I am trying to understand the three-tiered campus network design. I know that only big companies use this design, others usually use the collapsed-core. My question is a little bit dumb, where do we put all these things and the way we connect them?
Say for example, one big bank corporation, got one headquarter, and many other big building and also many small branches. Do we usually put the core switch on a separate data centre or in the headquarter or what? If that's the case (put core in data centre), then isn't the distribution switch that I assume we put in the other big building and small branches will have to be connected to the switch through at least two routers(one from the distribution site, one from the core site) in between, which means the connection is WAN? Am I correct or wrong? Can someone give me a light on this?
How about the internet, does internet traffic has to go through the core switch before going out? Design can be different for many people, but I am just hoping you can give me what's common.
I think this document would shed a lot of light on the topic for you....
With that said, the "campus design" is not all encompassing, you have SONA (which can be reviewed here: http://www.cisco.com/en/US/netsol/ns629/index.html ) that ties in all of the different pieces.
You have different designs for different "modules" or "pods" of your network, and then you have the overall architecture guides that explain how to tie in those various pods to form a resilient, available network.
As with all Cisco designs, they build in a lot of redundancy, because everyone wants the Five Nines - 99.999% availability. However, when the budget finally gets drawn up, that is usually a matter of debate - do we need that? That is where business justification comes into play.