Quick Summary
- A network topology is the arrangement of nodes and links -- the physical topology is how cables are laid out, the logical topology is how data actually flows
- The main types are star, bus, ring, mesh, and tree (hierarchical), plus hybrid mixtures of these
- The trade-offs come down to cost, scalability, fault tolerance, and how many single points of failure a layout has -- star dominates modern LANs, mesh is the most resilient
The Floor Plan Analogy
Designing a network is a lot like designing the floor plan of a building. You have to decide how the rooms (the devices) connect to each other. Do you run a cable from every room back to one central wiring closet? Do you string all the rooms along a single hallway they all share? Or do you build so many connecting corridors that there is always a backup route from any room to any other?
Each of those choices is a network topology -- the pattern of how nodes (computers, switches, routers) and links (the cables or wireless connections between them) are arranged. There is no single "best" floor plan. Each one trades off cost, scalability, fault tolerance, and single points of failure differently.
One more important idea: topology comes in two flavors. The physical topology is how the cables and devices are literally laid out. The logical topology is the path data actually takes through the network. They are not always the same -- you can have a physical star that behaves logically like something else, which is why engineers describe them separately.
The Main Topologies at a Glance
Here are the four classic shapes side by side. Notice how the links connect the nodes differently in each:
Try It Yourself: Topology Explorer & Failure Simulator
Pick a topology, then click a link in the diagram (or use the button) to cut it and see the impact for that layout. Reset restores everything.
Play with the simulator above and a pattern emerges: star and mesh contain a failure well in different ways, while bus and a basic ring are far more fragile to a single break. That trade-off -- resilience versus cost and simplicity -- is the heart of topology design.
The Topologies, One by One
Star — The Modern Default
In a star, every device connects to a central switch or hub. This is the dominant layout in modern LANs because it is easy to manage and a single cable failure affects only that one device. The catch: the central switch is a single point of failure -- if it dies, the whole segment goes down. The fix in practice is to make the switch itself redundant.
Bus — The Legacy Backbone
A bus connects all devices to one shared backbone cable. It is cheap and was used in early Ethernet such as 10BASE2. The downside is severe: a break anywhere in the backbone, or collisions on the shared medium, can take down the entire segment. The bus has largely disappeared from modern wiring for exactly this reason.
Ring — A Loop of Neighbors
In a ring, each node connects to exactly two neighbors, forming a loop, and data passes around the ring (think Token Ring and FDDI). A single break can disrupt the whole ring -- unless it is a dual or redundant ring, which can survive one break by sending traffic the other way around.
Mesh, Tree & Hybrid — Scaling Up
A mesh interconnects devices with many redundant paths; a full mesh
links every node to every other with n(n-1)/2 links, while a partial mesh
connects only some. It is very resilient but expensive -- used in WAN cores and the internet backbone.
A tree (hierarchical) layout is a star-of-stars in access, distribution, and core tiers,
the standard for campus design. Most real networks are hybrid -- a mix of all of these.
Topologies in the Real World
Different scales and goals call for different shapes. Here are three you will actually meet:
Your Home / Office LAN (Star)
Plug every laptop, printer, and access point into one switch or router and you have built a star. It is the default everywhere because adding or moving a device only touches one cable, and one bad cable bothers only one device. Just remember the switch is the single point of failure.
A Campus Network (Tree)
A university or large office uses a tree / hierarchical design: access switches in each room feed distribution switches per building, which feed a core. This star-of-stars scales cleanly and keeps local traffic local -- the standard model for enterprise and campus networks.
The Internet Backbone (Mesh)
Carrier and WAN cores -- and the internet backbone itself -- use a partial mesh. With many redundant paths between routers, traffic reroutes around a failed link or node, typically within seconds. It is expensive and complex, but for the backbone, resilience is worth every dollar.
Common Mistakes
A few misconceptions about topology cause real design errors. Watch out for these:
Confusing Physical with Logical
The physical topology (how cables run) and the logical topology (how data flows) are not always the same. A physical star wired to a switch can carry flows that differ from its shape. Assuming the cable layout tells you the whole story leads to surprises when you troubleshoot how traffic actually moves.
Ignoring the Single Point of Failure
A star is easy and cheap, but the central switch is a single point of failure. Building a critical network as one flat star with no redundant switch or uplink means one device failure takes everything down. Plan redundancy where uptime matters instead of discovering the gap during an outage.
Over-Building a Full Mesh
A full mesh needs n(n-1)/2 links, so cost and cabling explode as you add
nodes -- 10 nodes already need 45 links. Meshing everything "for safety" wastes money and complexity. Use
a partial mesh only where you actually need the redundant paths, and lean on tree or star elsewhere.
Physical vs. Logical Topology, Revisited
It is worth dwelling on this distinction because it confuses so many beginners. The physical topology answers "where do the cables go?" -- for example, every PC has a cable running back to one switch, which is physically a star. The logical topology answers "what path does the data actually take?" The same physical star can implement different logical flows depending on how the devices and protocols are configured. When you read a network diagram, always ask which view you are looking at: the wiring or the data flow.
How to Choose a Topology
Every topology choice is a balance of four forces: cost (cabling, hardware, and labor), scalability (how gracefully it grows), fault tolerance (how well it survives a failure), and the number of single points of failure. A small office optimizes for cost and simplicity, so it picks a star. A campus optimizes for scale, so it picks a tree of stars. A carrier backbone optimizes for fault tolerance, so it pays for a mesh. And because no single shape is perfect, almost every real-world network ends up as a hybrid -- stars at the edge, a tree in the middle, and mesh in the core.
The bottom line: a network topology is just the floor plan of how nodes and links are arranged. Learn the strengths and weaknesses of star, bus, ring, mesh, and tree -- especially where each one hides a single point of failure -- and you can read, design, and troubleshoot almost any network.
Frequently asked questions about network topologies
What is a network topology?
A network topology is the arrangement of nodes and the links between them. The physical topology describes how cables and devices are physically laid out, while the logical topology describes how data actually flows -- and the two can differ.
What are the main types of network topology?
The main types are star, bus, ring, mesh, and tree (hierarchical), plus hybrid combinations. Star connects every device to a central switch, bus shares one backbone, ring links each node to two neighbors in a loop, mesh adds redundant paths, and tree layers stars into access, distribution, and core tiers.
What is the difference between a star and a mesh topology?
In a star, every device connects to one central switch, so it is cheap and easy to manage but the central device is a single point of failure. In a mesh, devices interconnect with many redundant paths, so traffic can reroute around a failure -- but it costs more and is more complex to wire and maintain.
What is the difference between physical and logical topology?
Physical topology is how cables and devices are physically connected, such as every PC running back to a central switch. Logical topology is the path data actually takes through the network. The same physical star can carry different logical flows, which is why the two views are described separately.
Which topology is most fault tolerant?
Mesh is the most fault tolerant, because its many redundant paths let traffic reroute around a failed link or node while keeping the network connected. A full mesh connects every node to every other with n(n-1)/2 links, which is why it is reserved for WAN cores and the internet backbone.
Try the networking lab
You just learned how star, bus, ring, and mesh layouts handle a failure. Now cut cables in a star and a ring of switches, trace again after each cut, and see which PCs stay connected.
Launch the network lab →