PROJECT TITLE :
On Nonblocking Multicast Fat-Tree Data Center Networks with Server Redundancy
Fat-tree networks have been widely adopted as network topologies in information center networks (DCNs). But, it's expensive for fat-tree DCNs to support nonblocking multicast communication, due to the large variety of core switches required. Since multicast is a vital communication pattern in many cloud services and nonblocking multicast communication will guarantee the high performance of such services, reducing the cost of nonblocking multicast fat-tree DCNs is terribly vital. On the other hand, server redundancy is ubiquitous in nowadays’s knowledge centers to produce high availability of services. In this paper, we explore server redundancy in information centers to cut back the price of nonblocking multicast fat-tree knowledge center networks (DCNs). First, we gift a multirate network model that accurately describes the communication surroundings of the fat-tree DCNs. We have a tendency to then show that the sufficient condition on the amount of core switches needed for nonblocking multicast communication under the multirate model can be significantly reduced when the fat-tree DCNs are two-redundant, i.e., each server in the data center has exactly one redundant backup. We also study the overall redundant fat-tree DCNs where servers might have completely different numbers of redundant backups depending on the availability necessities of services they supply, and show that a higher redundancy level any reduces the cost of nonblocking multicast fat-tree DCNs. Then, to finish our analysis, we tend to contemplate a sensible faulty data center, where one or more active servers might fail at any time. We tend to offer a strategy to re-balance the active servers among edge switches after server failures so that the identical nonblocking condition still holds. Finally, we have a tendency to give a multicast routing algorithm with linear time complexity to configure multicast connections in fat-tree DCNs.
Did you like this research project?
To get this research project Guidelines, Training and Code... Click Here