Network topologies and traffic characteristics

The four topology shapes, then how the traffic riding them behaves:

Diagram comparing bus, ring, star, and mesh topologies

The topologies

Bus

Every workstation connects to a common transmission medium: a single cable called the backbone or bus. All devices share the same medium, so only one device can transmit at a time, and a break in the backbone can bring down the whole segment.

Ring

Computers and other network devices are cabled in succession, and the last device connects back to the first to form a circle. Each device connects to exactly two neighbors and has no direct connection to any others; traffic travels around the ring to reach non-adjacent devices.

Star

The most common physical topology: a central device connects to all other network devices via point-to-point links. Also called hub-and-spoke.

There are no direct physical connections between spoke devices; all traffic passes through the central device.

An extended star replaces one or more spoke devices with a device that has its own spokes: multiple star topologies whose central devices are connected to each other.

Mesh

A device can be connected to more than one other device, so multiple paths exist between nodes. Redundant links increase reliability and let the network self-heal around a failed link.

A full mesh connects every device to every other device: the most redundant and the most expensive. A partial mesh gives only some devices redundant links, and sometimes shows up in WANs.

Topologies in practice

A physical star with a switch as the central device is by far the most common LAN implementation today. When a switch interconnects the devices, both the physical and logical topologies are stars.

Mesh is not typically used within a LAN. Partial mesh shows up in WANs, and full mesh is reserved for cases where the cost is justified by the redundancy.

Traffic characteristics

One-way requirements:

Quality of Service (QoS)

QoS is a way to identify different types of applications so that, when there’s a traffic jam on the network, certain types of traffic get expedited ahead of others.

Batch applications

FTP, TFTP, inventory updates. No direct human interaction; bandwidth is important but not critical.

Interactive applications

An inventory inquiry. Human-to-machine interaction; response time and bandwidth are important but not critical.

Real-time applications

VoIP and video. Human-to-human interaction, and end-to-end latency is critical.