Network Management Strategies for the CIO
Requires Free Membership to View
When you write a book on network design, you have to come up with a process for designing switched networks, and that's where all the terminology comes in. The first time I saw the 3-layer hierarchy was in one of the Cisco textbooks by Priscilla Oppenheimer, although Cisco has a number of white papers on switched network design using the 3-layer model (which came first, the chicken or the egg?)
An edge switch is where the users connect to and uplinks to access switches.
An access switch is where the edge switches connect to and uplinks to core switches.
A core switch is the big momma in you data center.
So if you have a campus, you have edge switches on each floor, and access switches in the bottom floor which collapses the building backbone. The access switches then uplink to your core switch for campus access.
The edge switch should be cost-effective, stackable, and of moderate performance.
The access switch should be feature-rich (QoS, routing, access control, medium density), have medium performance and good flexibility. The access switch (according to the text books) is where your routing, QoS, filtering, accounting (etc etc) should be done. This way your solution scales as it grows.
The core switch should be fast. Oh yeah, and have enough ports to collapse your access switches.
In real life, three layers is overkill except for the largest of networks. Often, we only put an "access" switch and connect all of the edge switches to it.
This was first published in July 2002
Join the conversationComment
Share
Comments
Results
Contribute to the conversation