For a long time all the big projects related to the network, whether it is web-project or a large enterprise DC represented the same structure. It was characteristic of the tree architecture, distinguished only by the size and density of the tree "branches" due to the different requirements of reliability and performance. But the digital world is not standing still, and the rapidly growing and developing, with it not only in increasing the volume and speed, but also by changing its structure. Various Big Data, clouds and distributed calculations led to the fact that the network has become necessary to transfer large amounts of data between a large number of end nodes, and, preferably, with minimum delay.
All this has led to the fact that the traditional tree-like architecture consisting of levels of access, aggregation and core, was openly slip and falter. There is a need to replace it. What?
For a start, let's try to characterize the network structure in the so-called "traditional» Enterprise-projects:
But the same characteristics for systems of modern data centers, working with projects such as Web 2.0 cloud, Big Data, distributed computing and similar modern large-scale projects: