Introduction to LAN switching
Welcome to the world of LANS, in the previous couple of chapters, we learnt about routers and how they are used in the networking environments. In the next couple of chapters, we will learn about LANs and the main device we will use will be the SWITCH. The concepts we will learn will also tie in with routing concepts.
In today’s business environment, businesses need information to survive. With technology, this has been made possible, the use of new methods of communicating such as the use of voice, video, data which is transmitted over networks is crucial. As such, we need to design LANs with these needs in mind. In this chapter, we will discuss some of the considerations to make while we design the LAN. We will look at the hierarchical LAN model and its benefits, some design considerations as well as the benefits of well-designed LANs. This chapter is meant to introduce you to the world of LANs.
LAN design concepts
CISCO not only designs and produces network equipment, but also they focus on developing the most optimal way to use their devices, as such when designing a LAN network, they recommend that a hierarchical model. In this type of architecture, there are a few things that have to be observed:
- Network segmentation and broadcast traffic management – this is mainly through the use of VLANs
- Easy configuration and management of the switches
These concepts will be explored in more detail as we explore LAN design.
Hierarchical layered model in LAN design
As mentioned earlier, the design of a LAN network is critical to communication within the enterprise, when using the hierarchical model as recommended by CISCO, there are three layers that we should implement depending on the size of the organization.
- Core layer
- Distribution layer
- Access layer.
The figure below shows how the implementation of this hierarchy can be achieved.
Starting from the bottom, we have the access layer. This is the layer that connects to end user devices such as PCs, printers, IP phones among others.
The distribution layer, is meant to aggregate the data from the access layer. This layer controls the traffic in the lower levels and prioritizes traffic based on organizational policies that have been implemented during configuration of the switches. Typically, this level should be redundant and made up of faster switches than the access layer.
The core layer, is responsible for high-speed switching in the network. Typically, this layer should consist of the fastest switches in the network and offer the highest bandwidth since communication to other networks from the lower levels is forwarded through these switches.
Benefits of a hierarchical model
- Scalability – when you implement a network a hierarchical network model, expansion is simplified since all the roles are well defined. For example, if you have 5 access layer switches, connected to 2 distribution layer switches, you can add the access layer switches until all the ports on the distribution switches are filled up.
- Redundancy – this is achieved when the switches in each layer are connected to two or more devices at another level. If one device at the higher level in the hierarchy fails, the lower level switch automatically fails over to the other switch. Redundancy is achieved at the distribution and core layers.
- Performance – it is recommended that core layer switches should have very fast switching abilities. The distribution switches should also be very fast and redundant. The result of using very fast core and distribution layer switches would guarantee very fast networks.
- Security – the security of the network is enhanced since at each layer of the model, there are several security measures that can be put in place; for example switch ports at the access layer can be configured with port security, segmentation of the distribution layer using VLANs is also another security feature.
- Manageability is the ability to make configuration changes in the network, the use of the hierarchical model eases management of the switches. For example, making changes on one layer would be simplified since we can assume that the role of switches in that layer all perform similar functions, further, the modular design means that management does not mean that the network is down due to maintenance due to redundancy.
Considerations when choosing a switch
- When deciding the switch we should implement for our LANs, there are several considerations that we need to take in mind. These might be influenced by the organizational policies while others might be influenced by the technological needs.
- Switches with fixed configurations are switches that cannot be modified by adding additional modules, these are lower level switches and are ideal for the access layer functions.
- For more flexibility, we might need modular switches, these switches typically allow us to install modules such as more switching ports, these would be ideal for rapidly expanding networks that need to be changed frequently.
- To provide high bandwidth, we may need to interconnect special types of switches which have a stackable ability using a backplane cable. These would be ideal for high bandwidth requirements in a large network at the core layer.
- Port density – this is the number of ports on a switch. In many cases you will find switches with 24 or 48 port switches. This can be a design consideration since you may need to consider the inter-switch connections.
- Forwarding rates are the processing capabilities of the switch. The forwarding rate is measured by calculating how much data the switch can process in a second. This is different from the bandwidth that is available on its ports.
- In most modern networks, the use of IP phones is prevalent, most of these devices get power over the LAN interfaces connected to switches using a technology called POE (Power over Ethernet). As such, when deciding which devices to buy, PoE should be a feature that should not be overlooked.
- In recent times, switch designs have been changed so as to support layer 3 functionality, as you may already know, switches work at layer 2 of the OSI model, however, implementing layer 3 switches gives more options such as routing, IP addressing and other options.
Access layer switch features
There are several features that a switch at each level of the hierarchical model should have. As we mentioned earlier, the access layer is the lowest level in the hierarchical LAN architecture, at this level user devices gain access to the network over a number of devices. As such, the features at this level include: VLAN support on the switches, Fast Ethernet and Gigabit Ethernet links, PoE and support for link aggregation so as to increase the switching speed.
Security is important in our networks, at this layer, we can implemnent several security measures such as port security to control access to the network.
CISCO recommends that VLANs be localized to a switch, the switches at this level should have support for VLANs for a variety of purposes.
Link aggregation is the ability to use multiple links at the same time. This is a more effective way to use the bandwidth available on the switches.
To support multiple devices on a single port, PoE is an important feature, it allows us to use the switch to power certain devices in our network such as IP phones and Wireless controllers.
The ports on access layer switches should be fast enough to support the evolving bandwidth needs of the enterprise. As such, Fast Ethernet which offer speeds of up to 100Mbps and Gigabit Ethernet links which offer speeds of up to 1Gbps should be used.
Distribution layer features
At the distribution layer, communication across the various access layer switches should be supported, this means that these switches should offer more features than the access layer switches. Features such as redundancy, faster ports than the access layer, layer 3 support should be implemented at this layer.
- The use of security policies is a security feature that should be implemented at the distribution layer, some of these may include the use of access lists.
- Inter-vlan routing which is making communication between different VLANs possible should be available at this layer.
- The ports at this layer should be very fast, typically, Gigabit Ethernet and 10 gigabit Ethernet links should be used. These ports should be aggregated and redundancy should be implemented between the switches.
- At this layer, we need to prioritize the traffic from our access layer, as such, QOS (Quality of Service) mechanisms should be implemented.
NOTE: at the distribution layer, the use of layer 3 capable switches is highly recommended so as to support most of the features mentioned above.
Core layer features
The core layer of the network is the main link between our internetwork and other networks such as external networks. At this layer of the hierarchical model, there should be very fast switching, security policies, redundancy, layer 3 functionality and quality of service. In some organizations, the core layer may not be needed if the network is small.
- At the core layer, we should have very fast switches, typically operating at 10 gigabit speeds and above. This is to support the requirements of all the access and distribution layer switches.
- At this level, the use of security policies to control access should be implemented. This means that the switches at this layer should have layer 3 support.
- The core layer is sometimes implemented as the gateway to external networks and therefore redundancy is also an important element.
In the forthcoming chapters, we will discuss some of these concepts in detail through the networks that we will design and implement. The concepts in this chapter are meant to give you a firm foundation on the LAN architecture as recommended by CISCO. For more on this, you should conduct more research to discover best practices when it comes to designing and implementing LAN networks.