Logical and Physical Network Design

When referring to a 'logical' network we are talking about the logical addressing used to describe the network itself or the networks it connects to.

A logical network layout clearly shows the IP Addresses associated with each part of the network. In most cases, the logical network is a simple Class C network such as 192.168.0.0 with the default subnetmask of 255.255.255.0. This network allows up to 254 hosts to be connected directly to it without the need of any routing.

One of the most important steps when designing a logical network is the future growth and expansion, a point that most people tend to overlook and face problems such as running out of available IP Addresses. When I design a network, I always leave room for an addition 50-60% growth of its current size. In practice, this means that if you asked me to design a network with 60 workstations, I'll design it with at least 100 workstations, giving you 40 extra IP Addresses. Of course, if we are talking about one or two networks at most, then you simply assign as I previously said, one full Class C network.

In addition, it's generally a good practice to leave the first 10 or 20 IP Addresses for critical network devices/hosts, such as gateways, routers, switches, servers and printers, and hand out the rest of the remaining IP's to the users.

Coming to the second part of your question, the Physical design, it refers to the actual layout of the physical part of our network. This includes the cables, switches, workstations etc. A physical layout/map usually involves a diagram of the actual floor the way it would be seen if you were on the ceiling, looking down towards to the floor.

The classical physical design of most networks involve a central rack located in a computer room or a restricted access room (you can check my rack out if you like! Where all wiring from the walls terminate and connect at the back of a patch panel. From the front side of the patch panel,