For technicians that are beginning their network careers today, there is often a struggle to learn things that have just always automagically worked for them. One example of this is the crossover cable. Most modern network devices have a function called auto-mdix which will allow the incorrect use of straight through and crossover cables. As a result no one really has to learn the correct use. The exception is, of course, the CCNA exam.
For the CCNA exams, it is important to understand which cable is appropriate, a crossover or straight through cable. So how does one know which cable to use? The answer is quite simple, but is often confused and made complicated. The confusion is a result of people trying to memorize all of the scenarios for which a crossover is used. That is the wrong approach.
To easily understand where to use a crossover cable, we need to examine a couple of key concepts. The first concept is that there are Ethernet NICs that are wired like a hub or switch and there are NICs that are wired like a PC. So we basically have these two categories. Any time that we are connecting a device from one category to a device in the other category, we use a standard straight through cable. If we ever need to connect two devices that are in the same category, we should be thinking about a crossover cable.
Key Concept–Two devices in the same category use a crossover cable
Looking at our two categories, our first category is hubs and switches. Is there anything else? Well I guess someone could argue that there are routers with ESW (ethernet switch modules or cards), but their name pretty much tells us that they are a switch that is inserted into the router. As such, these specific cards act like a switch. So the simple and correct answer is that this first category only includes switches (and hubs if you have any of those left).
What is in the second category? This is really easy–everything else. To be complete, this includes the following: routers (except ESW ports), print servers, PCs, access-points, servers, and any other Ethernet connected device. If it isn’t a switch or a hub it goes into category two.
Now we can go back to the previous concept and apply it to these categories. This concept stated that when connecting two devices in the same category, a crossover cable was appropriate. So let’s test our knowledge.
For each of the connections, what is the appropriate cable type?
Hub<—>Router (answer: straight through–hub in category 1, router in category 2)
Hub<—>Hub (answer: crossover–both devices in category 1)
PC<—>Print Server (answer: crossover–both devices in category 2)
PC<—>Router (answer: crossover–both devices in category 2)
Switch<—>Hub (answer: crossover–both devices in category 1)
Switch<—>Switch (answer: crossover–both devices in category 1)
For those who want to take a look at the pinout for a crossover cable there is this and many other examples on the internet.
Understanding the use case for a crossover cable is not that difficult. What is difficult is trying to memorize all of the use cases. I hate memorizing, so I always recommend simply understanding that we have two categories of Ethernet devices. A crossover cable is only needed when we are interconnecting two devices in the same category. It is also worth mentioning that in practice auto-mdix has been available for many years and makes this discussion a non issue in most real world cases.