Network Topologies


The topology of a network refers to how the physical layout is organized ”that is, how the nodes on the network are physically wired together. Some topologies are cheaper to implement than others; some require more maintenance; and some have single points of failure, making them less reliable and robust. When designing a network, it is important to consider how big the network is going to be, the levels of traffic, and whether any failover requirements exist. Each of these aspects influences the decision of which topology to use. The following sections describe some of the more common topologies.

Bus Topology

In a bus topology, hosts are connected to a single cable ”also known as a backbone ”that runs through an area. Each end of the bus is terminated to avoid data signals being reflected back along the network. This type of configuration makes maintenance difficult because the network cabling must be disconnected to add a new connection. The bus itself is also a single point of failure ”a fault on the bus renders the whole network unusable ”but it is a cheap option and easy to install for smaller networks. The bus topology is normally used in broadcast networks, allowing only one host to transmit at a time. Figure 1.1 shows an example of a typical bus configuration.

Figure 1.1. A bus configuration.

graphics/01fig01.gif

Star Topology

The star topology is the most common topology, mainly because it is cheap to install and easy to maintain. The hosts are connected to a central hub , which enables new hosts to be added without the need to interrupt the network's functionality. The number of hosts, or nodes, that can be used in a star network is limited by the number of connections on the central hub, although multiple hubs can be connected together to form larger networks. As with the bus configuration, a single point of failure, the hub itself, does exist. Figure 1.2 shows an example of a simple star network.

Figure 1.2. A star configuration.

graphics/01fig02.gif

Ring Topology

The ring network is a configuration in which the hosts are connected in a loop. Each host is directly connected to its two adjacent hosts, one on either side. This topology offers slightly better resilience than bus or star topologies because a single failure results only in a degradation of the network; it takes two cable breaks, in separate segments, before any hosts become disconnected from the network. Figure 1.3 shows an example of a simple ring configuration.

Figure 1.3. A ring configuration.

graphics/01fig03.gif

Two rings can also be used to provide added resilience. The Fiber Distributed Data Interface (FDDI) is the most common example of this, in which the second ring continues to function if the primary ring fails.

Mixed Topologies

Networks commonly use more than one topology because it is more practical and cost effective; this is called a mixed topology. For example, a department might use a bus configuration as its main backbone but have multiple star networks connecting the hosts in each section. In some colleges and universities, dual-ring networks, such as FDDI, are used as the backbone. Figure 1.4 shows how a mixed configuration may be used.

Figure 1.4. An example of a mixed configuration.

graphics/01fig04.gif

Virtual LAN Topologies

Virtual local area networks (VLANs) have become more popular, but they require the use of an intelligent central switch or hub. VLANs are similar in functionality to having multiple star networks, but in this case, they all use the same physical switch as the central hub. A VLAN is created by assigning specific ports on the central hub to separate networks. The hosts on each VLAN communicate with each other as if they were on the same physical network segment. An advantage of the VLAN concept is that the network traffic on one VLAN is separate from the network traffic on another VLAN, making more efficient use of the available bandwidth. A further advantage is that hosts can be moved between VLANs without requiring any hardware configuration. Of course, the central switch, or hub, remains a single point of failure. Figure 1.5 demonstrates how an intelligent hub can be used to create two separate VLANs, (the hosts are shaded and labeled to show to which network they belong).

Figure 1.5. An example of a VLAN configuration.

graphics/01fig05.gif



Solaris 9 Network Administration Exam Cram 2 (Exam Cram CX-310-044)
Solaris 9 Network Administrator Exam Cram 2 (Exam CX-310-044)
ISBN: 0789728702
EAN: 2147483647
Year: 2003
Pages: 174
Authors: John Philcox

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net