 10GbE connections, which is currently the fastest available Ethernet connectivity option.

The availability of the core layer is also critical, so you should build in as much redundancy as you can. Layer 3 redundancy typically has a faster convergence than Layer 2 redundancy in the event of hardware failure.

Core layer-capable switches have the ability to swap cooling fans without having to turn the switch off. QoS is an important part of the services provided by core layer switches