Looking Back at Traditional

Looking Back at Traditional
Campus Networks
In the 1990s, the traditional campus network started as one LAN and grew and grew until segmentation
needed to take place just to keep the network up and running. In this era of rapid
expansion, response time was secondary to just making sure the network was functioning.
Besides, the majority of applications were store-and-forward, such as e-mail, and there was little
need for advanced quality of service options.
By looking at the technology, you can see why keeping the network running was such a challenge.
Typical campus networks ran on 10BaseT or 10Base2 (thinnet). As a result, the network
was one large collision domain—not to mention even one large broadcast domain. Despite these
limitations, Ethernet was used because it was scalable, effective, and somewhat inexpensive
compared to other options. (IBM “owned” Token Ring, and getting it installed frequently
meant getting in IBM to do it—sometimes expensive and often impractical.) ARCnet was used
in some networks, but Ethernet and ARCnet are not compatible, and the networks became two
separate entities. ARCnet soon became history. Token Ring became marginalized. Ethernet
became king.
382
Chapter 12 
The Campus Network
Because a campus network can easily span many buildings, bridges were used to connect the
buildings; this broke up the collision domains, but the network was still one large broadcast
domain. More and more users were attached to the hubs used in the network, and soon the performance
of the network was considered extremely slow.