Let’s compare bridging vs. switching….
To begin our exciting journey into the world of network switches and switching on computer networks, we have to first go back into really recent history. Now remember everything in I/T advances so fast that when you talk about “history” in most cases you’re only talking about a few years ago, anyway.
So we’re going to focus on bridging vs. switching in this short post.
To understand bridging, we have to go back a little bit in the history of I/T to why bridges were developed and used. And to understand that, we have to show how hubs originated and were used. I’m going to go over that history really quickly and really fast here so we can immediately get to where we are NOW with technology.
Bridging vs Switching – The Hard Truth
Now, if you haven’t watched it, you’re welcome to check out this post on Hubs and how they actually work (and why you don’t really find them used on networks in the present day).
Real quick: In the early days of computer networks, you would have what were called “Mainframe” computers. You would then have “Controllers” connected to these mainframes and off of those controllers you would have “Dumb Terminals” These were where the end-user(s) would access the mainframe and everything on it.
Now it’s important to understand these mainframe setups because at the time, to allow remote sites to connect to the mainframe computer, they would need Bridges.
And it’s important here that you get that this is where the network bridge was originally developed and used.
I’ll come back to bridges shortly…
Eventually, that mainframe setup morphed into what we now refer to as a “collapsed backbone” network like the one shown here. It was called a “collapsed backbone” network because hubs were developed and used to connect PCs (which were quickly on the rise in fame at the time) directly to the corporate backbone to reach any network servers or other devices on the network.
Again, if you watch the video I created on Hubs, you’ll see why they couldn’t keep up with the demands for use that we humans require of computers and networks – especially by today’s standards. But hubs were used at the time because they were the only thing available for connecting multiple PCs to a computer network at the same time.
There’s also a lot of history on how the servers on those early collapsed backbone networks ran off of Token Ring connections and didn’t function like Netware servers do. And I may create a video talking more about the history of those technologies, too, in the near future.
Suffice it to say you don’t need to know all of this history to pass the Network+ Certification Exam…at least not yet.
The key thing to keep in mind about these corporate collapsed backbone networks is that they grew and grew and grew and as they did, network services became slower and slower.
There are multiple reasons this happened. The biggest of those was LAN services needing faster service, software demands from new and innovative software applications needed more bandwidth and usage by themselves and networks were becoming saturated and bogged down.
This is where bridges come back on the scene. Bridges were used to break up collision domains to help handle the saturation occurring on these networks.
However, there were some limitations:
Bridges have a limited number of physical ports for use on their physical hardware
Bridges could not provide the network services needed to handle everything
Layer 2 Switches to the Rescue
Layer 2 switches helped save the day by breaking up collision domains on the network into a separate collision domain for each port on the switch.
The real benefit to Layer 2 switches arriving on the scene was that they could provide hundreds of ports – instead of the limited number of ports available on bridges.
So those early networks began to morph into something like this…
You would have hubs connected on each port of a switch to allow more devices (typically PCs) to connect in each area of the network.
And that’s the first and primary way that Layer 2 Switches started being used to solve the problem of ever-larger growing networks and saturation of those networks.
Now this leads us to the other big reason why Layer 2 Switches replaced Bridges on computer networks.
Bridges were designed to use software to create and manage a filter table in their own memory for making decisions about network traffic that arrived on their physical ports.
Layer 2 Switches on the other hand, use ASICs (Application Specific Integrated Circuits) to create and manage their filter tables.
Now, I’ll stop right here and remind you that it’s still ok to think of Layer 2 Switches as multi-port Bridges. That’s because they both break up collision domains. But the efficiency of Layer 2 Switches compared to Bridges and the fact that Switches offer a lot more physical ports for connecting devices than Bridges do shows why Switches won out over Bridges, and still do to this day.
I also want to inject here that Layer 2 Switches and Bridges are much faster than Routers (when we get into those in later videos), because Switches and Bridges don’t have to examine the Network layer header information at the beginning of a packet inside a frame. They simply look at the frame header Source and Destination MAC Addresses and make decisions to either forward the entire frame or drop it based on that information.
Last (but not least) is the fact that both Bridges and Switches could use the loop prevention protocol known as Spanning Tree – but the real key here is that only 1 instance of Spanning Tree can be used on a Bridge, whereas multiple instances of Spanning Tree Protocol (STP) can be used on a Layer 2 Switch.
I’ll get more into detail about Spanning Tree as we go along. In fact, if you haven’t watched this video yet, I’d invite you to watch it where it talks about what Spanning Tree is and why you have to have a loop-prevention protocol like Spanning Tree to prevent your entire network from locking up and no longer passing any traffic.
So that’s the gritty details of Bridging vs. Switching on computer networks. Here are the most important details to keep in mind for when you take your Network+ Certification Exam (and don’t hold me to these when you get into your own exam, as you may get multiple questions about this or you may get no questions at all):
Remember that Bridges are software-based, while switches are hardware-based (they use ASIC chips to make their forwarding decisions).
Also remember that a switch can really be thought of as a more capable multi-port Bridge. You can have a lot more physical port connections on a switch than you ever did with a Bridge.
Bridges can only run 1 instance of Spanning Tree Protocol, while Switches can typically run multiple instances of Spanning Tree Protocol.
Bridges and Switches forward Layer 2 broadcasts (and I’ll get more into broadcasts in future videos).
Bridges and Switches learn MAC addresses from the source address on each frame that arrives on an inbound port.
And really last but not least is that fact that Bridges and Switches both make forwarding decisions based on Layer 2 addresses.