Network Topologies Flashcards
Bus Topology
A single line of devices connected together by one shared network cable. Computers connect to the network by physically tapping into the network cable using special adapters.
This leads to an important requirement of bus networks: the ends of the cable must be properly terminated. At both ends of the network cable, special electrical resistors called terminators must be attached to absorb stray electrical signals on the wire. If these terminators are removed or if the network cable is accidentally cut, the electrical signals will not be properly absorbed and will bounce back along the wire, causing communication loss, a condition called signal reflection.
Ring Topology
In this topology, the cable enters a “ring in” port on the network card of the computer and exits a “ring out” port on its way to the next computer in the ring. By definition, a ring is a closed-loop, and the ring topology is no exception. When building a ring topology, even if the computers are all physically arranged in a straight line, the network cable will always connect to itself. The cable exiting the “ring out” port on the last computer will be fed into the “ring in” port on the first computer, thereby closing the loop.
Ring topologies are generally more reliable than bus topologies. However, like a bus network, if the ring is broken, network communication will fail. In cases where the network must be highly available, a dual-ring topology can be used. In this case, there are two sets of cables, and each computer has two network cards, one for each ring. In the event of a single cable break, the second ring can take over, allowing network traffic to continue to flow. However, the dual-ring topology has its greatest benefit in the event that both rings are simultaneously cut. In this case, the two loose ends on either side of the cable break can be connected together, merging the two broken rings into one much larger, but continuous, ring, where traffic can flow.
This topology was created to combat one of the more challenging aspects of the bus network: traffic collisions. When traffic collisions occur, all traffic must pause and wait for the line to clear before anyone can send again. This creates delays and degrades the performance of the network.
Star Topology
In this topology, also known as a hub-and-spoke network, is an improvement upon the bus topology previously described. Unlike the single straight line of the bus topology, a star network is composed of a central network device, such as an Ethernet switch, connected to various network devices, such as servers, computers, and printers, by individual network cables.
Each device is only connected to the central switch. All device-to-device communication is sent through the switch at the center of the network and then forwarded by the switch to the proper destination.
One key advantage of this topology is versatility. Instead of running cables from computer to computer, in a star topology, network cables are often run in the walls to a central closet. In fact, the owner of the building may decide to run many more network cables than are actually needed to anticipate future growth. This gives the network administrator the flexibility to move computers around the office without re-cabling the network. The owner must only connect the computer to the network outlet on the wall and then patch in the corresponding network cable to the central switch. This also means that a break in a single network cable will only impact one network device rather than the entire network. However, if the central switch fails, the network fails, resulting in loss of communication for all the devices in that network.
Today, star networks are the most common type of network found in local area network (LAN) environments.
Mesh Topology
Mesh topologies are often drawn as a web of direct connections between computers or nodes in a network. However, those connections may be permanent or constructed dynamically, as nodes need to talk to other nodes. A mesh topology permits nodes to communicate with each other; the topology may be either a full mesh, where every node has access to all other nodes, or a partial mesh network, where each node is only able to connect to a subset of the other nodes.
Mesh topologies are not necessarily constructed using physical network cables. The nodes may connect using Wi-Fi or radio signals or by virtual links such as virtual private networks (VPNs). Another example of a mesh network is a collection of routers that are able to communicate with each other and learn the best path for traffic to take when passing from node to node in the mesh.
Mesh networks are typically used where communication within a network must be highly available and redundancy is needed. The nodes within a mesh network can communicate with each other, and these connections can be changed dynamically if one node were to fail. This behavior is often referred to as a self-healing network because the nodes in the mesh are aware of each other and can establish new connections around failed nodes as needed. Common use cases include wireless networks at home and in the office, as well as large collections of routers, such as on the internet.
